NQF

Agenda Synopsis

Full Agenda

Measure Index

• Discharge to Community-Post Acute Care (PAC) Home Health Quality Reporting Program (Required under the IMPACT Act) (Public comments received:9; MUC ID: MUC15-523)
  
  
• Drug Regimen Review Conducted with Follow-Up for Identified Issues-Post Acute Care (PAC) Home Health Quality Reporting Program (Required under the IMPACT Act) (Public comments received:7; MUC ID: MUC15-1127)
  
  
• Falls risk composite process measure (Public comments received:5; MUC ID: MUC15-207)
  
  
• Improvement in Dyspnea in Patients with a Primary Diagnosis of Congestive Heart Failure, Chronic Obstructive Pulmonary Disease and/or Asthma (Public comments received:5; MUC ID: MUC15-235)
  
  
• Medicare Spending Per Beneficiary-Post Acute Care (PAC) Home Health Quality Reporting Program (Required under the IMPACT Act) (Public comments received:5; MUC ID: MUC15-1134)
  
  
• Potentially Preventable 30-Day Post-Discharge Readmission Measure for Home Health Quality Reporting Program (Required under the IMPACT Act) (Public comments received:11; MUC ID: MUC15-234)
  
  

• Hospice and Palliative Care Composite Process Measure (Public comments received:6; MUC ID: MUC15-231)
  
  
• Hospice Visits When Death Is Imminent (Public comments received:6; MUC ID: MUC15-227)
  
  

• Discharge to Community-Post Acute Care (PAC) Inpatient Rehabilitation Facility Quality Reporting Program (Required under the IMPACT Act) (Public comments received:7; MUC ID: MUC15-408)
  
  
• Drug Regimen Review Conducted with Follow-Up for Identified Issues-Post Acute Care (PAC) Inpatient Rehabilitation Facility Quality Reporting Program (Required under the IMPACT Act) (Public comments received:5; MUC ID: MUC15-1128)
  
  
• Medicare Spending per Beneficiary-Post Acute Care (PAC) Inpatient Rehabilitation Facility Quality Reporting Program (Required under the IMPACT Act) (Public comments received:4; MUC ID: MUC15-287)
  
  
• Potentially Preventable 30-Day Post-Discharge Readmission Measure for Inpatient Rehabilitation Facility Quality Reporting Program (Required under the IMPACT Act) (Public comments received:8; MUC ID: MUC15-496)
  
  
• Potentially Preventable Within Stay Readmission Measure for Inpatient Rehabilitation Facilities (Public comments received:6; MUC ID: MUC15-497)
  
  

• Compliance with Spontaneous Breathing Trial (SBT) (including Tracheostomy Collar Trial (TCT) or Continuous Positive Airway Pressure (CPAP) Breathing Trial)) by Day 2 of the LTCH Stay (Public comments received:1; MUC ID: MUC15-400)
  
  
• Discharge to Community-Post Acute Care (PAC) Long-Term Care Hospital Quality Reporting Program (Required under the IMPACT Act) (Public comments received:6; MUC ID: MUC15-414)
  
  
• Drug Regimen Review Conducted with Follow-Up for Identified Issues-Post Acute Care (PAC) Long-Term Care Hospital Quality Reporting Program (Required under the IMPACT Act) (Public comments received:2; MUC ID: MUC15-1129)
  
  
• Medicare Spending per Beneficiary-Post Acute Care (PAC) Long-Term Care Hospital Quality Reporting Program (Required under the IMPACT Act) (Public comments received:3; MUC ID: MUC15-289)
  
  
• Percent of Patients Who Received an Antipsychotic (AP) Medication (Public comments received:1; MUC ID: MUC15-530)
  
  
• Potentially Preventable 30-Day Post-Discharge Readmission Measure for Long-Term Care Hospital Quality Reporting Program (Required under the IMPACT Act) (Public comments received:5; MUC ID: MUC15-498)
  
  
• Ventilator Weaning (Liberation) Rate (Public comments received:1; MUC ID: MUC15-398)
  
  

• Application of IRF Functional Outcome Measure: Change in Mobility Score for Medical Rehabilitation Patients (NQF #2634) (Public comments received:9; MUC ID: MUC15-527)
  
  
• Application of IRF Functional Outcome Measure: Change in Self-Care Score for Medical Rehabilitation Patients (NQF #2633) (Public comments received:10; MUC ID: MUC15-236)
  
  
• Application of IRF Functional Outcome Measure: Discharge Mobility Score for Medical Rehabilitation Patients (NQF #2636) (Public comments received:9; MUC ID: MUC15-529)
  
  
• Application of IRF Functional Outcome Measure: Discharge Self-Care Score for Medical Rehabilitation Patients (NQF #2635) (Public comments received:9; MUC ID: MUC15-528)
  
  
• Discharge to Community-Post Acute Care (PAC) Skilled Nursing Facility Quality Reporting Program (Required under the IMPACT Act) (Public comments received:9; MUC ID: MUC15-462)
  
  
• Drug Regimen Review Conducted with Follow-Up for Identified Issues-Post Acute Care (PAC) Skilled Nursing Facility Quality Reporting Program (Required under the IMPACT Act) (Public comments received:9; MUC ID: MUC15-1130)
  
  
• Medicare Spending per Beneficiary-Post Acute Care (PAC) Skilled Nursing Facility Quality Reporting Program (Required under the IMPACT Act) (Public comments received:6; MUC ID: MUC15-291)
  
  
• Percent of Skilled Nursing Facility Residents Who Newly Received an Antipsychotic Medication (Public comments received:3; MUC ID: MUC15-1133)
  
  
• Percent of Skilled Nursing Facility Residents Who Self-Report Moderate to Severe Pain (Public comments received:2; MUC ID: MUC15-1131)
  
  
• Percent of Skilled Nursing Facility Residents Who Were Assessed and Appropriately Given the Influenza Vaccine (Public comments received:4; MUC ID: MUC15-1132)
  
  
• Potentially Preventable 30-Day Post-Discharge Readmission Measure for Skilled Nursing Facility Quality Reporting Program (Required under the IMPACT Act) (Public comments received:10; MUC ID: MUC15-495)
  
  

• Skilled Nursing Facility 30-Day Potentially Preventable Readmission Measure (SNFPPR) (required by PAMA) (Public comments received:8; MUC ID: MUC15-1048)
  
  

Full Measure Information

Measure Specifications

• NQF Number (if applicable):
  
  
• Description: This measure describes the risk-standardized rate of Medicare fee-for-service (FFS) patients/residents/persons who are discharged to the community, and do not have an unplanned (re)admission to an acute care hospital or LTCH in the 31 days following discharge to community, and remain alive during the 31 days following discharge to community.
  
  
• Numerator: This measure does not have a simple form for the numerator and denominator. The numerator is defined as the risk-adjusted estimate of the number of patients/residents/persons included in the measure who are discharged to the community, and do not have an unplanned (re)admission to an acute care hospital or LTCH on the day of discharge or in the 31 days following discharge to community, and remain alive during the 31 days following discharge to community. The numerator estimate includes risk adjustment for patient/resident/person characteristics, and a statistical estimate of the facility/agency effect beyond case mix.
  
  
• Denominator: The denominator is computed with the same model used for the numerator. It is the model developed using all non-excluded facility/agency stays/episodes in the national data. The measure includes all facility/agency stays/episodes in the measurement period that are observed in national Medicare FFS data and do not fall into an excluded category. For a particular facility/agency, the model is applied to the patient/resident/person population, but the facility/agency effect term is 0. In essence, it is the number of discharges to community that would be expected for that patient/resident/person population at the average facility/agency.
  
  
• Exclusions: i) Age under 18 years; (ii) Discharges to psychiatric hospital; (iii) Discharges against medical advice; (iv) Discharges to federal hospitals or disaster alternative care sites; (v) Patients not continuously enrolled in Part A FFS Medicare for the 12 months prior to the HHA admission date, and at least 31 days after HHA discharge date; (vi) Patients whose prior short-term acute-care stay was for non-surgical treatment of cancer (only applies to those patients whose HHA episode was preceded by an acute care discharge in the past 30 days); (vii) Discharges to hospice; (viii) HHA episodes stays that end in transfer to another HHA; (ix) HHA episodes with a missing risk adjustment authorization code.
  
  
• HHS NQS Priority: Making Care Safer, Patient and Family Engagement, Communication and Care Coordination
  
  
• HHS Data Source: Claims
  
  
• Measure Type: Outcome
  
  
• Steward: Centers for Medicare & Medicaid Services
  
  
• Endorsement Status: Never Submitted
  
  
• Is the measure specified as an electronic clinical quality measure? No
  
  

Preliminary Analysis of Measure

• Preliminary analysis result: Encourage Continued Development
  
  
• Preliminary analysis summary
  • Contribution to program measure set:A study of 66,510 Medicare beneficiaries during pre- and post-HH episodes, revealed that 64 percent of beneficiaries discharged from HH did not use any other Medicare-reimbursed acute or post-acute services in the 30 days following HH discharge. Significant numbers of patients were admitted to inpatient facilities (29 percent) and lesser numbers to skilled nursing facilities (7.6 percent), inpatient rehabilitation (1.5 percent) and home health (7.2 percent) or hospice (3.3 percent) within 30 days of HH discharge (Wolff et al., 2008). http://www.ncbi.nlm.nih.gov/pubmed/18953231The value of this measure is to gather information on use of health care services during an extended period (3O days) following an episode of care. It evaluates whether agencies have prepared patients/caregivers to care for themselves through obtaining the knowledge, resources and confidence after their care in home health episode. It evaluates whether agencies have provided the patient/caregiver with the “right” tools and whether agencies have adequately evaluated the patient/caregiver capacity to maintain their level of health and functioning.Over 70 percent of patients with COPD enrolled in the VA home care telehealth program had a significant reduction in the numbers of ED visits, hospital admissions and total exacerbations (Alrajab, Smith el al., 2012). This study suggests the value of telehealth in reducing hospitalizaton and ED visits following a home health episode of care.http://www.ncbi.nlm.nih.gov/pubmed/23082792
  • Impact on quality of care for patients:Keeping patients out of the institutional care following home health care as well as tracking the status of patients will potentially improve the transition of care for patients helping them stay in the community. Patients would benefit from home health agencies providing the resources and knowledge for patients to keep them in the community without a hospitalization or use of other institutional settings.
  
  
  
• Does the measure address a program goal or objective? Yes. This measure assesses use of services post-discharge from PAC settings and thus can be used as a reflection of the effectivenss and efficiency of services and impact on future service utlization.
  
  
• Is this a high-value measure? Yes. This is an outcome measure that evaluates the use of other health care resources 30 days after discharge from Home Health (cost/usemeasure). It evaluates whether the goals of home health have been achieved; that patient’s reach their highest level of independence and functioning, and have the knowledge and confidence to assume their own care after discharge from a home health agency.
  
  
• Does this measure fill a gap in the program measure set? Yes. There are currently no measures that evaluate the status of patients and use of health care resources within 30 days of discharge from home health
  
  
• Measure development status: Early Development
  
  
• Is the measure fully tested for the program's setting and level of analysis? No. Information on the testing of this measure was not provided.
  
  
• Since no, could the measure be revised to use in the setting or at level of analysis under consideration? Yes. This measure could be tested at the home health agency level of analysis via claims/administrative data.
  
  
• Is the measure currently in use? No.
  
  
• Does the MUC contribute to alignment and efficient use of measurement resources (burden and cost of measurement)? Yes. This measure captures information when the patient is discharged from home health and for 30-days following that discharge. It is being proposed to address the IMPACT Act requirements and will be used across all post-acute setting except hospice.
  
  
• Is the measure NQF endorsed for the program's setting and level of analysis? Never Submitted
  
  
• Does the measure address a high-priority quality issue in the dual eligible beneficiary population? Yes.
  
  

Rationale for measure provided by HHS
The ultimate goals of post-acute care are avoiding institutionalization and returning patients to their previous level of independence and functioning, with discharge to community being the primary goal for the majority of post-acute patients. For many, home is a symbol of independence, privacy, and competence. Discharge to community is considered a valuable outcome to measure because it is a multifaceted measure that captures the patient’s functional status, cognitive capacity, physical ability, and availability of social support at home. There is considerable variation in discharge to community rates within and across post-acute settings. Studies show geographic variation, variation across patient socioeconomic characteristics (for example, race and ethnicity), and variation by facility characteristics (for profit vs. nonprofit, freestanding vs. hospital-based, urban vs. rural). In the IRF setting, discharge to community rates vary across providers, ranging from about 60% to 75%. The 2015 MedPAC report shows that, in FY 2013, the facility-level, mean risk-adjusted discharge to community rate for IRFs within 100 days of admission was 75.8%, and the mean observed rate was 74.7%. Discharge to community rates also vary widely in the SNF setting, ranging from as low as 31% to as high as 65%. The 2015 MedPAC report shows a mean risk-adjusted discharge to community rate of 37.5% for SNFs within 100 days of admission, and mean observed rate of 40.1%. A multicenter study of 23 LTCHs reported that only 28.8% of 1,061 patients who were ventilator-dependent on admission were discharged to home or assisted living facility. A study of 66,510 Medicare beneficiaries during pre- and post-HH episodes, revealed that 64 percent of beneficiaries discharged from HH did not use any other Medicare-reimbursed acute or post-acute services in the 30 days following HH discharge. Significant numbers of patients were admitted to inpatient facilities (29 percent) and lesser numbers to skilled nursing facilities (7.6 percent), inpatient rehabilitation (1.5 percent) and home health (7.2 percent) or hospice (3.3 percent) within 30 days of HH discharge (Wolff et al., 2008).

Measure Specifications

• NQF Number (if applicable):
  
  
• Description: Percentage of stays Inpatient Rehabilitation Facility (IRF), Long Term Care Facility (LTCH), and Skilled Nursing Facility (SNF) or care episodes Home Health (HH) in which a drug regimen review was conducted at the Admission (IRF, LTCH or SNF)/ Start of Care (SOC)/ Resumption of Care (ROC) (HH) and timely follow-up with a physician occurred each time potential clinically significant medication issues were identified throughout the stay (IRF, LTCH, or SNF) or care episode (HH).
  
  
• Numerator: Number of stays or care episodes where the medical record contains documentation of a drug regimen review conducted at admission or start-of-care or resumption-of-care with all potential clinically significant medication issues identified during the course of care and followed-up with a physician or physician designee.
  
  
• Denominator: Care episodes or stays ending during the reporting period. Assessment timing is as follows: Beginning of care episode or stay: • HH – SOC or ROC • SNF – Admission • IRF - Admission • LTCH – Admission End of care episode or stay: • HH – Transfer, Discharge, or Death at Home • SNF – Discharge, or expired • IRF – Discharge, or expired • LTCH – Discharge, or expired
  
  
• Exclusions: Denominator Exclusion: NONE Numerator Exclusion: NONE
  
  
• HHS NQS Priority: Making Care Safer, Communication and Care Coordination
  
  
• HHS Data Source: Patient Assessment Data
  
  
• Measure Type: Process
  
  
• Steward: Centers for Medicare & Medicaid Services
  
  
• Endorsement Status: Never Submitted
  
  
• Is the measure specified as an electronic clinical quality measure? No
  
  

Preliminary Analysis of Measure

• Preliminary analysis result: Encourage Continued Development
  
  
• Preliminary analysis summary
  • Contribution to program measure set:This measure of medication reconciliation addresses an important aspect of care for patients as they transition from one setting to another. During transitions of care, there can be many changes to the drug regimen. Identifying medication issues and resolving these with a physician or physician-designee are important steps to prevent issues such as; adverse drug reaction, ineffective drug therapy, drug interactions, duplicate therapy, dosage errors etc., any of which could lead to hospitalization or re-hospitalization. Medication reconciliation is currently one of the Joint Commission National Patient Safety Goals –which is to maintain and communicate accurate patient medication information.
  • Impact on quality of care for patients:The improvement in the patient’s medical condition, decrease in medication errors or events as well as a decrease in use of the Emergency Department (ED) and hospitalization are all outcomes that would improve the quality of care for patients.
  
  
  
• Does the measure address a program goal or objective? Yes. This clinical measure addresses the program goals of effectiveness and safety.
  
  
• Is this a high-value measure? Yes. This is a process measure with strong links to outcomes. Medication review and reconcilliation can reduce hospitalizations and re-hospitalizations, reduce adverse events, and improve health.
  
  
• Does this measure fill a gap in the program measure set? Yes. This measure would address the high priority domain of communication and care coordination. CMS has noted the particular need for a measure of medication reconciliation for this program.
  
  
• Measure development status: Early Development
  
  
• Is the measure fully tested for the program's setting and level of analysis? Yes . The measure is being tested in Post-Acute Care Settings, and utlizes items from OASIS tested in Home Health settings. Item level testing of the measure has occurred.
  
  
• Is the measure currently in use? No.
  
  
• Does the MUC contribute to alignment and efficient use of measurement resources (burden and cost of measurement)? Yes. Yes this measure can contribute to the efficient use of measures as the items are already in the OASIS data set and regularly collected by Home Health Agencies. This measure would also be used in the SNF, IRF, and LTCH programs. This would promote alignment between settings and allow for easier comparisons of providers.
  
  
• Is the measure NQF endorsed for the program's setting and level of analysis? No. Never Submitted
  
  
• Does the measure address a high-priority quality issue in the dual eligible beneficiary population? Yes.
  
  

Rationale for measure provided by HHS
Medication review in post-acute care is generally considered to include medication reconciliation for all medications and medication review for what poses as potential clinically significant medication issues for the patient/resident. As a process measure, medication reconciliation and medication review for potential clinically significant medication issues are expected to reduce re-hospitalizations, reduce adverse events related to medications and improve health outcomes.

Measure Specifications

• NQF Number (if applicable):
  
  
• Description: Percentage of patients who were assessed for falls risk and whose care plan reflects the assessment and was implemented as appropriate.
  
  
• Numerator: Number of patients who were assessed for falls risk and whose risk was incorporated in the care plan based on assessment results and whose care plan was implemented (must meet all 3 conditions)
  
  
• Denominator: Number of home health episodes of care ending with a discharge during the reporting period, other than those covered by generic or measure-specific exclusions.
  
  
• Exclusions: Episodes of care ending with a transfer to an inpatient setting or death are excluded from the denominator. HHA's with denominator counts of less than 20 in the sample will be excluded from public reporting owing to small sample size.
  
  
• HHS NQS Priority: Making Care Safer
  
  
• HHS Data Source: OASIS-C1
  
  
• Measure Type: Composite
  
  
• Steward: Centers for Medicare & Medicaid Services
  
  
• Endorsement Status: Never Submitted
  
  
• Is the measure specified as an electronic clinical quality measure? No
  
  

Preliminary Analysis of Measure

• Preliminary analysis result: Encourage Continued Development
  
  
• Preliminary analysis summary
  • Contribution to program measure set:Falls among older people are high risk events associated with mortality, injury, and substantial amounts of health care resource use. A Cochrane review of 111 RCTs reports a 30% fall rate among community dwelling older people with evidence that multifactorial assessment and interventions reduce the rate of falls but not the risk of falls. Some studies demonstrated that both the risk of falls and rate of falls were reduced with certain interventions. http://www.ncbi.nlm.nih.gov/pubmed/19370674Another study shows that assessment and the adoption of strategies to prevent falls were effective. The Tinetti et al., (2008) study showed that the adoption of effective risk assessments and strategies for the prevention of falls (e.g., medication reduction and balance and gait training) produced an 11% reduction in serious fall in person over 70 yrs of age. The outcomes were rates of serious fall-related injuries (hip and other fractures, head injuries, and joint dislocations) and fall-related use of medical services per 1000 person-years among persons who were 70 years of age or older. One of the settings for this study was home health agencies. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3472807/pdf/nihms410481.pdf
  • Impact on quality of care for patients:This measure will encourage home health agencies to promote patient safety by conducting fall risk assessment and implementation of a plan of care to prevent falls for patients aged 65 or older. It will promote patient safety and potentially lower the number of falls and related complications.
  
  
  
• Does the measure address a program goal or objective? Yes. Falls risk and related clinical intervention assessments are considered safety measures which meet the program’s goals.?
  
  
• Is this a high-value measure? Yes. This is a composite measure –measuring 3 components of falls—patients who were assessed for falls risk; whose risk was incorporated in the care plan based on assessment results; and whose care plan was implemented. Composite measures are considered high-value.
  
  
• Does this measure fill a gap in the program measure set? Yes. This measures meets the high priority domains of making care safer and goal attainment.
  
  
• Measure development status: Early Development
  
  
• Is the measure fully tested for the program's setting and level of analysis? No. Unclear. Measure will utilize OASIS-1 data which are submitted by home health agencies.
  
  
• Since no, could the measure be revised to use in the setting or at level of analysis under consideration? Yes.
  
  
• Is the measure currently in use? No.
  
  
• Does the MUC contribute to alignment and efficient use of measurement resources (burden and cost of measurement)? Yes. This will align with other PAC providers with the development and support of measures focused on fall risk and major injury due to falls across settings.
  
  
• Is the measure NQF endorsed for the program's setting and level of analysis? No. Never Submitted
  
  
• Does the measure address a high-priority quality issue in the dual eligible beneficiary population? Yes.
  
  

Rationale for measure provided by HHS
See literature review for NQF #0537 about the importance of assessing falls among home health patients and developing interventions.

Measure Specifications

• NQF Number (if applicable):
  
  
• Description: Percentage of home health episodes of care during which a patient with a primary diagnosis of CHF, asthma and/or COPD became less short of breath or dyspneic.
  
  
• Numerator: Number of home health episodes of care where a patient with a primary diagnosis of CHF and/or COPD has less dyspnea at discharge than at start (or resumption) of care.
  
  
• Denominator: Number of home health episodes of care ending with a primary diagnosis of CHF and/or COPD with a discharge during the reporting period, other than those covered by generic or measure-specific exclusions.
  
  
• Exclusions: Episodes of care ending with a transfer to an inpatient setting or death are excluded from the denominator. HHA's with denominator counts of less than 20 in the sample will be excluded from public reporting owing to small sample size.
  
  
• HHS NQS Priority: Making Care Safer
  
  
• HHS Data Source: OASIS-C1
  
  
• Measure Type: Outcome
  
  
• Steward: Centers for Medicare & Medicaid Services
  
  
• Endorsement Status: Never Submitted
  
  
• Is the measure specified as an electronic clinical quality measure? No
  
  

Preliminary Analysis of Measure

• Preliminary analysis result: Encourage Continued Development
  
  
• Preliminary analysis summary
  • Contribution to program measure set:Dyspnea is the subjective experience – of discomfort with breathing and is described as breathlessness. Dyspnea is a cardinal symptom of Chronic Obstructive Pulmonary Disease (COPD), as well as Heart Failure (HF) and Asthma. Although the underlying etiology for these diseases is different, research has shown that evidence-based practices can improve the patient’s ability to breath. Dyspnea for three disease groups is related to the previous measure 0179 -Improvement in Dyspnea interfering with activity. Dyspnea interfering with activity is an important health status indicator that impacts quality of life and substantially affects a patient’s ability to engage in a wide variety of activities. The etiology of dyspnea interfering with activity varies (disease-related and/or related to deconditioning from an extended time of limited activity like bedrest), but a high proportion of home health care patients are affected based on the data reported by home health care agencies where 70% of patients are reported as having some dyspnea interfering with activity.Dyspnea interfering with activity has been identified as a risk factor for hospitalization among Medicare home care patients in one large study (n = 922) of home health care http://www.ncbi.nlm.nih.gov/pubmed/17099104Research supports the benefits of beta-2 long and short acting agonists, anticholinergics broncodilators as well as other interventions that can improve the COPD patient’s shortness of breath. http://www.goldcopd.org/ For HF patients there are a number of strategies to improve the symptoms of HF including dyspnea. These include the ability to self- manage care, dietary restriction, daily weighing, exercise and medication adherence http://www.ncbi.nlm.nih.gov/pubmed/17099104
  • Impact on quality of care for patients:Reporting of this measure is important for home health care patients. This symptom affects patients and can be debilitating. It can lead to frequent hospitalizations and poor quality of life. There has been improvement in this measure over time, suggesting that agencies are improving care for this outcome. There are number of best practice improvement packages developed by Home Health Quality Improvement (HHQI) in the scope of work focused on dyspnea and interventions to improve dyspnea.
  
  
  
• Does the measure address a program goal or objective? Yes. This is a clinical measure assessing symptom improvement for a patient population, and thus meets the program goal of effectiveness of care.?
  
  
• Is this a high-value measure? Yes. This is an outcome measure that addresses an important aspect of patient quality of life.
  
  
• Does this measure fill a gap in the program measure set? No. This is a modification to an existing measure to include three diseases: Heart Failure, COPD and Asthma instead of all diseases. This update reflects input from the Pulmonary and Critical Care Steering Committee that this meausre should be limited to cardiopulmonary conditions.
  
  
• Measure development status: Early Development
  
  
• Is the measure fully tested for the program's setting and level of analysis? No.
  
  
• Since no, could the measure be revised to use in the setting or at level of analysis under consideration? Yes. This measure is a modification of a previously endorsed NQF measure (0179: Improvement in Dyspnea – Endorsement Removed), thus should have the potential for revision and testing at the home health agency level.
  
  
• Is the measure currently in use? No.
  
  
• Does the MUC contribute to alignment and efficient use of measurement resources (burden and cost of measurement)? Yes. This measure is proposed to use data from OASIS, thus should be collected as current practice and not contribute to measure burden or additional costs of collection.
  
  
• Is the measure NQF endorsed for the program's setting and level of analysis? No. Never Submitted
  
  
• Does the measure address a high-priority quality issue in the dual eligible beneficiary population? Yes.
  
  

Rationale for measure provided by HHS
See literature for NQF measure #0179 about the importance of dyspnea and the potential for home health to affect outcomes.

Measure Specifications

• NQF Number (if applicable):
  
  
• Description: The MSPB-PAC Measure for HHAs evaluates providers’ efficiency relative to the efficiency of the national median HHA provider. Specifically, the MSPB-PAC Measure assesses the cost to Medicare for services during an episode of care, which consists of a treatment period and an associated services period. The episode is triggered by the initiation of a 60 day HHA service period. The treatment period begins at the trigger and ends on the last day of the service period. The associated services period begins at the trigger and ends 30 days after the end of the treatment period. These periods constitute the episode window during which beneficiaries’ Medicare services are counted toward the episode. The MSPB-PAC episode includes all services during the episode window that are attributable to the HHA provider and those rendered by other providers, except those services during the associated services period that are clinically unrelated to HHA responsibilities (e.g., planned care and routine screening).
  
  
• Numerator: The numerator is the attributed provider’s average MSPB-PAC Amount. The MSPB-PAC Amount for each HHA provider depends on two factors: i) the average of the ratio of standardized episode spending level and expected episode spending for each HHA provider; and ii) the average standardized episode spending across all HHA providers. To calculate the MSPB-PAC Amount for each HHA, one finds the average of the ratio of the standardized episode spending over the expected episode spending, and then multiplies this quantity by the average episode spending level across all HHAs.
  
  
• Denominator: The denominator for a HHA’s MSPB-PAC Measure is the weighted median MSPB-PAC Amount across all episodes for HHAs nationally.
  
  
• Exclusions: The measure excludes the following episodes: • Any episode that is triggered by HHA Request for Anticipated Payment (RAP) claims. • Any episode that is triggered by an HHA claim that happens outside the 50 states or DC. • Any episode that is triggered by an HHA claim for which we see Part C crossover claims. • Any episode for which standard allowed amount of the HHA claim could not be calculated or is equal to 0. • Any episode in which a beneficiary is not enrolled in Medicare Fee-for-Service for the entirety of the lookback period plus the episode window or is enrolled in Part C for any part of the lookback plus episode window. • Any episode in which a beneficiary has a primary payer other than Medicare for any part of the lookback plus episode window. • Any episode for which the lookback period extends beyond our observation period.
  
  
• HHS NQS Priority: Making Care Affordable
  
  
• HHS Data Source: Claims
  
  
• Measure Type: Cost/Resource Use
  
  
• Steward: Centers for Medicare & Medicaid Services
  
  
• Endorsement Status: Never Submitted
  
  
• Is the measure specified as an electronic clinical quality measure? No
  
  

Preliminary Analysis of Measure

• Preliminary analysis result: Encourage Continued Development
  
  
• Preliminary analysis summary
  • Contribution to program measure set:The Medicare Spending Per Beneficiary is a measure that compares the cost for each episode per beneficiary of home health care with all other home health agencies. This measure could incentivize agencies to lower the cost of care per patient. Agencies could look for ways to standardize care and processes to ensure consistency of practices. It will be important to evaluate whether agencies “cherry pick” patients who have fewer needs for care.
  • Impact on quality of care for patients:The potential impact is that patients may receive additional support, education and services to enable them to transition to the community and stay in the community without hospitalizations, emergency department use as well as admissions to SNFs and home health care.
  
  
  
• Does the measure address a program goal or objective? Yes. This measure would address the IMPACT Act requirement to implement a measure of Medicare Spending Per Beneficiary.
  
  
• Is this a high-value measure? Yes. This is an outcome and utilization measure (cost/resource use) that focuses on the efficient use of health care resources.
  
  
• Does this measure fill a gap in the program measure set? Yes. This measure addresses the high priority domain of making care affordable. CMS has noted the need to assess medical costs based on post-acute care episodes.
  
  
• Measure development status: Early Development
  
  
• Is the measure fully tested for the program's setting and level of analysis? No. The information received indicates testing utlizies claims data from PAC settings which should include home health, but no special mention of home health agencies.
  
  
• Since no, could the measure be revised to use in the setting or at level of analysis under consideration? Yes. This measure will calculated from the episode of care claims data which should be available across PAC settings.
  
  
• Is the measure currently in use? No.
  
  
• Does the MUC contribute to alignment and efficient use of measurement resources (burden and cost of measurement)? Yes. This measure would contribute to alignment as it will be used across all of PAC settings except Hospice. In addition, since it will be calculated from claims data it should not pose a burden to home health agencies.
  
  
• Is the measure NQF endorsed for the program's setting and level of analysis? N/A. Never Submitted
  
  
• Does the measure address a high-priority quality issue in the dual eligible beneficiary population? Yes.
  
  

Rationale for measure provided by HHS
Medicare payments to PAC have grown at a consistently higher rate than other major Medicare sectors. Between 2001 and 2013, Medicare PAC spending grew at an annual rate of 6.1 percent and doubled to $59.4 billion.

Measure Specifications

• NQF Number (if applicable):
  
  
• Description: All-condition risk-adjusted potentially preventable hospital readmission rates.
  
  
• Numerator: This measure does not have a simple form for the numerator and denominator. The numerator is defined as the risk-adjusted estimate of the number of unplanned, potentially preventable readmissions that occurred within 30 days from discharge from the prior proximal acute hospitalization. The numerator, as defined, includes risk adjustment for patient characteristics and a statistical estimate of the facility effect beyond patient mix.
  
  
• Denominator: The denominator is computed with the same model used for the numerator. It is the model developed using all non-excluded episodes in the national data. The measure includes all episodes in the measurement period that are observed in national Medicare FFS data and do not fall into an excluded category. For a particular agency, the model is applied to the patient population, but the agency effect term is 0. In essence, it is the number of unplanned potentially preventable readmissions that would be expected for that patient population at the average agency.
  
  
• Exclusions: i) Patients who are under 18 years old; (ii) Patients not continuously enrolled in Part A FFS Medicare for the 12 months prior to the HH episode admission date, and at least 30 days after discharge date; (iii) Patients who died during the HH episode; (iv) Patients with a missing risk adjustment authorization code; (v) Patients who leave HH against medical advice; (vi) Patients transfer at the end of a stay to another setting; (vii) Patients who did not have a short-term acute care stay within 30 days prior to the HH episode admission date; (viii) Patient who had the following principal diagnoses in the prior proximal hospitalization: medical (nonsurgical) treatment of cancer; primary psychiatric diseases; rehabilitation care/fitting of prostheses and for the adjustment of devices.
  
  
• HHS NQS Priority: Making Care Safer
  
  
• HHS Data Source: Administrative claims
  
  
• Measure Type: Outcome
  
  
• Steward: Centers for Medicare & Medicaid Services
  
  
• Endorsement Status: Never Submitted
  
  
• Is the measure specified as an electronic clinical quality measure? No
  
  

Preliminary Analysis of Measure

• Preliminary analysis result: Encourage Continued Development
  
  
• Preliminary analysis summary
  • Contribution to program measure set:This measure is valuable as it is an indicator of the ability of home health agencies to keep patients out of the hospital following the proximal acute hospitalization. It evaluates whether patients are adequately prepared to care for themselves, maintain their level of care and have the knowledge and resources to stay in the community. Some of the resources and knowledge for the patient include the ability to: obtain and take medications correctly, obtain supplies and care for non-healed wound, and obtain home supplies such as oxygen. Patients will also need to know the “red flags” for their condition and when and where to seek help when needed.
  • Impact on quality of care for patients:Keeping patients out of the hospital is one of the key goals during and following a home health episode of care. Patients would benefit from home health agencies providing the resources and knowledge for patients to keep them in the community without a re-hospitalization.
  
  
  
• Does the measure address a program goal or objective? Yes. Measures of potentially avoidable admissions and readmissions are considered safety measures based on their focus of reducing harms caused in and by the health care delivery system. Safety measures meet the programs goals and objectives.
  
  
• Is this a high-value measure? Yes. As an outcome and efficiency measure, it meets the criteria to be a high-value measure.
  
  
• Does this measure fill a gap in the program measure set? Yes. This risk-adjusted measure would be the first to evaluate potentiallly preventable re-hospitalization in the 30 days following discharge from home health care.
  
  
• Measure development status: Early Development
  
  
• Is the measure fully tested for the program's setting and level of analysis? No. Information received indicates testing has not been done.
  
  
• Since no, could the measure be revised to use in the setting or at level of analysis under consideration? Yes. This measure is being developed to promote alignment across the various PAC settings, and thus should be tested in the home health environmnet in order to assess the agency level of performance and scientific acceptability.
  
  
• Is the measure currently in use? No.
  
  
• Does the MUC contribute to alignment and efficient use of measurement resources (burden and cost of measurement)? Yes. Yes it measures whether patients post –discharge from home health are readmitted to the hospital within 30 days. It will be used in all PAC settings except hospices.
  
  
• Is the measure NQF endorsed for the program's setting and level of analysis? No. Never Submitted
  
  
• Does the measure address a high-priority quality issue in the dual eligible beneficiary population? Yes.
  
  

Rationale for measure provided by HHS
This is the environmental scan conducted that demonstrates potentially preventable readmissions is a concern for community dwelling individuals and that home health interventions can reduce the risk of readmission.

Measure Specifications

• NQF Number (if applicable):
  
  
• Description: This measure will assess percentage of hospice patients who received care processes consistent with guidelines at admission. This is a composite measure based on select measures from 7 NQF-endorsed measures: NQF #1641, NQF #1647, NQF #1634, NQF #1637, NQF #1639, NQF #1638, NQF #1617.
  
  
• Numerator: The numerator is patients who meet the numerator criteria for all of the select measures of the 7 NQF-endorsed measures: 1641, 1647 (modified), 1634, 1637, 1639, 1638, and 1617. Specifically, these measures are: NQF #1641 Hospice and Palliative Care – Treatment Preferences NQF #1647 (modified) Beliefs/Values Addressed (if desired by the patient) NQF #1634 Hospice and Palliative Care – Pain Screening NQF #1637 Hospice and Palliative Care – Pain Assessment NQF #1639 Hospice and Palliative Care – Dyspnea Screening NQF #1638 Hospice and Palliative Care – Dyspnea Treatment NQF #1617 Patients Treated with an Opioid Who Are Given a Bowel Regimen
  
  
• Denominator: All hospice patients
  
  
• Exclusions: Patients under 18 years of age
  
  
• HHS NQS Priority: Making Care Safer, Patient and Family Engagement, Communication and Care Coordination, Effective Prevention and Treatment
  
  
• HHS Data Source: This is a composite measure based on 7 NQF-endorsed measures. The data to calculate the 7 NQF-endorsed measures are collected through the Hospice Item Set (HIS). CMS implemented the HIS, a standardized, patient-level data collection instrument, as part of the Hospice Quality Reporting Program (HQRP) in the FY 2014 Hospice Wage Index final rule (78 FR 48234–48281). Medicare-certified hospices are required to submit an HIS-Admission record and an HIS-Discharge record for each patient admission on or after July 1, 2014. Thus, the data elements needed for this measure are readily available at national level.
  
  
• Measure Type: Composite
  
  
• Steward: Centers for Medicare & Medicaid Services
  
  
• Endorsement Status: Never Submitted
  
  
• Is the measure specified as an electronic clinical quality measure? No
  
  

Preliminary Analysis of Measure

• Preliminary analysis result: Encourage Continued Development
  
  
• Preliminary analysis summary
  • Contribution to program measure set:The inclusion of this measure in the Hospice Reporting Program will promote alignment with the existing measures and reduce redundancies in reporting. It provides an opportunity for the integration of multiple identified gap priority areas to be addressed and standardized.
  • Impact on quality of care for patients:The measure will promote standardization of the collection and reporting of data prioritized by the hospice community to be of importance for the clinical treatment of hospice patients and recognition of hospice patient/family/caregiver goals.
  
  
  
• Does the measure address a program goal or objective? Yes. This composite measure will assess percentage of hospice patients who received care processes consistent with guidelines at admission and includes aspects of care identified as important. The primary goals of the hospice program re to make the hospice patient as physically and emotionally comfortable as possible with minimal disruption to nomral activities while remaining primarily in the home environment. This composite measure meets the program goals. .
  
  
• Is this a high-value measure? Yes. Composite measures are considered to be of high-value.
  
  
• Does this measure fill a gap in the program measure set? Yes. If this composite measure is replacing the exisiting single measures used in the program, it will ensure the concepts of care coordination, communication and symptom management continue to be addressed in the program.
  
  
• Measure development status: Early Development
  
  
• Is the measure fully tested for the program's setting and level of analysis? Yes. This composite measure combines multiple measures currently in use in the Hospice Quality Reporting Program.
  
  
• Is the measure currently in use? No. While the measure is a modification of existing measures, it is not in use in its proposed composite format.
  
  
• Does the MUC contribute to alignment and efficient use of measurement resources (burden and cost of measurement)? Yes. The composite measure will utilize the Hospice Item Set which is a standardized, patient-level instrument in use across hospice settings.
  
  
• Is the measure NQF endorsed for the program's setting and level of analysis? No. Never Submitted
  
  
• Does the measure address a high-priority quality issue in the dual eligible beneficiary population? No.
  
  

Rationale for measure provided by HHS
Treatment Preferences and Spiritual Care The Hospice and Palliative Care - Treatment Preferences measure addresses patient autonomy for patients with high severity of illness and risk of death, including seriously and incurably ill patients enrolled in hospice or hospital-based palliative care. The National Priorities Partnership has identified palliative and end-of-life care as one of its national priorities. A goal of this priority is to ensure that all patients with life-limiting illness have the right to express preferences that guide use of invasive or life-sustaining forms of treatment. The affected populations are large; in 2009, 1.56 million people with life-limiting illness received hospice care.(NHPCO 2010) In 2008, 58.5% of US hospitals with 50 or more beds had some form of palliative care service, and national trends show steady expansion of these services.(Center to Advance Palliative Care 2010) Patients and family caregivers rate control over treatment decisions as a high priority when living with serious and life-limiting illnesses. (Singer et al 1999) From a recent systematic review of clinical trials, moderate evidence supports multicomponent interventions to increase advance directives, and "care planning through engaging values, involving skilled facilitators, and focusing on key decision makers.” These studies found improved outcomes of patient-physician communication, improved satisfaction with care, and increased hospice enrollment.( Lorenz et al 2008) The more recently published Coping with Cancer Study, a prospective observational study of over 300 patients with advanced cancer, found that communication of patient treatment preferences was associated with use of treatments honoring those preferences and wish lesser use of aggressive, high-cost treatments.(Wright et al 2010; 2008) Spiritual care also has been shown to be a critical element of quality of life at the end of life.(Boston et al 2011; Cohen et al 1996; Puchalski et al 2009; Steinhauser et al 2000) References Boston P, Bruce A, Schrieber R. Existential suffering in the palliative care setting: an integrated literature review. J Pain Symptom Manage. 2011 Mar;41(3):604-18. Epub 2010 Dec 8. Center to Advance Palliative Care http://www.capc.org/news-and-events/releases/04-05-10 Cohen SR, Mount BM, Tomas JJN, Mount LF. Existential well-being is an important determinant of quality of life. Cancer 1996; 77:576-86. Lorenz KA, Lynn J, Dy SM et al. Evidence for improving palliative care at the end of life: a systematic review. Ann Intern Med 2008: 148:147-159. NHPCO Facts and figures: hospice care in America 2010 edition http://www.nhpco.org/files/public/Statistics_Research/Hospice_Facts_Figures_Oct-2010.pdf Puchalski C, Ferrell B, Virani R, Otis-Green S, Baird P, Bull J, Chochinov H, Handzo G, Nelson-Becker H, Prince-Paul M, Pugliese K, Sulmasy D. Improving the quality of spiritual care as a dimension of palliative care: the report of the Consensus Conference. J Palliat Med. 2009 Oct;12(10):885-904. Review. Singer PA, Martin DK, Kelner M. Quality end-of-life care: patients´ perspective. JAMA 1999; 281: 163-168. Steinhauser KE, Christakis NA, Clipp EC, McNeilly M, McIntyre L, Tulsky JA. Factors considered important at the end of life by patients, family, physicians, and other care providers. JAMA 2000 Nov 15;284(19):2476-82. Wright AA, Zhang B, Ray A et al. Associations between end-of-life discussions, patient mental health, medical care near death, and caregiver bereavement adjustment. JAMA 2008; 300:1665-1673. Wright AA, Mack JW, Kritek PA, Balboni TA, Massaro AF, Matulonis UA, Block SD, Prigerson HG. Influence of patients’ preferences and treatment site on cancer patients’ end of life care. Cancer. 2010 Oct 1;116(19):4656-63. Pain Research on care of patients with serious incurable illness and those nearing the end of life shows they experience high rates of pain (40-70% prevalence) and other physical, emotional, and spiritual causes of distress. (SUPPORT, 1995; Gade et al 2008) The National Priorities Partnership has identified palliative and end-of-life care as one of its national priorities. A goal of this priority is to ensure that all patients with life-limiting illness have access to effective treatment for symptoms such as pain and shortness of breath. The affected populations are large; in 2009, 1.56 million people with life-limiting illness received hospice care. (NHPCO, 2010) In 2008, 58.5% of US hospitals with 50 or more beds had some form of palliative care service, and national trends show steady expansion of these services.(Center to Advance Palliative Care 2010) Patients and family caregivers rate pain management as a high priority when living with serious and life-limiting illnesses. (Singer, 1999) The consequences of inadequate screening, assessment and treatment for pain include physical suffering, functional limitation, and development of apathy and depression. (Gordon 2005) References: Center to Advance Palliative Care http://www.capc.org/news-and-events/releases/04-05-10 Gade G, Venohr I, Conner D, et al. Impact of an inpatient palliative care team: a randomized control trial. J Palliat Med. 2008;11(2):180–190. Gordon DB, Dahl JL, Miaskowski C et al. American Pain Society recommendations for improving the quality of acute and cancer pain management. Arch Intern Med 2005; 165:1574-1580. NHPCO Facts and figures: hospice care in America 2010 edition http://www.nhpco.org/files/public/Statistics_Research/Hospice_Facts_Figures_Oct-2010.pdf Singer PA, Martin DK, Kelner M. Quality end-of-life care: patients´ perspective. JAMA 1999; 281: 163-168. The Writing Group for the SUPPORT Investigators. A controlled trial to improve care for seriously ill hospitalized patients. The study to understand prognosis and preferences for outcomes and risks of treatments (SUPPORT). JAMA. 1995;274:1591-1598. http://www.nationalprioritiespartnership.org/PriorityDetails.aspx?id=608 Shortness of Breath Dyspnea is a common symptom in serious illness, more common than pain for patients with chronic obstructive lung disease, lung cancer, cystic fibrosis, and restrictive lung diseases such as pulmonary fibrosis.(Luce et al 2001) Unlike pain, dyspnea severity is associated with the risk of death.(Olajidae et al 2007) Between 50-70% of patients with advanced lung cancer experience dyspnea near the end of life. As detailed in a recent systematic review, opioids, oxygen and non-pharmacologic nursing interventions demonstrate efficacy in randomized controlled trials of treatment for dyspnea in cancer and in other serious illness.( Ben-Aharon et al 2008; Lorenz et al 2008) Unfortunately, dyspnea is often persistent and under-treated in advanced cancer and other end-stage diseases.( Roberts et al 1993 ) References: Ben-Aharon I, Gafter-Gvili A, Paul M et al. Interventions for alleviating cancer-related dyspnea: a systematic review. J Clin Oncol 2008; 26:2396-2404. Lorenz KA, Lynn J, Dy SM et al. Evidence for improving palliative care at the end of life: a systematic review. Ann Intern Med 2008; 148:147-159. Luce JM, Luce JA. Management of dyspnea in patients with far-advanced lung disease. JAMA 2001; 285:1331-1337. Olajidae O, Hanson LC, Usher BM et al. Validation of the Palliative Performance Score in the acute tertiary hospital setting. J Palliat Med 2007; 10:111-117 Roberts DK, Thorne SE, Pearson C. Cancer Nurs 1993; 16:310-320 Bowel Regimen Opioids are commonly used in the management of moderate to severe pain, and constipation is a common adverse effect. (Myotoku 2010; Tuteja 2010; Pappagallo 2001) A systematic review evaluating the extent and management of opioid-related side effects in both cancer and non-cancer patients indicated that tolerance is not developed to opioid-induced constipation and confirmed the need for prophylaxis. (McNicol 2003) Risk of constipation is further aggravated by immobility and dehydration in older people with pain. The American Pain Society and American Geriatrics Society as well as expert consensus opinion recognize the frequency of constipation with opioid use and the necessity for prophylactic therapy. (APS 2005; RANO 2002; AGS 2002; APS 2002; Weiner 2001; Davis 2003; Etzioni 2007; Dy 2008) A study of 194,017 emergency department visits made by 76,759 cancer patients in the final 6 months of life revealed that 3,392 visits were made for constipation. (Barbera 2010) A Cochrane systematic review of 26 studies of patients at least 18 years old taking opioids for at least 6 months for non-cancer pain revealed gastrointestinal complaints (e.g., constipation, nausea, dyspepsia) as the most commonly reported side effect. (Noble 2010) References: AGS Panel on Persistent Pain in Older Persons. The management of persistent pain in older persons. J Am Geriatr Soc 2002;50(6 Suppl):S205-24 American Pain Society (APS). Guideline for the management of cancer pain in adults and children. 2005 American Pain Society (APS). Guideline of the management of pain in osteoarthritis, rheumatoid arthritis, and juvenile chronic arthritis. 2002. Barbera L, Taylor C, Dudgeon D. Why do patients with cancer visit the emergency department near the end of life? Can Med Assoc J 2010;182(6):563-569 Davis MP, Srivastava M. Demographics, assessment and management of pain in the elderly. Drugs Aging 2003;20(1):23-57 Dy SM, Asch SM, Naeim A, et al. Evidence-based standards for cancer pain. J Clin Oncol 2008;26(23):3879-3885 Etzioni S, Chodosh J, Ferrell BA, et al. Quality indicators for pain management in vulnerable elders. JAGS 2007;55:S403-S408 McNicol E, Horowicz-Mehler N, Fisk RA et al. Management of opioid side effects in cancer-related and chronic noncancer pain: a systematic review. J Pain 2003;4(5):231-56 Myotoku M, Nakanishi A, Kanematsu M, et al. Reduction in opioid side effects by prophylactic measures of palliative care team may result in improved quality of life. J Pall Care 2010;13(4):401-406 Noble M, Treadwell JR, Tregear SJ, et al. Long-term opioid treatment for chronic noncancer pain. Cochrane Database Sys Rev 2010;(1):CD006605 Pappagallo M. Incidence, prevalence, and management of opioid bowel dysfunction. Am J Surg 2001;182(5A Suppl):11s-8s Registered Nurses Association of Ontario (RNAO). Assessment and management of pain. 2002. (Nursing Best Practice Guideline: Shaping the Future of Nursing) Tuteja AK, Biskupiak J, Stoddard GJ, et al. Opioid-induced bowel disorders and narcotic bowel syndrome in patients with chronic non-cancer pain. Neurogastroenterol Motil 2010;22:424-e96 Weiner DK, Hanlon JT. Pain in nursing home residents: management strategies. Drugs Aging 2001;18(1):13-29

Measure Specifications

• NQF Number (if applicable):
  
  
• Description: This measure will assess hospice staff visits to patients and caregivers in the last week of life.
  
  
• Numerator: The numerator of this measure will be the number of patients in the denominator who receive hospice staff visits in the last week of life. Members of the hospice staff whose visits are considered for the measure include: nurses (registered nurse, licensed professional nurse or nurse practitioner if acting in the role of a nurse), hospice aides, physicians (or nurse practitioner or physician assistant if acting as the attending physician), chaplains or spiritual counselors, therapists (physical therapist, occupational therapist or speech language therapist), medical social workers, and volunteers.
  
  
• Denominator: The denominator is the number of hospice patients who are discharged as expired within a defined target period.
  
  
• Exclusions: Patients who received continuous home care or general inpatient care only in the last week of life.
  
  
• HHS NQS Priority: Making Care Safer, Effective Prevention and Treatment, Making Care Affordable, Patient and Family Engagement, Communication and Care Coordination,
  
  
• HHS Data Source: Hospice Item Set (HIS)
  
  
• Measure Type: Process
  
  
• Steward: Centers for Medicare & Medicaid Services
  
  
• Endorsement Status: Never Submitted
  
  
• Is the measure specified as an electronic clinical quality measure? No
  
  

Preliminary Analysis of Measure

• Preliminary analysis result: Encourage continued development
  
  
• Preliminary analysis summary
  • Contribution to program measure set:This measure meets the priorities, needs and requirements of the HQRP. Process measures focused on assessment are generally many steps removed from relevant outcomes. However, there aren’t currently any measures in the program that are related to the assessment of hospice staff visits to patients and caregivers in the last week of life. This measure could be considered to fill a gap in that it adds to the relatively limited set if measures specific to this area.
  • Impact on quality of care for patients:http://www.qualityforum.org/Publications/2006/12/A_National_Framework_and_Preferred_Practices_for_Palliative_and_Hospice_Care_Quality.aspx NQF’s Framework for Preferred Practices for Palliative Care recommends that signs and symptoms of impending death are recognized, communicated and educated, and care appropriate for the phase of illness is provided. Assessing hopice staff visits in the last week of life can be linked to hospices more proactively checking on their patients, which creates better opportunities of recognizing signs of impending death, communicating and eduation about symptoms and providing appropriate care.
  
  
  
• Does the measure address a program goal or objective? Yes. This measure addresses the program goal of advanced care planning and treatment.?
  
  
• Is this a high-value measure? Yes. This is a process measure assessing hospice staff visits to patients and caregivers in the last week of life. Assessing hopice staff visits in the last week of life could be linked to hospices more proactively checking on their patients, which creates better opportunities for recognizing signs of impending death, communicating and eduating about symptoms and providing appropriate care.
  
  
• Does this measure fill a gap in the program measure set? Yes. There are currently only eight measures in the HQRP; none of the measures are related to the assessment of hospice staff visits to patients and caregivers in the last week of life. This measure could be considered to fill a gap in that it adds to the relatively limited set if measures specific to this area. MAP has previously noted gaps in the HQRP around pain, goal attainment, patient engagement, care coordination, and depression. This measure could help address a number of these areas.
  
  
• Measure development status: Early Development. Note, there are potential changes to the measure specifications under consideration by CMS as recommended by a Caregiver TEP.
  
  
• Is the measure fully tested for the program's setting and level of analysis? No. Testing on this measure has not yet been completed. The setting for testing is hospice. The pilot test started in June 2015.
  
  
• Since no, could the measure be revised to use in the setting or at level of analysis under consideration? N/A.
  
  
• Is the measure currently in use? No.
  
  
• Does the MUC contribute to alignment and efficient use of measurement resources (burden and cost of measurement)? No. This measure will be collected utilizing data from the Hospice Item Set and is specific to care provided by hospice services. The Hospice Item Set is required for this program, thus additional data collection burden is not anticipated.
  
  
• Is the measure NQF endorsed for the program's setting and level of analysis? No. Never Submitted
  
  
• Does the measure address a high-priority quality issue in the dual eligible beneficiary population? Yes.
  
  

Rationale for measure provided by HHS
The literature supports hospice visits when death is imminent as a high priority in end-of-life care by showing the last week of life as the point in the terminal illness trajectory with the highest symptom burden. Particularly during the last few days before death, patients experience myriad physical and emotional symptoms, necessitating close care and attention from the integrated hospice team. Physical symptoms with high prevalence in the last week of life include fatigue, pain, dyspnea, respiratory secretions/death rattle, anorexia, dry mouth, nausea and/or vomiting, affecting a quarter to more than 80 percent of imminently dying patients. The specific prevalence of each symptom varies across studies, reflecting the heterogeneity of the samples and the range of assessment techniques used.(Lynn, Teno et al. 1997, Klinkenberg, Willems et al. 2004, Kehl and Kowalkowski 2012) Psychosocial symptoms with high prevalence in the last week of life include confusion, anxiety, depression and delirium, affecting a third to more than half of imminently dying patients.(Klinkenberg, Willems et al. 2004) A study of after-death interviews with close relatives of terminal patients found that 75 percent of patients experienced at least two symptoms requiring management in the last week of life.(Klinkenberg, Willems et al. 2004) The symptom burden typically increases significantly in the last few days of life compared to the previous stage,(Currow, Smith et al. 2010) further supporting care of the imminently dying patient as a high priority aspect of healthcare. Studies focusing on the expectations of patients and families also demonstrate the importance of care and attention from the hospice team in the days leading up to death. Caregivers of dying patients agree overwhelmingly with the importance of preparation at the end of life. Hospice assistance, ranging from legal to logistical to emotional, is paramount in preparing hospice patients and their families for imminent death. (Steinhauser, Christakis et al. 2000) Bereaved family members and friends from a variety of settings identified the provision of physical comfort and emotional support to dying patients and their families as fundamental aspects of high-quality care.(Steinhauser, Christakis et al. 2000) References: Currow, D.C., et al., Do the Trajectories of Dyspnea Differ in Prevalence and Intensity By Diagnosis at the End of Life? A Consecutive Cohort Study. Journal of Pain and Symptom Management, 2010. 39(4): p. 680-690. Kehl, K.A. and J.A. Kowalkowski, A Systematic Review of the Prevalence of Signs of Impending Death and Symptoms in the Last 2 Weeks of Life. American Journal of Hospice & Palliative Medicine, 2012. 30(6): p. 601-616. Klinkenberg, M., et al., Symptom Burden in the Last Week of Life. J Pain Symptom Manage., 2004. 27(1): p. 5-13. Lynn, J., et al., Perceptions by Family Members of the Dying Experience of Older and Seriously Ill Patients. Annals of Internal Medicine, 1997. 126(2): p. 97-106. Steinhauser, K.E., et al., Factors Considered Important at the End of Life by Patients, Family, Physicians, and Other Care Providers. JAMA, 2000. 284(19): p. 2476-2482.

Measure Specifications

• NQF Number (if applicable):
  
  
• Description: This measure describes the risk-standardized rate of Medicare fee-for-service (FFS) patients/residents/persons who are discharged to the community following a post-acute stay/episode, and do not have an unplanned (re)admission to an acute care hospital or LTCH in the 31 days following discharge to community, and remain alive during the 31 days following discharge to community.
  
  
• Numerator: This measure does not have a simple form for the numerator and denominator. The numerator is defined as the risk-adjusted estimate of the number of patients/residents/persons included in the measure who are discharged to the community, and do not have an unplanned (re)admission to an acute care hospital or LTCH on the day of discharge or in the 31 days following discharge to community, and remain alive during the 31 days following discharge to community. The numerator estimate includes risk adjustment for patient/resident/person characteristics, and a statistical estimate of the facility/agency effect beyond case mix.
  
  
• Denominator: The denominator is computed with the same model used for the numerator. It is the model developed using all non-excluded facility/agency stays/episodes in the national data. The measure includes all facility/agency stays/episodes in the measurement period that are observed in national Medicare FFS data and do not fall into an excluded category. For a particular facility/agency, the model is applied to the patient/resident/person population, but the facility/agency effect term is 0. In essence, it is the number of discharges to community that would be expected for that patient/resident/person population at the average facility/agency.
  
  
• Exclusions: i) Age under 18 years; (ii) No short-term acute care stay within 30 days prior to IRF admission ; (iii) Discharges to psychiatric hospital; (iv) Discharges against medical advice; (v) Discharges to federal hospitals or disaster alternative care sites; (vi) Patients not continuously enrolled in Part A FFS Medicare for the 12 months prior to IRF admission date, and at least 31 days after IRF discharge date; (vii) Patients whose prior short-term acute-care stay was for non-surgical treatment of cancer; (viii) Discharges to hospice; (ix) IRF stays that end in transfer to another IRF; (x) IRF stays with claims data that are problematic (e.g., anomalous records for stays that overlap wholly or in part, or are otherwise erroneous or contradictory); (xi) Patients who received care from a provider located outside of the US, Puerto Rico, or a US territory..
  
  
• HHS NQS Priority: Making Care Safer, Patient and Family Engagement, Making Care Affordable
  
  
• HHS Data Source: Claims
  
  
• Measure Type: Outcome
  
  
• Steward: Centers for Medicare & Medicaid Services
  
  
• Endorsement Status: Never Submitted
  
  
• Is the measure specified as an electronic clinical quality measure? No
  
  

Preliminary Analysis of Measure

• Preliminary analysis result: Encourage continued development
  
  
• Preliminary analysis summary
  • Contribution to program measure set:This measure meets the priorities, needs and requirements of the IRF QRP, addresses a high-impact area of function and resource use, and meets the requirements of the IMPACT Act. This measure would be unique to the IRF QRP in terms of addressing this quality issue (discharge to community); the program does not currently include any resource use measures in this area.
  • Impact on quality of care for patients:Restoring functional status is a primary focus on IRF care and returning home is very important to patients. This is a resource use measure that assesses discharge to community and is being proposed to meet the requirements of the IMPACT Act. IRF discharge rates vary across providers, ranging from 60% to 75%. MedPAC found in FY 2013 the average rate for discharge to the community for IRFs within 100 days was around 75%. Implementing this measure could help consumers make choices about post-acute care that are aligned with their goals of returning to the community.
  
  
  
• Does the measure address a program goal or objective? Yes. This measure addressess the IRF QRP goal of successful return to the community post-discharge.
  
  
• Is this a high-value measure? Yes. Restoring functional status is a primary focus on IRF care and returning home is very important to patients. This is a resource use measure that assesses discharge to community and is being proposed to meet the requirements of the IMPACT Act. IRF discharge rates vary across providers, ranging from 60% to 75%. MedPAC found in FY 2013 the average rate for discharge to the community for IRFs within 100 days was around 75%.
  
  
• Does this measure fill a gap in the program measure set? Yes. Currently, there are no other measures that address discharge to community in the IRF QRP. MAP previously recommended that the program measure set is too limited and could be enhanced by addresing core measure concepts such as functional status and resource use measures. This measure also addresses care coordination and transition planning concepts.
  
  
• Measure development status: Early Development
  
  
• Is the measure fully tested for the program's setting and level of analysis? No. This measure has not yet been tested.
  
  
• Since no, could the measure be revised to use in the setting or at level of analysis under consideration? N/A.
  
  
• Is the measure currently in use? No. This measure is not currently being used in a program.
  
  
• Does the MUC contribute to alignment and efficient use of measurement resources (burden and cost of measurement)? Yes. This measure would be unique to the IRF QRP in terms of measuring discharge to community; the program does not currently include any resource use measures in this area. The measure would also promote alignment of similar measure concepts cross programs.
  
  
• Is the measure NQF endorsed for the program's setting and level of analysis? No. Never Submitted
  
  
• Does the measure address a high-priority quality issue in the dual eligible beneficiary population? Yes.
  
  

Rationale for measure provided by HHS
The ultimate goals of post-acute care are avoiding institutionalization and returning patients to their previous level of independence and functioning, with discharge to community being the primary goal for the majority of post-acute patients. For many, home is a symbol of independence, privacy, and competence. Discharge to community is considered a valuable outcome to measure because it is a multifaceted measure that captures the patient’s functional status, cognitive capacity, physical ability, and availability of social support at home. There is considerable variation in discharge to community rates within and across post-acute settings. Studies show geographic variation, variation across patient socioeconomic characteristics (for example, race and ethnicity), and variation by facility characteristics (for profit vs. nonprofit, freestanding vs. hospital-based, urban vs. rural). In the IRF setting, discharge to community rates vary across providers, ranging from about 60% to 75%. The 2015 MedPAC report shows that, in FY 2013, the facility-level, mean risk-adjusted discharge to community rate for IRFs within 100 days of admission was 75.8%, and the mean observed rate was 74.7%. Discharge to community rates also vary widely in the SNF setting, ranging from as low as 31% to as high as 65%. The 2015 MedPAC report shows a mean risk-adjusted discharge to community rate of 37.5% for SNFs within 100 days of admission, and mean observed rate of 40.1%. A multicenter study of 23 LTCHs reported that only 28.8% of 1,061 patients who were ventilator-dependent on admission were discharged to home or assisted living facility. A study of 66,510 Medicare beneficiaries during pre- and post-HH episodes, revealed that 64 percent of beneficiaries discharged from HH did not use any other Medicare-reimbursed acute or post-acute services in the 30 days following HH discharge. Significant numbers of patients were admitted to inpatient facilities (29 percent) and lesser numbers to skilled nursing facilities (7.6 percent), inpatient rehabilitation (1.5 percent) and home health (7.2 percent) or hospice (3.3 percent) within 30 days of HH discharge (Wolff et al., 2008).

Measure Specifications

• NQF Number (if applicable):
  
  
• Description: Percentage of stays Inpatient Rehabilitation Facility (IRF), Long Term Care Facility (LTCH), and Skilled Nursing Facility (SNF) or care episodes Home Health (HH) in which a drug regimen review was conducted at the Admission (IRF, LTCH or SNF)/ Start of Care (SOC)/ Resumption of Care (ROC) (HH) and timely follow-up with a physician occurred each time potential clinically significant medication issues were identified throughout the stay (IRF, LTCH, or SNF) or care episode (HH).
  
  
• Numerator: Number of stays or care episodes where the medical record contains documentation of a drug regimen review conducted at admission or start-of-care or resumption-of-care with all potential clinically significant medication issues identified during the course of care and followed-up with a physician or physician designee.
  
  
• Denominator: Care episodes or stays ending during the reporting period. Assessment timing is as follows: Beginning of care episode or stay: • HH – SOC or ROC • SNF – Admission • IRF - Admission • LTCH – Admission End of care episode or stay: • HH – Transfer, Discharge, or Death at Home • SNF – Discharge, or expired • IRF – Discharge, or expired • LTCH – Discharge, or expired
  
  
• Exclusions: Denominator Exclusion: NONE Numerator Exclusion: NONE
  
  
• HHS NQS Priority: Making Care Safer, Communication and Care Coordination
  
  
• HHS Data Source: Patient assessment data
  
  
• Measure Type: Process
  
  
• Steward: Centers for Medicare & Medicaid Services
  
  
• Endorsement Status: Never Submitted
  
  
• Is the measure specified as an electronic clinical quality measure? No
  
  

Preliminary Analysis of Measure

• Preliminary analysis result: Encourage Continued Development
  
  
• Preliminary analysis summary
  • Contribution to program measure set:This measure of medication reconciliation addresses an important aspect of care for patients as they transition from one setting to another. During transitions of care, there can be many changes to the drug regimen. Identifying medication issues and resolving these with a physician or physician-designee are important steps to prevent issues such as; adverse drug reaction, ineffective drug therapy, drug interactions, duplicate therapy, dosage errors etc., any of which could lead to hospitalization or re-hospitalization. Medication reconciliation is currently one of the Joint Commission National Patient Safety Goals –which is to maintain and communicate accurate patient medication information.
  • Impact on quality of care for patients:The improvement in the patient’s medical condition, decrease in medication errors or events as well as a decrease in use of the Emergency Department (ED) and hospitalization are all outcomes that would improve the quality of care for patients.
  
  
  
• Does the measure address a program goal or objective? Yes. This clinical measure addresses the program goals of effectiveness and safety. ?
  
  
• Is this a high-value measure? Yes. This is a process measure with strong links to outcomes. Medication review and reconcilliation can reduce hospitalizations and re-hospitalizations, reduce adverse events, and improve health.
  
  
• Does this measure fill a gap in the program measure set? Yes. This measure would address the high priority domain of communication and care coordination. CMS has noted the particular need for a measure of medication reconciliation for this program.
  
  
• Measure development status: Early Development
  
  
• Is the measure fully tested for the program's setting and level of analysis? No. This measure has not yet been tested. Testing is to be conducted for Post-Acute Care settings in 2015.
  
  
• Since no, could the measure be revised to use in the setting or at level of analysis under consideration? Yes. It is anticipated that this measure will be fully tested for each setting of care for which it is proposed.
  
  
• Is the measure currently in use? No. This measure has not currently being used in a program.
  
  
• Does the MUC contribute to alignment and efficient use of measurement resources (burden and cost of measurement)? Yes . This measure is proposed to meet the requirements of the IMPACT Act and is in development to ensure alignment acorss PAC settings of care, thus should promote efficient use of resources.
  
  
• Is the measure NQF endorsed for the program's setting and level of analysis? No. Never Submitted
  
  

Rationale for measure provided by HHS
Medication review in post-acute care is generally considered to include medication reconciliation for all medications and medication review for what poses as potential clinically significant medication issues for the patient/resident. As a process measure, medication reconciliation and medication review for potential clinically significant medication issues are expected to reduce re-hospitalizations, reduce adverse events related to medications and improve health outcomes.

Measure Specifications

• NQF Number (if applicable):
  
  
• Description: The MSPB-PAC Measure for IRFs evaluates providers’ efficiency relative to the efficiency of the national median IRF provider. Specifically, the MSPB-PAC Measure assesses the cost to Medicare for services during an episode of care, which consists of a treatment period and an associated services period. The episode is triggered by an admission to an IRF stay. The treatment period begins at the trigger and ends at discharge. The associated services period begins at the trigger and ends 30 days after the end of the treatment period (i.e., discharge). These periods constitute the episode window during which beneficiaries’ Medicare services are counted toward the episode. The MSPB-PAC episode includes all services during the episode window that are attributable to the IRF provider and those rendered by other providers, except those services during the associated services period that are clinically unrelated to IRF responsibilities (e.g., planned care and routine screening).
  
  
• Numerator: The numerator is the attributed provider’s average MSPB-PAC Amount. The MSPB-PAC Amount for each IRF provider depends on two factors: i) the average of the ratio of standardized episode spending level and expected episode spending for each IRF provider; and ii) the average standardized episode spending across all IRF providers. To calculate the MSPB-PAC Amount for each IRF, one finds the average of the ratio of the standardized episode spending over the expected episode spending, and then multiplies this quantity by the average episode spending level across all IRFs.
  
  
• Denominator: The denominator for an IRF’s MSPB-PAC Measure is the weighted median MSPB-PAC Amount across all episodes for IRFs nationally.
  
  
• Exclusions: The measure excludes the following episodes: • Any episode that is triggered by an IRF stay that happens outside the 50 states or DC. • Any episode that is triggered by an IRF stay for which we see Part C crossover claims. • Any episode for which standard allowed amount of the IRF stay could not be calculated or is equal to 0. • Any episode in which a beneficiary is not enrolled in Medicare Fee-for-Service for the entirety of the lookback period plus the episode window or is enrolled in Part C for any part of the lookback plus episode window. • Any episode in which a beneficiary has a primary payer other than Medicare for any part of the lookback plus episode window. • Any episode for which the lookback period extends beyond our observation period.
  
  
• HHS NQS Priority: Making Care Affordable
  
  
• HHS Data Source: Claims
  
  
• Measure Type: Cost/Resource Use
  
  
• Steward: Centers for Medicare & Medicaid Services
  
  
• Endorsement Status: Never Submitted
  
  
• Is the measure specified as an electronic clinical quality measure? No
  
  

Preliminary Analysis of Measure

• Preliminary analysis result: Encourage Continued Development
  
  
• Preliminary analysis summary
  • Contribution to program measure set:The Medicare Spending per beneficiary measure assesses the cost to Medicare for services performed by PAC providers and other healthcare providers during an episode.This measure could incentivize providers to lower the cost of care per patient. Providers could look for ways to standardize care and processes to ensure consistency of practices. This measure would address an IMPACT Act requirement.
  • Impact on quality of care for patients:The Medicare Payment Advisory Commission (MedPAC ) has found significant regional variation in post acute care spending. This measure would allow comparisons between providers and incent providers to lower costs.
  
  
  
• Does the measure address a program goal or objective? Yes. This measure would address the IMPACT Act requirement to implement a measure of Medicare Spending Per Beneficiary.
  
  
• Is this a high-value measure? Yes. This is an outcome and utilization measure (cost/resource use) that focuses on the efficient use of health care resources.
  
  
• Does this measure fill a gap in the program measure set? Yes. This measure addresses the high priority domain of making care affordable. CMS has noted the need to assess medical cists based on post-acute care episodes.
  
  
• Measure development status: Early Development
  
  
• Is the measure fully tested for the program's setting and level of analysis? No. The information received indicates testing is occuring across PAC settings, which should include IRFs. .
  
  
• Since no, could the measure be revised to use in the setting or at level of analysis under consideration? Yes. This measure will be calculated from the episode of care claims data.
  
  
• Is the measure currently in use? No.
  
  
• Does the MUC contribute to alignment and efficient use of measurement resources (burden and cost of measurement)? Yes. This measure would contribute to alignment as it will be used across all of PAC settings except Hospice; in addition, since it will be calculated from claims data it should not pose a burden to any of the PAC settings.
  
  
• Is the measure NQF endorsed for the program's setting and level of analysis? N/A. Never Submitted
  
  
• Does the measure address a high-priority quality issue in the dual eligible beneficiary population? Yes.
  
  

Rationale for measure provided by HHS
Medicare payments to PAC have grown at a consistently higher rate than other major Medicare sectors. Between 2001 and 2013, Medicare PAC spending grew at an annual rate of 6.1 percent and doubled to $59.4 billion.

Measure Specifications

• NQF Number (if applicable):
  
  
• Description: All-condition risk-adjusted potentially preventable hospital readmission rates
  
  
• Numerator: This measure does not have a simple form for the numerator and denominator. The numerator is defined as the risk-adjusted estimate of the number of unplanned potentially preventable readmissions that occurred within 30 days from IRF discharge. The numerator, as defined, includes risk adjustment for patient characteristics and a statistical estimate of the facility effect beyond patient mix.
  
  
• Denominator: The denominator is computed with the same model used for the numerator. It is the model developed using all non-excluded stays in the national data. The measure includes all stays in the measurement period that are observed in national Medicare FFS data and do not fall into an excluded category. For a particular facility, the model is applied to the patient population, but the facility effect term is 0. In essence, it is the number of unplanned potentially preventable readmissions that would be expected for that patient population at the average facility
  
  
• Exclusions: 1. Patients who died during the IRF stay 2. Patients less than 18 years old. 3. Patients who were transferred to the same level of care or a hospital at the end of their IRF stay. 4. Patients who were not continuously enrolled in Part A FFS Medicare for the 12 months prior to the IRF admissions, and at least 30 days after IRF discharge. 5. Patients who did not have a short-term acute-care stay within 30 days prior to the IRF admission date. 6. Patients who leave the IRF against medical advice 7. Patients for whom the prior short-term acute-care stay was for the nonsurgical treatment of cancer. 8. Patients who were transferred to a federal hospital from the IRF. 9. Patients who received care from a provider located outside of the US, Puerto Rico, or a US territory. 10. IRF stays with data that are problematic (e.g., anomalous records for hospital stays that overlap wholly or in part or are otherwise erroneous or contradictory).
  
  
• HHS NQS Priority: Communication and Care Coordination
  
  
• HHS Data Source: Claims: This measure is based on Medicare claims data for fee-for-service beneficiaries using IRF services. Several variables from the inpatient claims and SNF claims will be used including admission/discharge dates and diagnostic/procedure information.
  
  
• Measure Type: Outcome
  
  
• Steward: Centers for Medicare & Medicaid Services
  
  
• Endorsement Status: Never Submitted
  
  
• Is the measure specified as an electronic clinical quality measure? No
  
  

Preliminary Analysis of Measure

• Preliminary analysis result: Encourage Continued Development
  
  
• Preliminary analysis summary
  • Contribution to program measure set:This measure would address a gap in assessing care coordination and hospital readmissions. Additionally, this measure would meet an IMPACT Act requirement.
  • Impact on quality of care for patients:MedPAC estimates that 76% of hospital readmissions may be potentially preventable. Risk-standardized readmission rates across IRFs have been found to range from 11 to 16 percent. Implementing this measure could help reduce variation and close this performance gap.
  
  
  
• Does the measure address a program goal or objective? Yes. This measure of potentially avoidable readmissions addresses the program goal of successful discharge to the community. ?
  
  
• Is this a high-value measure? Yes. As an outcome and efficiency measure, as well as one that focuses on care coordination, it meets the criteria to be a high-value measure.
  
  
• Does this measure fill a gap in the program measure set? Yes. This measure would address a gap in the high priority domain of communication and care coordination. CMS has noted the need to assess care transitions and rehospitalizations. Additionally, this measure is intended to meet an IMPACT act requirement.
  
  
• Measure development status: Early Development
  
  
• Is the measure fully tested for the program's setting and level of analysis? Yes. Measure is being tested utilizing IRF claims.
  
  
• Is the measure currently in use? No. It should be noted that while this specific measure is not in current use, it is very similar to NQF #2502 which focuses on all-cause re-admissions post-discharge from an IRF.
  
  
• Does the MUC contribute to alignment and efficient use of measurement resources (burden and cost of measurement)? Yes. The measure will promote alignment for the consistent use of measures across programs. It relies on claims data for calculation thus should not introduce significant burden on facilities for reporting.
  
  
• Is the measure NQF endorsed for the program's setting and level of analysis? No. Never Submitted
  
  
• Does the measure address a high-priority quality issue in the dual eligible beneficiary population? Yes.
  
  

Rationale for measure provided by HHS
The peer-reviewed literature specific to potentially preventable readmissions following IRF discharge is limited. However, MedPAC has estimated that 76 percent of 30-day readmissions for Medicare beneficiaries overall were due to five potentially preventable conditions (heart failure, electrolyte imbalance, respiratory infection, sepsis, and urinary tract infection (MedPAC 2007).

Measure Specifications

• NQF Number (if applicable):
  
  
• Description: All-condition risk-adjusted potentially preventable hospital readmission rates occurring during an IRF stay
  
  
• Numerator: This measure does not have a simple form for the numerator and denominator. The numerator is defined as the risk-adjusted estimate of the number of unplanned potentially preventable readmissions that occurred during an IRF stay. The numerator, as defined, includes risk adjustment for patient characteristics and a statistical estimate of the facility effect beyond patient mix.
  
  
• Denominator: The denominator is computed with the same model used for the numerator. It is the model developed using all non-excluded stays in the national data. The measure includes all stays in the measurement period that are observed in national Medicare FFS data and do not fall into an excluded category. For a particular facility, the model is applied to the patient population, but the facility effect term is 0. In essence, it is the number of unplanned potentially preventable readmissions that would be expected for that patient population at the average facility
  
  
• Exclusions: 1. Patients who died during the IRF stay. 2. Patients less than 18 years old. 3. Patients who were not continuously enrolled in Part A FFS Medicare for the 12 months prior to the IRF admissions, and at least 30 days after IRF discharge. 4. Patients who did not have a short-term acute-care stay within 30 days prior to the IRF admission date. 5. Patients who leave the IRF against medical advice 6. Patients for whom the prior short-term acute-care stay was for the nonsurgical treatment of cancer. 7. Patients who were transferred to a federal hospital from the IRF. 8. Patients who received care from a provider located outside of the US, Puerto Rico, or a US territory. 9. IRF stays with data that are problematic (e.g., anomalous records for hospital stays that overlap wholly or in part or are otherwise erroneous or contradictory).
  
  
• HHS NQS Priority: Communication and Care Coordination
  
  
• HHS Data Source: Claims: This measure is based on Medicare claims data for fee-for-service beneficiaries using IRF services. Several variables from the inpatient claims will be used including admission/discharge dates and diagnostic/procedure information.
  
  
• Measure Type: Outcome
  
  
• Steward: Centers for Medicare & Medicaid Services
  
  
• Endorsement Status: Never Submitted
  
  
• Is the measure specified as an electronic clinical quality measure? No
  
  

Preliminary Analysis of Measure

• Preliminary analysis result: Encourage Continued Development
  
  
• Preliminary analysis summary
  • Contribution to program measure set:This risk-adjusted measure would be the first to evaluate potentiallly preventable re-hospitalization in while a patient is in an inpatient rehabilitation facility.
  • Impact on quality of care for patients:MedPAC estimates that 76 percent of 30-day readmissions for Medicare beneficiaries overall were due to five potentially preventable conditions (heart failure, electrolyte imbalance, respiratory infection, sepsis, and urinary tract infection (MedPAC 2007). By focusing on improved clinical management of patients with potentially preventable conditions, IRFs have an opportunity to reduce readmission rates for patients under their care.
  
  
  
• Does the measure address a program goal or objective? Yes. This measure meets the program goal of promoting appropriate care coordination and effective care to prevent unplanned re-admissions.
  
  
• Is this a high-value measure? Yes. As an outcome and efficiency measure, as well as one that focuses on care coordination, it meets the criteria to be a high-value measure.
  
  
• Does this measure fill a gap in the program measure set? Yes. This risk-adjusted measure would be the first to evaluate potentiallly preventable re-hospitalization while a patient is in an inpatient rehabilitation facility. It also is intended to meet requirements set forth in the IMPACT Act.
  
  
• Measure development status: Early Development
  
  
• Is the measure fully tested for the program's setting and level of analysis? Yes. Measure is being tested utilizing IRF claims.
  
  
• Is the measure currently in use? No. It should be noted that while this specific measure is not in current use, it is very similar to NQF #2502 which focuses on all-cause readmissions post-discharge from an IRF.
  
  
• Does the MUC contribute to alignment and efficient use of measurement resources (burden and cost of measurement)? Yes. The measure will promote alignment for the consistent use of measures across programs. It relies on claims data for calculation thus should not introduce significant burden on facilities for reporting.
  
  
• Is the measure NQF endorsed for the program's setting and level of analysis? No. Never Submitted
  
  
• Does the measure address a high-priority quality issue in the dual eligible beneficiary population? Yes.
  
  

Rationale for measure provided by HHS
The peer-reviewed literature specific to potentially preventable readmissions during an IRF stay is limited. However, MedPAC has estimated that 76 percent of 30-day readmissions for Medicare beneficiaries overall were due to five potentially preventable conditions (heart failure, electrolyte imbalance, respiratory infection, sepsis, and urinary tract infection (MedPAC 2007).

Measure Specifications

• NQF Number (if applicable):
  
  
• Description: This measure assesses facility-level compliance with Spontaneous Breathing Trial (SBT), including Tracheostomy Collar Trial (TCT) or Continuous Positive Airway Pressure (CPAP) breathing trial, by Day 2 of the LTCH stay for patients on invasive mechanical ventilation (IMV) support upon admission, and for whom at admission weaning attempts were expected or anticipated. Compliance is calculated and reported separately for the following two components: 1. the percentage of patients who were assessed for readiness for SBT (including TCT or CPAP breathing trial) by Day 2 of the LTCH stay, 2. the percentage of patients found ready for SBT (including TCT or CPAP breathing trial) for whom an SBT (including TCT or CPAP breathing trial) was performed by Day 2 of LTCH stay.
  
  
• Numerator: This measure assesses facility-level compliance with Spontaneous Breathing Trial (SBT), including Tracheostomy Collar Trial (TCT) or Continuous Positive Airway Pressure (CPAP) breathing trial, by Day 2 of the LTCH stay for patients on invasive mechanical ventilation (IMV) support upon admission, and for whom at admission weaning attempts were expected or anticipated. The numerators for the two (2) components are: the number of patients who were assessed for readiness for SBT (including TCT or CPAP breathing trial) by Day 2 of the LTCH stay, 2. the number of patients found ready for SBT (including TCT or CPAP breathing trial) for whom an SBT (including TCT or CPAP breathing trial) was performed by Day 2 of LTCH stay.
  
  
• Denominator: The target population (denominator) for this measure is the total number of patients admitted to the Long Term Care Hospital (LTCH) during the reporting period who were on invasive mechanical ventilation support upon upon admission to the LTCH,for whom weaning attempts were expected or anticipated at admission.
  
  
• Exclusions: This measure excludes patients with missing data and invasively mechanically ventilated patients identified as non-weaning at the time of admission to an LTCH. Patients who may be considered non-weaning include patients who are considered chronically ventilated as defined by evidence-based guidelines for ventilator liberation or patients with an acute or chronic condition that negates at admission any expectation or anticipation of weaning attempts (e.g. progressive neuromuscular disease such amyotrophic lateral sclerosis, or irreversible neurological injury or disease or dysfunction such as high (C2) spinal cord injury). Consideration of a patient as non-weaning must be based on documentation found in the patient’s medical record by Day 2 of LTCH Stay.
  
  
• HHS NQS Priority: Making Care Safer
  
  
• HHS Data Source: LTCH CARE data set. Assessment items related to mechanical ventilation support will be added to the current assessment tool for this setting (LTCH CARE Data Set) to ensure operationalization of this measure.
  
  
• Measure Type: Process
  
  
• Steward: Centers for Medicare & Medicaid Services
  
  
• Endorsement Status: Never Submitted
  
  
• Is the measure specified as an electronic clinical quality measure? No
  
  

Preliminary Analysis of Measure

• Preliminary analysis result: Encourage Continued Development
  
  
• Preliminary analysis summary
  • Contribution to program measure set:This measure has high value potential for the progam measure set and can allow for better complaince with ventilator process elements during LTCH stay, improve patient safety, health outcomes, and decrease costs at the facility level.
  • Impact on quality of care for patients:The potential impact of this measure is to improve patient safety outcomes and decrease costs. This measure was reviewed by the MAP PAC/LTC Workgroup in the 2014-2015 Pre-rulemaking Cycle (#X3705). The workgroup encouraged this measure for continued development because it addresses an important patient safety priority for LTCHs. It is estimated that 25% of ventilated patients in LTCHs acquire ventilator-associated pneumonia. There is evidence for interventions developed to decrease incidence of ventilator-associated pneumonia and improve ventilator care. VAP and VAE are associated with substantial morbidity, mortality, and excess healthcare costs. Furthermore, during the public comment period, MAP received two comments in support of MAP's recommendation noting its importance to patient safety and suggesting that the measure be further developed with adjustment for sociodemographic status.
  
  
  
• Does the measure address a program goal or objective? Yes. This measure addresses a high priority domain of the program (effective prevention and treatment related to ventilaror use), supports alignment across programs under the IMPACT Act, and addresses program gap in measurement.
  
  
• Is this a high-value measure? Yes. This process measure addresses an important patient safety priority for LTCHs, is non-duplicative of what is already in the program, and captures a broad population in the LTCH setting.
  
  
• Does this measure fill a gap in the program measure set? Yes. This measure fills a gap in the program as it relates to quality of care for patients admitted to LTCH for ventilator weaning which represents a large population in the LTCH setting. Currently, there is only one other measure in the program related to ventilator use: NQF#2632 Functional Outcome Measure Change in Mobility Among LTCH Patients Requiring Ventilator Support. This is an NQF endorsed measure that addresses the change in mobility score between admission and discharge among patients requiring ventilator support at admission. Additionally, the National Healthcare Safety Network (NHSN)Ventilator-Associated Event (VAE) Outcome Measure assesses ventilator associated infection rates. This compliance measure can allow for improved quality and outcomes for patients on ventilators.
  
  
• Measure development status: Early Development
  
  
• Is the measure fully tested for the program's setting and level of analysis? Yes. The level of analysis is at the facility level within a LTCH.
  
  
• Is the measure currently in use? No. New measure, never used in a program.
  
  
• Does the MUC contribute to alignment and efficient use of measurement resources (burden and cost of measurement)? Yes. This measure supports alignment by filling a gap in ventilator related measures in the program, addresses an NQS priority for making care safer by reducing harm caused in the delivery of care, and targets a large portion of the patient population in the LTCH setting. This measure does not require a burden on data collection. It uses the LTCH Care Data Set as an assessment tool for analysis.
  
  
• Is the measure NQF endorsed for the program's setting and level of analysis? No. Never Submitted
  
  
• Does the measure address a high-priority quality issue in the dual eligible beneficiary population? No .
  
  

Rationale for measure provided by HHS
Patients on invasive mechanical ventilation comprise a substantial proportion of LTCH patient admissions, and thus present a critical focus for assessment of high quality care. In Fiscal Year 2012, the LTCH MS-DRGs for “Respiratory system diagnosis with ventilator support 96+ hours” (MS-DRG-LTCH 207) and “Respiratory system diagnosis with ventilator support < 96 hours” (MS-DRG-LTCH 208) accounted for over 16,000 discharges, or greater than 13% of discharges. (MedPAC 2014). Mechanically ventilated patients are at higher risk of mortality, ventilator-associated pneumonia (Cook et al, 1998; Papazian et al., 1996; Vincent et al., 1995), delirium (Ely et al., 2001), ventilator associated lung injury (Meade et al., 1995 and 1997; Slutsky and Trembley, 1998), and other ventilator-associated events. The cost of invasive mechanical ventilation in LTCHs is considerable, estimated at $1.3 billion in 2006 (Kahn et al., 2010). Discontinuation of invasive mechanical ventilation is associated with improved patient outcomes, including lower post-discharge mortality (Aboussouan et al. 2008; Dermot Frengley et al. 2014; Hassenpflug, Steckart, and Nelson 2011). Citations: Aboussouan, L. S., Lattin, C. D., and Kline, J. L. (2008). 'Determinants of long-term mortality after prolonged mechanical ventilation'. Lung 186 (5):299-306, doi 10.1007/s00408-008-9110-x. Cook, D. J., Walter, S. D., Cook, R. J., Griffith, L. E., Guyatt, G. H., Leasa, D., Jaeschke, R. Z., and Brun-Buisson, C. (1998). 'Incidence of and risk factors for ventilator-associated pneumonia in critically ill patients'. Ann Intern Med 129 (6):433-40. Dermot Frengley, J., Sansone, G. R., Shakya, K., and Kaner, R. J. (2014). 'Prolonged mechanical ventilation in 540 seriously ill older adults: effects of increasing age on clinical outcomes and survival'. J Am Geriatr Soc 62 (1):1-9, doi 10.1111/jgs.12597. Ely EW, Inouye SK, Bernard GR, et al. Delirium in mechanically ventilated patients: validity and reliability of the confusion assessment method for the intensive care unit (CAM-ICU). JAMA. 2001 Dec 5;286(21):2703-10. PMID: 11730446. Hassenpflug, M., Steckart, J., and Nelson, D. (2011). Post-ICU Mechanical Ventilation: Extended Care Facility Residents Transferred From Intensive Care To Long-Term Acute Care. In, American Thoracic Society 2011 International Conference. Denver, Colorado. Kahn, J. M., Benson, N. M., Appleby, D., Carson, S. S., and Iwashyna, T. J. (2010). 'Long-term acute care hospital utilization after critical illness'. JAMA 303 (22):2253-9, doi 10.1001/jama.2010.761. Meade, M. O., and Cook, D. J. (1995). 'The aetiology, consequences and prevention of barotrauma: a critical review of the literature'. Clin Intensive Care 6 (4):166-73. Meade, M. O., Cook, D. J., Kernerman, P., and Bernard, G. (1997). 'How to use articles about harm: the relationship between high tidal volumes, ventilating pressures, and ventilator-induced lung injury'. Crit Care Med 25 (11):1915-22. MedPAC. (2014). Chapter 11. Long-term Care Hospital Services. In: Report to the Congress: Medicare Payment Policy. In. Medicare Payment Advisory Commission, Washington, DC. Papazian, L., Bregeon, F., Thirion, X., Gregoire, R., Saux, P., Denis, J. P., Perin, G., Charrel, J., Dumon, J. F., Affray, J. P., and Gouin, F. (1996). 'Effect of ventilator-associated pneumonia on mortality and morbidity'. Am J Respir Crit Care Med 154 (1):91-7, doi 10.1164/ajrccm.154.1.8680705. Slutsky, A. S., and Tremblay, L. N. (1998). 'Multiple system organ failure. Is mechanical ventilation a contributing factor?'. Am J Respir Crit Care Med 157 (6 Pt 1):1721-5, doi 10.1164/ajrccm.157.6.9709092. Vincent, J. L., Bihari, D. J., Suter, P. M., Bruining, H. A., White, J., Nicolas-Chanoin, M. H., Wolff, M., Spencer, R. C., and Hemmer, M. (1995). 'The prevalence of nosocomial infection in intensive care units in Europe. Results of the European Prevalence of Infection in Intensive Care (EPIC) Study. EPIC International Advisory Committee'. JAMA 274 (8):639-44.

Measure Specifications

• NQF Number (if applicable):
  
  
• Description: This measure describes the risk-standardized rate of Medicare fee-for-service (FFS) patients/residents/persons who are discharged to the community following a post-acute stay/episode, and do not have an unplanned (re)admission to an acute care hospital or LTCH in the 31 days following discharge to community, and remain alive during the 31 days following discharge to community.
  
  
• Numerator: This measure does not have a simple form for the numerator and denominator. The numerator is defined as the risk-adjusted estimate of the number of patients/residents/persons included in the measure who are discharged to the community, and do not have an unplanned (re)admission to an acute care hospital or LTCH on the day of discharge or in the 31 days following discharge to community, and remain alive during the 31 days following discharge to community. The numerator estimate includes risk adjustment for patient/resident/person characteristics, and a statistical estimate of the facility/agency effect beyond case mix.
  
  
• Denominator: The denominator is computed with the same model used for the numerator. It is the model developed using all non-excluded facility/agency stays/episodes in the national data. The measure includes all facility/agency stays/episodes in the measurement period that are observed in national Medicare FFS data and do not fall into an excluded category. For a particular facility/agency, the model is applied to the patient/resident/person population, but the facility/agency effect term is 0. In essence, it is the number of discharges to community that would be expected for that patient/resident/person population at the average facility/agency.
  
  
• Exclusions: i) Age under 18 years; (ii) No short-term acute care stay within 30 days prior to LTCH admission ; (iii) Discharges to psychiatric hospital; (iv) Discharges against medical advice; (v) Discharges to federal hospitals or disaster alternative care sites; (vi) Patients not continuously enrolled in Part A FFS Medicare for the 12 months prior to the LTCH stay admission date, and at least 31 days after PAC discharge date; (vii) Patients whose prior short-term acute-care stay was for non-surgical treatment of cancer; (viii) Discharges to hospice; (ix) LTCH stays that end in transfer to another LTCH; (x) LTCH stays with claims data that are problematic (e.g., anomalous records for stays that overlap wholly or in part, or are otherwise erroneous or contradictory); (xi) Patients who received care from a provider located outside of the US, Puerto Rico, or a US territory.
  
  
• HHS NQS Priority: Making Care Safer, Patient and Family Engagement, Making Care Affordable
  
  
• HHS Data Source: Claims
  
  
• Measure Type: Outcome
  
  
• Steward: Centers for Medicare & Medicaid Services
  
  
• Endorsement Status: Never Submitted
  
  
• Is the measure specified as an electronic clinical quality measure? No
  
  

Preliminary Analysis of Measure

• Preliminary analysis result: Encourage Continued Development
  
  
• Preliminary analysis summary
  • Contribution to program measure set:This is a high value measure that addresses the multifaceted care coordination needs of discharge patients to the community. Although the developers noted anticipated performance gaps in the measure due to variation in discharge and readmission rates based on facility and patient characteristics, this measure has the potential to assess the degree to which patients who are not prepared to live in a community are being inappropriately discharged.
  • Impact on quality of care for patients:The potential impact of this measure could lead to improved discharge to community rates, decreased costs, and increases quality of care within the facility to ensure patients are appropriately discharged. This measure can also support the care coordination needs of the patients upon discharge and improve overall patient health outcomes.
  
  
  
• Does the measure address a program goal or objective? Yes. This measure addresses the Communication/Care Coordination high priority domain for the LTCH QRP which assesses patient care transitions and rehospitaltizations, discharge to the community, and potentially preventable readmissions. This measure also supports alignment across programs under the IMPACT Act, and addresses program gap in measurement.
  
  
• Is this a high-value measure? Yes. This is a cost and resource use measure that assesses quality of care in the facility, proper discharge planning to the community, and patient goal attainment.
  
  
• Does this measure fill a gap in the program measure set? Yes. Currently there are two measures in the program that relate to readmissions including NQF#2631 Application of Percent of Long-Term Care Hospital Patients with an Admission and Discharge Functional Assessment and a Care Plan That Addresses Function and NQF#2512 All-Cause Unplanned Readmission Measure for 30 Days Post-Discharge from Long-Term Care Hospitals. MUC 414 promotes alignment with these two measures by assessing discharge patients with positive health outcomes once placed into the community. In addition, this measure relates to MUC 498 that assesses potentially avaoidable readmissions.
  
  
• Measure development status: Early Development
  
  
• Is the measure fully tested for the program's setting and level of analysis? No. No testing data available.
  
  
• Since no, could the measure be revised to use in the setting or at level of analysis under consideration? NA. NA
  
  
• Is the measure currently in use? No. New measure, never used in a program.
  
  
• Does the MUC contribute to alignment and efficient use of measurement resources (burden and cost of measurement)? Yes. This measure contributes to alignment by addressing the NQS priorities of making care safer by reducing harm caused in the delivery of care, ensuring that each person and family is engaged as partners in their care, and promoting effective communication and coordination of care. Data is collected through Medicare Claims Data and is an efficient use of measurement resources because this information is readily available.
  
  
• Is the measure NQF endorsed for the program's setting and level of analysis? No. Never Submitted
  
  
• Does the measure address a high-priority quality issue in the dual eligible beneficiary population? No.
  
  

Rationale for measure provided by HHS
The ultimate goals of post-acute care are avoiding institutionalization and returning patients to their previous level of independence and functioning, with discharge to community being the primary goal for the majority of post-acute patients. For many, home is a symbol of independence, privacy, and competence. Discharge to community is considered a valuable outcome to measure because it is a multifaceted measure that captures the patient’s functional status, cognitive capacity, physical ability, and availability of social support at home. There is considerable variation in discharge to community rates within and across post-acute settings. Studies show geographic variation, variation across patient socioeconomic characteristics (for example, race and ethnicity), and variation by facility characteristics (for profit vs. nonprofit, freestanding vs. hospital-based, urban vs. rural). In the IRF setting, discharge to community rates vary across providers, ranging from about 60% to 75%. The 2015 MedPAC report shows that, in FY 2013, the facility-level, mean risk-adjusted discharge to community rate for IRFs within 100 days of admission was 75.8%, and the mean observed rate was 74.7%. Discharge to community rates also vary widely in the SNF setting, ranging from as low as 31% to as high as 65%. The 2015 MedPAC report shows a mean risk-adjusted discharge to community rate of 37.5% for SNFs within 100 days of admission, and mean observed rate of 40.1%. A multicenter study of 23 LTCHs reported that only 28.8% of 1,061 patients who were ventilator-dependent on admission were discharged to home or assisted living facility. A study of 66,510 Medicare beneficiaries during pre- and post-HH episodes, revealed that 64 percent of beneficiaries discharged from HH did not use any other Medicare-reimbursed acute or post-acute services in the 30 days following HH discharge. Significant numbers of patients were admitted to inpatient facilities (29 percent) and lesser numbers to skilled nursing facilities (7.6 percent), inpatient rehabilitation (1.5 percent) and home health (7.2 percent) or hospice (3.3 percent) within 30 days of HH discharge (Wolff et al., 2008).

Measure Specifications

• NQF Number (if applicable):
  
  
• Description: Percentage of stays Inpatient Rehabilitation Facility (IRF), Long Term Care Facility (LTCH), and Skilled Nursing Facility (SNF) or care episodes Home Health (HH) in which a drug regimen review was conducted at the Admission (IRF, LTCH or SNF)/ Start of Care (SOC)/ Resumption of Care (ROC) (HH) and timely follow-up with a physician occurred each time potential clinically significant medication issues were identified throughout the stay (IRF, LTCH, or SNF) or care episode (HH).
  
  
• Numerator: Number of stays or care episodes where the medical record contains documentation of a drug regimen review conducted at admission or start-of-care or resumption-of-care with all potential clinically significant medication issues identified during the course of care and followed-up with a physician or physician designee.
  
  
• Denominator: Care episodes or stays ending during the reporting period. Assessment timing is as follows: Beginning of care episode or stay: • HH – SOC or ROC • SNF – Admission • IRF - Admission • LTCH – Admission End of care episode or stay: • HH – Transfer, Discharge, or Death at Home • SNF – Discharge, or expired • IRF – Discharge, or expired • LTCH – Discharge, or expired
  
  
• Exclusions: Denominator Exclusion: NONE Numerator Exclusion: NONE
  
  
• HHS NQS Priority: Making Care Safer, Communication and Care Coordination
  
  
• HHS Data Source: Patient assessment data
  
  
• Measure Type: Process
  
  
• Steward: Centers for Medicare & Medicaid Services
  
  
• Endorsement Status: Never Submitted
  
  
• Is the measure specified as an electronic clinical quality measure? No
  
  

Preliminary Analysis of Measure

• Preliminary analysis result: Encourage Continued Development
  
  
• Preliminary analysis summary
  • Contribution to program measure set:This measure of medication reconciliation addresses an important aspect of care for patients as they transition from one setting to another. During transitions of care, there can be many changes to the drug regimen. Identifying medication issues and resolving these with a physician or physician-designee are important steps to prevent issues such as; adverse drug reaction, ineffective drug therapy, drug interactions, duplicate therapy, dosage errors etc., any of which could lead to hospitalization or re-hospitalization. Medication reconciliation is currently one of the Joint Commission National Patient Safety Goals –which is to maintain and communicate accurate patient medication information.Currently; there are not medication management measures in the LTCH QRP. This measure fills a significant gap area and is expected to reduce hospitalizations, adverse events related to medications, and improve health outcomes.
  • Impact on quality of care for patients:The potential impact of this measure is to promote medication reconciliation and medication review for clinically significant issues. This measure is intended to improve health outcomes and quality of care for patients in the LTCH setting.
  
  
  
• Does the measure address a program goal or objective? Yes. This clinical measure addresses the program goals of effectiveness and safety.
  
  
• Is this a high-value measure? Yes. This is a process measure with strong links to outcomes. Medication review and reconcilliation can reduce hospitalizations and re-hospitalizations, reduce adverse events, and improve health.
  
  
• Does this measure fill a gap in the program measure set? Yes. This measure would address the high priority domain of communication and care coordination. CMS has noted the particular need for a measure of medication reconciliation for this program.
  
  
• Measure development status: Early Development
  
  
• Is the measure fully tested for the program's setting and level of analysis? No . Level of analysis and testing data not provided.
  
  
• Is the measure currently in use? No. New measure, never used in a program
  
  
• Does the MUC contribute to alignment and efficient use of measurement resources (burden and cost of measurement)? Yes. This measure is proposed to meet the requirements of the IMPACT Act and is in development to ensure alignment acorss PAC settings of care, thus should promote efficient use of resources.
  
  
• Is the measure NQF endorsed for the program's setting and level of analysis? No. Never Submitted
  
  
• Does the measure address a high-priority quality issue in the dual eligible beneficiary population? No.
  
  

Rationale for measure provided by HHS
Medication review in post-acute care is generally considered to include medication reconciliation for all medications and medication review for what poses as potential clinically significant medication issues for the patient/resident. As a process measure, medication reconciliation and medication review for potential clinically significant medication issues are expected to reduce re-hospitalizations, reduce adverse events related to medications and improve health outcomes.

Measure Specifications

• NQF Number (if applicable):
  
  
• Description: The MSPB-PAC Measure for LTCHs evaluates providers’ efficiency relative to the efficiency of the national median LTCH provider. Specifically, the MSPB-PAC Measure assesses the cost to Medicare for services during an episode of care, which consists of a treatment period and an associated services period. The episode is triggered by an admission to an LTCH stay. The treatment period begins at the trigger and ends at discharge. The Measure is constructed differently for cases in which the LTCH stay is paid according to the standard MS-LTC-DRG versus cases in which the LTCH stay is paid a site neutral rate comparable to the IPPS payment rates. The associated services period for standard payment rate cases begins at the trigger and ends 30 days after the end of the treatment period (i.e., discharge). The associated services period for site neutral payment rate cases begins at the close of the treatment period and ends 30 days after, to parallel the MSPB-Hospital measure. For the standard and site neutral cases, these periods constitute the episode window during which beneficiaries’ Medicare services are counted toward the episode. For the standard cases, the MSPB-PAC episode includes all services during the episode window that are attributable to the LTCH provider and those rendered by other providers, except those services during the associated services period that are clinically unrelated to LTCH responsibilities (e.g., planned care and routine screening). For the site neutral cases, the MSPB-PAC episode includes all services during the episode window that are attributable to the LTCH provider and those rendered by other providers, except those services during the associated services period that are clinically unrelated to LTCH responsibilities (e.g., planned care and routine screening). As discussed above, there is a difference in the construction of the associated services period for these cases, in that it only begins at discharge and ends 30 days after.
  
  
• Numerator: The numerator is the attributed provider’s average MSPB-PAC Amount. The MSPB-PAC Amount for each LTCH provider depends on two factors: i) the average of the ratio of standardized episode spending level and expected episode spending for each LTCH provider; and ii) the average standardized episode spending across all LTCH providers. To calculate the MSPB-PAC Amount for each LTCH, one finds the average of the ratio of the standardized episode spending over the expected episode spending, and then multiplies this quantity by the average episode spending level across all LTCHs.
  
  
• Denominator: The denominator for an LTCH’s MSPB-PAC Measure is the weighted median MSPB-PAC Amount across all episodes for LTCHs nationally.
  
  
• Exclusions: The measure excludes the following episodes: • Any episode that is triggered by a LTCH stay that happens outside the 50 states or DC. • Any episode that is triggered by a LTCH stay for which we see Part C crossover claims. • Any episode for which standard allowed amount of the LTCH stay could not be calculated or is equal to 0. • Any episode in which a beneficiary is not enrolled in Medicare Fee-for-Service for the entirety of the lookback period plus the episode window or is enrolled in Part C for any part of the lookback plus episode window. • Any episode in which a beneficiary has a primary payer other than Medicare for any part of the lookback plus episode window. • Any episode for which the lookback period extends beyond our observation period.
  
  
• HHS NQS Priority: Making Care Affordable
  
  
• HHS Data Source: Claims
  
  
• Measure Type: Cost/Resource Use
  
  
• Steward: Centers for Medicare & Medicaid Services
  
  
• Endorsement Status: Never Submitted
  
  
• Is the measure specified as an electronic clinical quality measure? No
  
  

Preliminary Analysis of Measure

• Preliminary analysis result: Encourage Continued Development
  
  
• Preliminary analysis summary
  • Contribution to program measure set:The potential value of this measure is to assess the level of health care costs in the MSPB-PAC population and increase alignment across high priority gap areas. This measure has the potential to identify the determining factors of high cost of health care and can lead to new efficiency measures that can result in reduced costs.
  • Impact on quality of care for patients:The potential impact of this measure is to promote efficiency and reduce the cost of spending in the MSPB-PAC population.
  
  
  
• Does the measure address a program goal or objective? Yes. Addresses a high priority domain (Making Care Affordable), supports alignment across programs under the IMPACT Act, and addresses program gap in measurement.
  
  
• Is this a high-value measure? Yes. This is an efficiency based cost and resource use measure that is non-duplicative and captures a broad populaiton the PAC/LTC setting.
  
  
• Does this measure fill a gap in the program measure set? Yes. Currently, there are no health care spending measures in the LTCH QRP indicating the level of cost for this population.
  
  
• Measure development status: Early Development
  
  
• Is the measure fully tested for the program's setting and level of analysis? Yes. The level of analysis is at the facility level within the PAC population.
  
  
• Is the measure currently in use? No. New measure, never used in a program
  
  
• Does the MUC contribute to alignment and efficient use of measurement resources (burden and cost of measurement)? Yes. This measure contributes to the alignment across all care settings under the IMPACT Act, addresses the NQS priorities of making quality care more affordable for individuals, families, employers, and governments by developing and spreading new health care delivery models. This measure contributes to the efficient use of resources by using claims data to test the measure.
  
  
• Is the measure NQF endorsed for the program's setting and level of analysis? No. Never Submitted
  
  
• Does the measure address a high-priority quality issue in the dual eligible beneficiary population? No.
  
  

Rationale for measure provided by HHS
Medicare payments to PAC have grown at a consistently higher rate than other major Medicare sectors. Between 2001 and 2013, Medicare PAC spending grew at an annual rate of 6.1 percent and doubled to $59.4 billion.

Measure Specifications

• NQF Number (if applicable):
  
  
• Description: This measure reports the percentage of patients in a Long Term Care Hospital who receive antipsychotic medications during the target period.
  
  
• Numerator: The numerator is the number of Long Term Care Hospital patients receiving antipsychotic medications.
  
  
• Denominator: The denominator is the total of all patients in a Long Term Care Hospital during the target period except for those who meet exclusion criteria.
  
  
• Exclusions: Patients will be excluded from the denominator if they are diagnosed with any of the following conditions: schizophrenia, Tourette's syndrome, and Huntington's Disease.
  
  
• HHS NQS Priority: Making Care Safer, Effective Prevention and Treatment
  
  
• HHS Data Source: LTCH CARE data set
  
  
• Measure Type: Process
  
  
• Steward: Centers for Medicare & Medicaid Services
  
  
• Endorsement Status: Never Submitted
  
  
• Is the measure specified as an electronic clinical quality measure? No
  
  

Preliminary Analysis of Measure

• Preliminary analysis result: Encourage Continued Development
  
  
• Preliminary analysis summary
  • Contribution to program measure set:Currently, there are not medication management measures in the LTCH QRP. This measure fills a significant gap area within a vulnerable population and is expected to reduce hospitalizations, adverse events related to medications, and improve health outcomes.
  • Impact on quality of care for patients:The potential impact of this measure is to promote medication management within the older adult population taking antipsychotic medications. This measure is intended to improve health outcomes and quality of care for patients in the LTCH setting.This measure is similar to a currently endorsed NQF measure: Antipsychotic Use in Persons with Dementia (#2111),a health plan Part D measure and consideration may be given to alignment/harmonization since these measures have congruent rationale.
  
  
  
• Does the measure address a program goal or objective? Yes. Addresses a high priority domain (Communication/Care Coordination), supports alignment across programs under the IMPACT Act, and addresses program gap in measurement.
  
  
• Is this a high-value measure? Yes. Thi is a high value process measure that is not duplicative and captures a broad population in the LTCH setting.
  
  
• Does this measure fill a gap in the program measure set? Yes. Currently there are no medication management measures in the LTCH QRP.
  
  
• Measure development status: Early Development
  
  
• Is the measure fully tested for the program's setting and level of analysis? Yes. The level of analysis is at the clinician level within the LTCH setting.
  
  
• Is the measure currently in use? Yes. This measure is currently in use in SNF QRP (2012-Present) and is being proposed for the LTCH QRP
  
  
• Does a review of its performance history raise any concerns? Yes. The measure target population includes all long term care hospitals. The developers noted the high possibility of a performance gap due to increased use of antipsychotic medications in the older adult population and variation in facility level administration of medication and prescribing rates. This could lead to unintended consequences such as avoiding admission of patients with diagnoses that may put them at risk for needing antipsychotic medication.
  
  
• Does the MUC contribute to alignment and efficient use of measurement resources (burden and cost of measurement)? Yes. This measure contributes to the alignment across all care settings under the IMPACT Act, addresses two NQS priorities of making care safer by reducing harm caused in the delivery of care and promoting effective communication and coordination of care, starting with cardiovascular disease. This measure also contributes to the efficient use of measurement resource and will use LTCH CARE Data Set for testing analysis.
  
  
• Is the measure NQF endorsed for the program's setting and level of analysis? No. Never Submitted
  
  
• Does the measure address a high-priority quality issue in the dual eligible beneficiary population? No.
  
  

Rationale for measure provided by HHS
Antipsychotic medication use is common among older adults in all post acute care settings. Antipsychotic medications can be potentially dangerous for the elderly, especially for those whom the medications are clinically indicated. Of particular concern is the off-label use of these drugs for older adults with dementia or dementia-related psychoses or agitation (Jeste et al., 2008). The FDA issued a black box warning against prescribing atypical antipsychotic medications for older adults with dementia in 2005 (Rosack, 2005). The evidence on which the warning is based on a meta-analysis of 17 randomized trials with a total of 5,106 patients that identified an “approximately 1.6- to 1.7-fold increase in mortality in the combined studies” (Rosack, 2005). Three years later, the FDA (June 2008) extended the warning to all categories of antipsychotic drugs (conventional & atypical). In addition to elevated mortality risk, elevated risk for serious adverse events such as falls, somnolence, and abnormal gait are results from clinical trials of atypical antipsychotic (AP) medications (Rosack, 2005; FDA, 2008; Ballard & Margallo-Lana, 2004; Martin et al., 2003; Neil, Curran, and Wattis, 2003; Doody et al., 2001; Jackson-Siegal, 2004). Also, there is evidence of increased risk for cerebrovascular adverse events associated with certain atypical antipsychotic medications (e.g., risperidone, olanzapine, and aripiprazole) (Jeste et al., 2008). Regardless of the warnings and potential adverse events, the administration of antipsychotic therapy is common and frequent among mechanically ventilated patients or among patients with delirium (Al-Qadheeb et al., 2013).

Measure Specifications

• NQF Number (if applicable):
  
  
• Description: All-condition risk-adjusted potentially preventable hospital readmission rates
  
  
• Numerator: This measure does not have a simple form for the numerator and denominator. The numerator is defined as the risk-adjusted estimate of the number of unplanned, potentially preventable readmissions that occurred within 30 days from LTCH discharge. The numerator, as defined, includes risk adjustment for patient characteristics and a statistical estimate of the facility effect beyond patient mix.
  
  
• Denominator: The denominator is computed with the same model used for the numerator. It is the model developed using all non-excluded stays in the national data. The measure includes all stays in the measurement period that are observed in national Medicare FFS data and do not fall into an excluded category. For a particular facility, the model is applied to the patient population, but the facility effect term is 0. In essence, it is the number of unplanned potentially preventable readmissions that would be expected for that patient population at the average facility
  
  
• Exclusions: 1. Patients who died during the LTCH stay. 2. Patients less than 18 years old. 3. Patients who were transferred to the same level of care or a hospital at the end of their LTCH stay 4. Patients who were not continuously enrolled in Part A FFS Medicare for the 12 months prior to the LTCH admissions, and at least 30 days after LTCH discharge. 5. Patients who did not have a short-term acute-care stay within 30 days prior to the LTCH admission date. 6. Patients who leave the LTCH against medical advice 7. Patients for whom the prior short-term acute-care stay was for the nonsurgical treatment of cancer. 8. Patients who were transferred to a federal hospital from the LTCH. 9. Patients who received care from a provider located outside of the US, Puerto Rico, or a US territory. 10. LTCH stays with data that are problematic (e.g., anomalous records for hospital stays that overlap wholly or in part or are otherwise erroneous or contradictory).
  
  
• HHS NQS Priority: Communication and Care Coordination
  
  
• HHS Data Source: Claims: This measure is based on Medicare claims data for fee-for-service beneficiaries using LTCH services. Several variables from the inpatient claims will be used including admission/discharge dates and diagnostic/procedure information.
  
  
• Measure Type: Outcome
  
  
• Steward: Centers for Medicare & Medicaid Services
  
  
• Endorsement Status: Never Submitted
  
  
• Is the measure specified as an electronic clinical quality measure? No
  
  

Preliminary Analysis of Measure

• Preliminary analysis result: Encourage Continued Development
  
  
• Preliminary analysis summary
  • Contribution to program measure set:This is a high value measure that is not duplicative, captures a broader population of condition specific readmissions.
  • Impact on quality of care for patients:The potential impact of this measure could lead to decreased readmissions rates, decreased costs, and increases quality of care within the facility to ensure patients are properly diagnosed and discharged if needed. This measure can also support the care coordination needs of the patients upon discharge. Overall, this measure has potential to improve patient outcomes upon discharge to their communities from LTCH.
  
  
  
• Does the measure address a program goal or objective? Yes. Addresses a high priority domain (Communication/Care Coordination), supports alignment across programs under the IMPACT Act, and addresses program gap in measurement.
  
  
• Is this a high-value measure? Yes. This is an outcome measure that addresses the Communication/Care Coordination high priority domain for future measure consideration under the program.
  
  
• Does this measure fill a gap in the program measure set? Yes. This measure compliments the two measures currently in the program (NQF#2512 and NQF#2631) concerning patient admissions/readmissions. This MUC fills a gap by aligning across all conditions and examines potentially avoidable conditions for patient readmissions.
  
  
• Measure development status: Early Development
  
  
• Is the measure fully tested for the program's setting and level of analysis? No. Testing data not available
  
  
• Since no, could the measure be revised to use in the setting or at level of analysis under consideration? NA. NA
  
  
• Is the measure currently in use? No. New measure, never used in a program
  
  
• Does the MUC contribute to alignment and efficient use of measurement resources (burden and cost of measurement)? Yes. The measure is based on Medicare claims data for FFS beneficiaries using LTCH servicesHarmonized across different setting according to the IMPACT Act. Not duplicative due to specific care setting.
  
  
• Is the measure NQF endorsed for the program's setting and level of analysis? No. Never Submitted
  
  
• Does the measure address a high-priority quality issue in the dual eligible beneficiary population? Yes.
  
  

Rationale for measure provided by HHS
The peer-reviewed literature specific to potentially preventable readmissions following LTCH discharge is limited. However, MedPAC has estimated that 76 percent of 30-day readmissions for Medicare beneficiaries overall were due to five potentially preventable conditions (heart failure, electrolyte imbalance, respiratory infection, sepsis, and urinary tract infection (MedPAC 2007).

Measure Specifications

• NQF Number (if applicable):
  
  
• Description: For patients admitted to an LTCH on invasive mechanical ventilation support and for whom weaning attempts were expected or anticipated at admission, this measure reports: (1) percentage of patients fully weaned at discharge (alive) (Ventilator Weaning/Liberation Rate), and (2) percentage of patients not fully weaned at discharge (alive).
  
  
• Numerator: The numerator represents the number of patients within each category of weaning status at discharge. The numerator will be calculated separately according to each of the measure component groups below. Each numerator component is the number of patients in the following categories: (1) the number of patients reported as fully weaned on planned or unplanned discharge assessment (2) the number of patients reported as not fully weaned on planned or unplanned discharge assessment. A patient is considered fully weaned if s/he does not require any invasive mechanical ventilation support for at least 2 consecutive calendar days immediately prior to the date of discharge (alive or dead) from an LTCH.
  
  
• Denominator: The target population (denominator) for this measure is the total number of patients who were discharged (alive or dead) from an LTCH during the reporting period and who were on invasive mechanical ventilation support upon admission to the LTCH, for whom weaning attempts were expected or anticipated at admission.
  
  
• Exclusions: This measure excludes patients with missing data and invasively mechanically ventilated patients identified as non-weaning at the time of admission to an LTCH. Patients who may be considered non-weaning include patients who are considered chronically ventilated as defined by evidence-based guidelines for ventilator liberation or patients with an acute or chronic condition that negates at admission any expectation or anticipation of weaning attempts (e.g. progressive neuromuscular disease such amyotrophic lateral sclerosis, or irreversible neurological injury or disease or dysfunction such as high (C2) spinal cord injury). Consideration of a patient as non-weaning must be based on documentation found in the patient’s medical record by Day 2 of LTCH Stay.
  
  
• HHS NQS Priority: Making Care Safer
  
  
• HHS Data Source: LTCH CARE data set. Assessment items related to mechanical ventilation support will be added to the current assessment tool for this setting (LTCH CARE Data Set) to ensure operationalization of this measure.
  
  
• Measure Type: Outcome
  
  
• Steward: Centers for Medicare & Medicaid Services
  
  
• Endorsement Status: Never Submitted
  
  
• Is the measure specified as an electronic clinical quality measure? No
  
  

Preliminary Analysis of Measure

• Preliminary analysis result: Encourage Continued Development
  
  
• Preliminary analysis summary
  • Contribution to program measure set:This measure has high value potential for the program because successful weaning is associated with decreased morbidity, mortality, and resource use.
  • Impact on quality of care for patients:The potential impact of this measure is to improve patient safety outcomes and decrease costs. This measure was reviewed by the MAP PAC/LTC Workgroup in the 2014-2015 Pre-rulemaking cycle. This measure addresses an important safety priority for LTCHs. MedPAC estimates that 16% of LTCH patients use ventilator services. Weaning is the process of decreasing the amount of support a patient receives from the ventilator. Furthermore, during the public comment period, MAP received two comments in support of MAP's recommendation noting its importance to patient safety and suggesting that the measure be further developed with adjustment for sociodemographic status.Additional impact data may be available from the CMS measure developer contractor for this measure, RTI, as noted in the MUC form.
  
  
  
• Does the measure address a program goal or objective? Yes. This measure addresses a high priority domain of the program (Effective Prevention and Treatment related to ventilaror use), supports alignment across programs under the IMPACT Act, and addresses program gap in measurement.
  
  
• Is this a high-value measure? Yes. This is a high value outcome measure that addresses an important patient safety priority for LTCHs, is non-duplicative of what is already in the program, and captures a broad population in the LTCH setting.
  
  
• Does this measure fill a gap in the program measure set? Yes. This measure fills a gap in the program as it assesses the quality of care for patients admitted to LTCH on invasive mechanical ventilation support for ventilator weaning which represents a large population in the LTCH setting. Currently, there is only one other measure in the program related to ventilator use: NQF#2632 Functional Outcome Measure Change in Mobility Among LTCH Patients Requiring Ventilator Support. This is an NQF endorsed measure that addresses the change in mobility score between admission and discharge among patients requiring ventilator support at admission. Additionally, the National Healthcare Safety Network (NHSN)Ventilator-Associated Event (VAE) Outcome Measure assesses ventilator associated infection rates. This measure also relates to MUC 400 Compliance with Ventilator Process Elements during LTCH stay and can allow for improved quality and outcomes for patients on ventilators.
  
  
• Measure development status: Early Development. The information provided by CMS indicates updates on testing will be provided in September 2015; this information has not been received by NQF to date.
  
  
• Is the measure fully tested for the program's setting and level of analysis? Yes. The level of analysis is at the facility level within a LTCH.
  
  
• Is the measure currently in use? No. New measure, never used in a program.
  
  
• Does the MUC contribute to alignment and efficient use of measurement resources (burden and cost of measurement)? Yes. This measure supports alignment by filling a gap in ventilator related measures in the program, addresses an NQS priority for making care safer by reducing harm caused in the delivery of care, and targets a large portion of the patient population in the LTCH setting. This measure does not require a burden on data collection. It uses the LTCH Care Data Set as an assessment tool for analysis.
  
  
• Is the measure NQF endorsed for the program's setting and level of analysis? No. Never Submitted
  
  
• Does the measure address a high-priority quality issue in the dual eligible beneficiary population? No.
  
  

Rationale for measure provided by HHS
Patients on invasive mechanical ventilation comprise a substantial proportion of LTCH patient admissions, and thus present a critical focus for assessment of high quality care. In Fiscal Year 2012, the LTCH MS-DRGs for “Respiratory system diagnosis with ventilator support 96+ hours” (MS-DRG-LTCH 207) and “Respiratory system diagnosis with ventilator support < 96 hours” (MS-DRG-LTCH 208) accounted for over 16,000 discharges, or greater than 13% of discharges. (MedPAC 2014). Mechanically ventilated patients are at higher risk of mortality, ventilator-associated pneumonia (Cook et al, 1998; Papazian et al., 1996; Vincent et al., 1995), delirium (Ely et al., 2001), ventilator associated lung injury (Meade et al., 1995 and 1997; Slutsky and Trembley, 1998), and other ventilator-associated events. The cost of invasive mechanical ventilation in LTCHs is considerable, estimated at $1.3 billion in 2006 (Kahn et al., 2010). Discontinuation of invasive mechanical ventilation is associated with improved patient outcomes, including lower post-discharge mortality (Aboussouan et al. 2008; Dermot Frengley et al. 2014; Hassenpflug, Steckart, and Nelson 2011). Citations: Aboussouan, L. S., Lattin, C. D., and Kline, J. L. (2008). 'Determinants of long-term mortality after prolonged mechanical ventilation'. Lung 186 (5):299-306, doi 10.1007/s00408-008-9110-x. Cook, D. J., Walter, S. D., Cook, R. J., Griffith, L. E., Guyatt, G. H., Leasa, D., Jaeschke, R. Z., and Brun-Buisson, C. (1998). 'Incidence of and risk factors for ventilator-associated pneumonia in critically ill patients'. Ann Intern Med 129 (6):433-40. Dermot Frengley, J., Sansone, G. R., Shakya, K., and Kaner, R. J. (2014). 'Prolonged mechanical ventilation in 540 seriously ill older adults: effects of increasing age on clinical outcomes and survival'. J Am Geriatr Soc 62 (1):1-9, doi 10.1111/jgs.12597. Ely EW, Inouye SK, Bernard GR, et al. Delirium in mechanically ventilated patients: validity and reliability of the confusion assessment method for the intensive care unit (CAM-ICU). JAMA. 2001 Dec 5;286(21):2703-10. PMID: 11730446. Hassenpflug, M., Steckart, J., and Nelson, D. (2011). Post-ICU Mechanical Ventilation: Extended Care Facility Residents Transferred From Intensive Care To Long-Term Acute Care. In, American Thoracic Society 2011 International Conference. Denver, Colorado. Kahn, J. M., Benson, N. M., Appleby, D., Carson, S. S., and Iwashyna, T. J. (2010). 'Long-term acute care hospital utilization after critical illness'. JAMA 303 (22):2253-9, doi 10.1001/jama.2010.761. Meade, M. O., and Cook, D. J. (1995). 'The aetiology, consequences and prevention of barotrauma: a critical review of the literature'. Clin Intensive Care 6 (4):166-73. Meade, M. O., Cook, D. J., Kernerman, P., and Bernard, G. (1997). 'How to use articles about harm: the relationship between high tidal volumes, ventilating pressures, and ventilator-induced lung injury'. Crit Care Med 25 (11):1915-22. MedPAC. (2014). Chapter 11. Long-term Care Hospital Services. In: Report to the Congress: Medicare Payment Policy. In. Medicare Payment Advisory Commission, Washington, DC. Papazian, L., Bregeon, F., Thirion, X., Gregoire, R., Saux, P., Denis, J. P., Perin, G., Charrel, J., Dumon, J. F., Affray, J. P., and Gouin, F. (1996). 'Effect of ventilator-associated pneumonia on mortality and morbidity'. Am J Respir Crit Care Med 154 (1):91-7, doi 10.1164/ajrccm.154.1.8680705. Slutsky, A. S., and Tremblay, L. N. (1998). 'Multiple system organ failure. Is mechanical ventilation a contributing factor?'. Am J Respir Crit Care Med 157 (6 Pt 1):1721-5, doi 10.1164/ajrccm.157.6.9709092. Vincent, J. L., Bihari, D. J., Suter, P. M., Bruining, H. A., White, J., Nicolas-Chanoin, M. H., Wolff, M., Spencer, R. C., and Hemmer, M. (1995). 'The prevalence of nosocomial infection in intensive care units in Europe. Results of the European Prevalence of Infection in Intensive Care (EPIC) Study. EPIC International Advisory Committee'. JAMA 274 (8):639-44.

Measure Specifications

• NQF Number (if applicable): 2634
  
  
• Description: This quality measure estimates the risk-adjusted mean change in mobility score between admission and discharge among Skilled Nursing Facility residents. (The endorsed specifications of the measure are: This measure estimates the mean risk-adjusted mean change in mobility score between admission and discharge for Inpatient Rehabilitation Facility (IRF) Medicare patients.)
  
  
• Numerator: The measure does not have a simple form for the numerator and denominator. This measure estimates the risk-adjusted change in mobility score between admission and discharge among Skilled Nursing Facility residents age 21 and older. The change in mobility score is calculated as the difference between the discharge mobility score and the admission mobility score. (The endorsed specifications of the measure are: The measure does not have a simple form for the numerator and denominator. This measure estimates the risk-adjusted change in mobility score between admission and discharge among Inpatient Rehabilitation Facility (IRF) patients age 21 and older. The change in mobility score is calculated as the difference between the discharge mobility score and the admission mobility score.)
  
  
• Denominator: Skilled Nursing Facility residents included in this measure are at least 21 years of age, Medicare Fee-for-Service beneficiaries, are not independent with all of the mobility activities at the time of admission, and have complete stays.(The endorsed specifications of the measure are: Inpatient Rehabilitation Facility patients included in this measure are at least 21 years of age, Medicare beneficiaries, are not independent with all of the mobility activities at the time of admission, and have complete stays.)
  
  
• Exclusions: This quality measure has 8 exclusion criteria: 1. Residents with incomplete stays 2. Residents who are independent with all mobility activities at the time of admission 3. Residents with the following medical conditions: coma; persistent vegetative state; complete tetraplegia; locked-in syndrome; severe anoxic brain damage, cerebral edema, or compression of brain 4. Residents younger than 21 years 5. Residents discharged to hospice 6. residents who are not Medicare Fee-for-Service beneficiaries 7. Residents in swing beds in critical access hospitals 8. Residents who do not receive rehabilitation therapy services(The endorsed specifications of the measure are: This quality measure has 5 exclusion criteria:1) Patients with incomplete stays. Rationale: It can be challenging to gather accurate discharge functional status data for patients who experience incomplete stays. Patients with incomplete stays include patients who are unexpectedly discharged to an acute care setting (Short-stay Acute Hospital, Critical Access Hospital, Inpatient Psychiatric Facility, or Long-term Care Hospital) because of a medical emergency; patients who die or leave an Inpatient Rehabilitation Facility (IRF) against medical advice; and patients with a length of stay less than 3 days. 2) Patients who are independent with all mobility activities at the time of admission. Rationale: Patients who are independent with CARE mobility items at the time of admission are assigned the highest score on all the mobility items, and thus, would not be able to show functional improvement on this same set of items at discharge. 3) Patients with the following medical conditions: coma, persistent vegetative state; complete tetraplegia; locked-in syndrome or severe anoxic brain damage, cerebral edema or compression of brain.Rationale: These patients are excluded because they may have limited or less predictable mobility improvement with the selected mobility items.4) Patients younger than age 21.Rationale: There is only limited evidence published about functional outcomes for individuals younger than 21.5) Patients discharged to hospice.Rationale: Patient goals may change during the IRF stay.6) Patients not covered by the Medicare program.)
  
  
• HHS NQS Priority: Patient and Family Engagement, Communication and Care Coordination
  
  
• HHS Data Source: CARE data
  
  
• Measure Type: Outcome
  
  
• Steward: Centers for Medicare & Medicaid Services
  
  
• Endorsement Status: Never Submitted
  
  
• Is the measure specified as an electronic clinical quality measure? No
  
  

Preliminary Analysis of Measure

• Preliminary analysis result: Encourage Continued Development
  
  
• Preliminary analysis summary
  • Contribution to program measure set:Research has shown differences in SNF residents’ functional outcomes by geographic region and race/ethnicity after adjusting for key patient demographic characteristics and admission clinical status, which supports the need to monitor SNF residents’ functional outcomes.
  • Impact on quality of care for patients:The mobility and self-care quality measures will standardize the collection of functional status data, which can improve communication when patients are transferred between providers. Most SNF patients receive care in an acute care hospital prior to the SNF stay, and many SNF patients receive care from another provider after the SNF stay. Use of standardized clinical data to describe a patient´s status across providers can facilitate communication across providers. In describing the importance of functional status, the National Committee on Vital and Health Statistics Subcommittee on Health (2001) noted, “Information on functional status is becoming increasing essential for fostering healthy people and a healthy population. Achieving optimal health and well-being for Americans requires an understanding across the life space of the effects of people’s health conditions on their ability to do basic activities and participate in life situations, in other words, their functional status.” This quality measure will inform SNF providers about opportunities to improve care in the area of function and strengthen incentives for quality improvement related to patient function.
  
  
  
• Does the measure address a program goal or objective? Yes. The measure promotes alignment and standardization of patient assessment across PAC settings (IMPACT Act).
  
  
• Is this a high-value measure? Yes. This MUC is an outcome measure.
  
  
• Does this measure fill a gap in the program measure set? Yes. Functional status and goal attainment.
  
  
• Measure development status: Early Development
  
  
• Is the measure fully tested for the program's setting and level of analysis? Yes. Adaptation of endorsed IRF measure; currently in field-testing for SNF population.
  
  
• Is the measure currently in use? No. The measure from which this MUC is being adapted will be implemented in October, 2016.
  
  
• Does the MUC contribute to alignment and efficient use of measurement resources (burden and cost of measurement)? Yes. This MUC supports alignment and standardization of metrics across PAC/LTC programs.
  
  
• Is the measure NQF endorsed for the program's setting and level of analysis? No. Never Submitted
  
  
• Does the measure address a high-priority quality issue in the dual eligible beneficiary population? No.
  
  

Rationale for measure provided by HHS
During a Skilled Nursing Facility (SNF) stay, the goals of treatment include fostering the patient’s ability to manage his or her daily activities so that the patient can complete self-care and mobility activities as independently as possible and if feasible, return to a safe, active and productive life in a community-based setting. Previous research has found direct relationships between increased intense therapy services and improved functional outcomes in the SNF setting. Jette et. al (2005) found that higher physical and occupational therapy intensities were associated with greater odds of improving by at least 1 stage in the mobility and activities of daily living functional independence across each condition including patients with stroke, orthopedic conditions, and cardiovascular and pulmonary conditions. Similarly, a randomized control trial, of 26 SNF patients compared higher intensity rehabilitation to the standard-of-care found greater improvement for mobility activities including gait speed, longer walking distances, and a trend for improvement for self-care activities as measured by the Barthel index (Lenze et. al 2012). The mobility and self-care quality measures will standardize the collection of functional status data, which can improve communication when patients are transferred between providers. Most SNF patients receive care in an acute care hospital prior to the SNF stay, and many SNF patients receive care from another provider after the SNF stay. Use of standardized clinical data to describe a patient´s status across providers can facilitate communication across providers. In describing the importance of functional status, the National Committee on Vital and Health Statistics Subcommittee on Health (2001) noted, “Information on functional status is becoming increasing essential for fostering healthy people and a healthy population. Achieving optimal health and well-being for Americans requires an understanding across the life space of the effects of people’s health conditions on their ability to do basic activities and participate in life situations, in other words, their functional status.” This quality measure will inform SNF providers about opportunities to improve care in the area of function and strengthen incentives for quality improvement related to patient function. Jette, D. U., R. L. Warren, & C. Wirtalla. (2005). The relation between therapy intensity and outcomes of rehabilitation in skilled nursing facilities. Archives of Physical Medicine and Rehabilitation, 86 (3), 373-9. Lenze, E. J., Host, H. H., Hildebrand M. W., Morrow-Howell, N., Carpenter, B., Freedland, K. E., … Binder, E, F. (2012). Enhanced medical rehabilitation increases therapy intensity and engagement and improves functional outcomes in postacute rehabilitation of older adults: a randomized-controlled trial. Journal of the American Medical Directors Association. 13(8):708-12. National Committee on Vital and Health Statistics Subcommittee on Health. Classifying and Reporting Functional Status. 2001. Retrieved from http://www.ncvhs.hhs.gov/010617rp.pdf

Summary of NQF Endorsement Review

• Year of Most Recent Endorsement Review: 2015
  
  
• Project for Most Recent Endorsement Review: Person- and Family-Centered Care: Phase 2
  
  
• Review for Importance: The measure meets the Importance criteria (1a. Evidence, 1b. Performance Gap, 1c. High Impact) 1a. Evidence: Y-13; N-0; 1b. Performance Gap: H-3; M-8; L-2; I-0; 1c. High Priority: H-7; M-6; L-0; I-0 Rationale:
  • The developer noted that IRF measures are limited to Medicare only and that the Long-Term Care Hospital Quality Reporting Program was established as a Medicare program.
  • The Committee highlighted that there are talks about these quality measures becoming pay-for-performance measures; however, in IRFs there are currently requirements for pay for performance such as a two-percent reduction in payments for failure to submit certain quality data. The Committee questioned the connection between these specific measures and pay-for-performance measures. The developer clarified that the Inpatient Rehabilitation Quality Reporting Program assigns a penalty for failure to report, however it is not tied to a pay-for-performance program.
  
  
  
• Review for Scientific Acceptability: The measure meets the Scientific Acceptability criteria (2a. Reliability - precise specifications, testing; 2b. Validity - testing, threats to validity) 2a. Reliability: H-0; M-10; L-0; I-3 2b. Validity: H-1; M-9; L-1; I-2 Rationale:
  • The developers utilized different types of reliability including Inter-rater reliability and patient videos reliability. Items that did not test well during the PAC demo were not included. Test-retest reliability was not performed due to the instability of the patients’ function.
  • The Committee expressed concerns that reliability and validity data was at the care level and not at the facility level; however, since this is an outcome measure the Committee agreed that both reliability and validity should be considered moderate.
  • The developers confirmed that the data elements they are using in the risk adjustment model and that the observed or expected calculation comes from the assessment data and comorbidities from the claims data.
  
  
  
• Review for Feasibility: H-6; M-5; L-2; I-0 (4a. Clinical data generated during care delivery; 4b. Electronic sources; 4c.Susceptibility to inaccuracies/ unintended consequences identified 4d. Data collection strategy can be implemented) Rationale: The Committee questioned the length of time it takes to administer or grade the instrument. The developer noted that clinicians are assessing patients on the ability to complete the activities listed in the measure.
  
  
• Review for Usability: H-6; M-5; L-0; I-2 (Meaningful, understandable, and useful to the intended audiences for 3a. Public Reporting/Accountability and 3b. Quality Improvement) Rationale: The Committee had no concerns with the usability of the measure.
  
  
• Review for Related and Competing Measures:
  • The Committee considered this measure to potentially compete with 2321: Functional Change: Change in Mobility Score (UDSMR) and was asked to vote to determine whether these measures are directly competing and select the best in class measure. While the Committee agreed that these measures are competing, they did not achieve consensus on whether one measure is superior. Therefore, these measures will both continue on to NQF member vote as recommended and be brought forward to the CSAC on June 9th for resolution of the best-in-class decision.
  • The Committee also considered this measure to be related to 2636: Inpatient Rehabilitation Facility (IRF) Functional Outcome Measure: Discharge Mobility Score for Medical Rehabilitation Patients (CMS), however there were no recommendations for harmonization.
  
  
  
• Endorsement Public Comments: March 2, 2015- March 31, 2015

  Comments received: Measures 2634 and 2636 received two similar comments. The first commenter supported the underlying concept of the measures, stating that inpatient rehabilitation facilities need to be measured on outcomes based on functional improvement. However, the commenter suggested that an alternative measure that determines how the provider improved the patient’s life (mobility) would better incentivize a change in clinical practice and associated patient-level outcomes as opposed to measure 2634 and measure 2636. Another commenter concurred with the Committee’s concern with the validity and reliability of measures developed using a cross-sectional study design from a demonstration project, which did not follow the same patients across venues of care and thus limiting applicability across sites.

  NQF response: NQF is limited to reviewing measures that are submitted for endorsement. We have added this suggestion to the measure gap list in the report. Thank you for your comment.

  Developer response: Thank you for your comment. As discussed during the measure review on January 22, 2015 and documented in the Person- and Family-Centered Care Phase 2 Draft Report on page 11, the Post-Acute Care Payment Reform demonstration was a prospective cohort study, not a cross-sectional study. In addition to collecting admission and discharge data using the CARE Tool during the post-acute care stay, inpatient claims data for acute care stays prior to and following the post-acute care stay were linked to the CARE admission and discharge data. The reliability and validity of the CARE function items were presented and discussed during the January 21-22, 2015 meeting, and several committee members referred to our analysis as very good. We have also submitted provider-level reliability data to the committee for review, as requested during the January 21-22, 2015 meeting. The Improving Medicare Post Acute Care Transformation (IMPACT) Act directs the Secretary to specify quality measures on which PAC providers are required to submit standardized patient assessment data and other necessary data specified by the Secretary with respect to five quality domains, one of which is functional status, cognitive function, and changes in function and cognitive function.
  
  

• Endorsement Committee Recommendation: Standing Committee Recommendation for Endorsement: Y-11; N-2
  
  

Measure Specifications

• NQF Number (if applicable): 2633
  
  
• Description: This quality measure estimates the risk-adjusted mean change in self-care score between admission and discharge among SNF residents. (The endorsed specifications of the measure are: This measure estimates the risk-adjusted mean change in self-care score between admission and discharge for Inpatient Rehabilitation Facility (IRF) Medicare patients.)
  
  
• Numerator: The measure does not have a simple form for the numerator and denominator. This measure estimates the risk-adjusted change in self-care score between admission and discharge among SNF residents age 21 and older. The change in self-care score is calculated as the difference between the discharge self-care score and the admission self-care score. (The endorsed specifications of the measure are: The measure does not have a simple form for the numerator and denominator. This measure estimates the risk-adjusted change in self-care score between admission and discharge among Inpatient Rehabilitation Facility (IRF) Medicare patients age 21 or older. The change in self-care score is calculated as the difference between the discharge self-care score and the admission self-care score.)
  
  
• Denominator: SNF residents included in this measure are at least 21 years of age, Medicare Fee-for-Service beneficiaries, are not independent with all of the self-care activities at the time of admission, and have complete stays.(The endorsed specifications of the measure are: Inpatient Rehabilitation Facility patients included in this measure are at least 21 years of age, Medicare beneficiaries, are not independent on all of the self-care activities at the time of admission, and have complete stays.)
  
  
• Exclusions: This quality measure has 8 exclusion criteria: 1. Residents with incomplete stays 2. Residents who are independent with all self-care activities at the time of admission 3. Residents with the following medical conditions: coma; persistent vegetative state; complete tetraplegia; locked-in syndrome; severe anoxic brain damage, cerebral edema, or compression of brain 4. Residents younger than 21 years 5. Residents discharged to hospice 6. Residents who are not Medicare Fee-for-Service beneficiaries 7. Residents in swing beds in critical access hospitals 8. Residents who do not receive rehabilitation therapy services(The endorsed specifications of the measure are: This quality measure has 6 exclusion criteria:1) Patients with incomplete stays. Rationale: It can be challenging to gather accurate discharge functional status data for patients who experience incomplete stays. Patients with incomplete stays include patients who are unexpectedly discharged to an acute care setting (Short-stay Acute Hospital, Critical Access Hospital, Inpatient Psychiatric Facility, or Long-term Care Hospital), because of a medical emergency; patients who die or leave an Inpatient Rehabilitation Facility (IRF) against medical advice; patients discharged directly to another IRF and patients with a length of stay less than 3 days. 2) Patients who are independent with all self-care activities at the time of admission. Rationale: Patients who are independent with all the self-care items at the time of admission are assigned the highest score on all the self-care items, and thus, would not be able to show functional improvement on this same set of items at discharge. 3) Patients with the following medical conditions: coma; persistent vegetative state; complete tetraplegia; locked-in syndrome; or severe anoxic brain damage, cerebral edema or compression of brain.Rationale: These patients are excluded because they may have limited or less predictable mobility improvement with the selected self-care items.4) Patients younger than age 21.Rationale: There is only limited evidence published about functional outcomes for children.5) Patients discharged to hospice.Rationale: Patient goals may change during the IRF stay. 6) Patients who are not Medicare beneficiaries.Patients not covered by the Medicare program.)
  
  
• HHS NQS Priority: Patient and Family Engagement, Communication and Care Coordination
  
  
• HHS Data Source: CARE Data
  
  
• Measure Type: Outcome
  
  
• Steward: Centers for Medicare & Medicaid Services
  
  
• Endorsement Status: Never Submitted
  
  
• Is the measure specified as an electronic clinical quality measure? No
  
  

Preliminary Analysis of Measure

• Preliminary analysis result: Encourage Continued Development
  
  
• Preliminary analysis summary
  • Contribution to program measure set:Research has shown differences in SNF residents’ functional outcomes by geographic region and race/ethnicity after adjusting for key patient demographic characteristics and admission clinical status, which supports the need to monitor SNF residents’ functional outcomes.
  • Impact on quality of care for patients:The mobility and self-care quality measures will standardize the collection of functional status data, which can improve communication when patients are transferred between providers. Most SNF patients receive care in an acute care hospital prior to the SNF stay, and many SNF patients receive care from another provider after the SNF stay. Use of standardized clinical data to describe a patient´s status across providers can facilitate communication across providers. In describing the importance of functional status, the National Committee on Vital and Health Statistics Subcommittee on Health (2001) noted, “Information on functional status is becoming increasing essential for fostering healthy people and a healthy population. Achieving optimal health and well-being for Americans requires an understanding across the life space of the effects of people’s health conditions on their ability to do basic activities and participate in life situations, in other words, their functional status.” This quality measure will inform SNF providers about opportunities to improve care in the area of function and strengthen incentives for quality improvement related to patient function.
  
  
  
• Does the measure address a program goal or objective? Yes. The measure promotes alignment and standardization of patient assessment across PAC settings (IMPACT Act).
  
  
• Is this a high-value measure? Yes. This MUC is an outcome measure.
  
  
• Does this measure fill a gap in the program measure set? Yes. Functional status and goal attainment.
  
  
• Measure development status: Early Development
  
  
• Is the measure fully tested for the program's setting and level of analysis? Yes. Adaptation of endorsed IRF measure; currently in field-testing for SNF population.
  
  
• Is the measure currently in use? No. The measure from which this MUC is being adapted will be implemented in October, 2016.
  
  
• Does the MUC contribute to alignment and efficient use of measurement resources (burden and cost of measurement)? Yes. This MUC supports alignment and standardization of metrics across PAC/LTC programs.
  
  
• Is the measure NQF endorsed for the program's setting and level of analysis? No. Never Submitted
  
  
• Does the measure address a high-priority quality issue in the dual eligible beneficiary population? No.
  
  

Rationale for measure provided by HHS
During a Skilled Nursing Facility (SNF) stay, the goals of treatment include fostering the patient’s ability to manage his or her daily activities so that the patient can complete self-care and mobility activities as independently as possible and if feasible, return to a safe, active and productive life in a community-based setting. Previous research has found direct relationships between increased intense therapy services and improved functional outcomes in the SNF setting. Jette et. al (2005) found that higher physical and occupational therapy intensities were associated with greater odds of improving by at least 1 stage in the mobility and activities of daily living functional independence across each condition including patients with stroke, orthopedic conditions, and cardiovascular and pulmonary conditions. Similarly, a randomized control trial, of 26 SNF patients compared higher intensity rehabilitation to the standard-of-care found greater improvement for mobility activities including gait speed, longer walking distances, and a trend for improvement for self-care activities as measured by the Barthel index (Lenze et. al 2012). The mobility and self-care quality measures will standardize the collection of functional status data, which can improve communication when patients are transferred between providers. Most SNF patients receive care in an acute care hospital prior to the SNF stay, and many SNF patients receive care from another provider after the SNF stay. Use of standardized clinical data to describe a patient´s status across providers can facilitate communication across providers. In describing the importance of functional status, the National Committee on Vital and Health Statistics Subcommittee on Health (2001) noted, “Information on functional status is becoming increasing essential for fostering healthy people and a healthy population. Achieving optimal health and well-being for Americans requires an understanding across the life space of the effects of people’s health conditions on their ability to do basic activities and participate in life situations, in other words, their functional status.” This quality measure will inform SNF providers about opportunities to improve care in the area of function and strengthen incentives for quality improvement related to patient function. Jette, D. U., R. L. Warren, & C. Wirtalla. (2005). The relation between therapy intensity and outcomes of rehabilitation in skilled nursing facilities. Archives of Physical Medicine and Rehabilitation, 86 (3), 373-9. Lenze, E. J., Host, H. H., Hildebrand M. W., Morrow-Howell, N., Carpenter, B., Freedland, K. E., … Binder, E, F. (2012). Enhanced medical rehabilitation increases therapy intensity and engagement and improves functional outcomes in postacute rehabilitation of older adults: a randomized-controlled trial. Journal of the American Medical Directors Association. 13(8):708-12. National Committee on Vital and Health Statistics Subcommittee on Health. Classifying and Reporting Functional Status. 2001. Retrieved from http://www.ncvhs.hhs.gov/010617rp.pdf

Summary of NQF Endorsement Review

• Year of Most Recent Endorsement Review: 2015
  
  
• Project for Most Recent Endorsement Review: Person- and Family-Centered Care: Phase 2
  
  
• Review for Importance: The measure meets the Importance criteria (1a. Evidence, 1b. Performance Gap, 1c. High Impact) 1a. Evidence: Y-15; N-1; 1b. Performance Gap: H-2; M-12; L-1; I-1; 1c. High Priority: H-9; M-6; L-1; I-0 Rationale:
  • The Committee noted that the measure is proposed for use for Medicare only, and felt that this limits the use of the measure and potentially introduces duplication of efforts if using multiple tools for differing payer populations.
  • The Committee requested clarification on the intent of the measure and if it was a reflection of the care in the IRF or how the patient was prepare for integration back into the community. Specifically, they wanted to know if there is a connection between how a patient is doing at discharge and how they will do in the community. The developer indicated that information was provided in the supplemental information specific to the evidence behind the measure. CMS further explained this is another attempt to standardize measurement and allow tracking of patients as they traverse the care continuum and between settings. The measures allow the comparison of uniform assessment data, whether it’s self-care or mobility.
  • The Committee asked for the reasoning behind the proposal of four measures using essentially the same data. The developer indicated that when testing understanding of the measures with consumers, they were led to develop both a change score concept for use by facilities and then the percentage of patients that achieve a certain status to improve consumer understanding. They would have provided both in the same measure if the NQF submissions allowed. There was a suggestion that these two pairs of measures be considered “paired” measure to promote their use together. A member from the rehabilitation community indicated he would find the information provided from both levels of measurement useful. Internally they can be used for the facility for quality improvement and externally for use with consumers.
  • The Committee requested clarification of the 6-point measure scale. Based on input from an expert panel and comparison of current tools in use for similar purposes, the scale proposed was deemed the best fit for purpose. This became important because there is another tool in use by IRFs – the FIM– that is required for payment and uses a different scale; members indicated that facilities may find that confusing if there were different requirements for different programs. CMS indicated that a determination has not been made to convert to function items from the CARE Item Set [tool].
  
  
  
• Review for Scientific Acceptability: The measure meets the Scientific Acceptability criteria (2a. Reliability - precise specifications, testing; 2b. Validity - testing, threats to validity) 2a. Reliability: H-0; M-7; L-2; I-6 (consensus not reached) 2b. Validity: H-1; M-7; L-1; I-6 (consensus not reached) UPDATED VOTES FOR 2a. Reliability: H-5; M-10; L-3; I-0 2b. Validity: H-4; M-12; L-2; I-0 Rationale:
  • As raised with previous measures, the Committee indicated a strong interest in seeing scientific acceptability data at the facility level. A member notes that Crohnbach alphas provided are at the patient level. The developer indicated they could provide facility level error bars on splines for consideration.
  The Committee asked the developer to consider if it would be more accurate to assess change in function between admission and discharge versus coming up with an expected functional level and seeing if it could be achieved. The assumption is that the comparison to an expected score would be more game-able. The developer indicated they use every bit of data they have available and the true intent of the percent of patients measure is for consumer understandability.
  • The Committee acknowledged the wealth of data provided on the reliability and validity of the CARE tool. They continued to struggle with lack of data at the facility level. The developer directed the Committee to supplemental information they provided which may have come in after the Committee reviewed each measure. Supplemental information included the relationship between discharge scores and discharge back to the community and between CARE scores and length of stay.
  • The Committee noted that there was some data available, specifically generalized estimation equation data that have splines and error bars, and upon submission that data will be extremely helpful.
  • NQF staff clarified that this is not a unique situation and as measures become operationalized, more data becomes available and as this is a standing committee, that data will come back to this committee for further review. There is also the understanding that with the movement toward pay for performance, Committees want more data and NQF is trying to work those issues into the process.
  
  
  
• Review for Feasibility: H-4; M-8; L-3; I-0 (4a. Clinical data generated during care delivery; 4b. Electronic sources; 4c.Susceptibility to inaccuracies/ unintended consequences identified 4d. Data collection strategy can be implemented) Rationale: The Committee had no questions or concerns on the feasibility of this measure
  
  
• Review for Usability: H-3; M-7; L-3; I-2 (Meaningful, understandable, and useful to the intended audiences for 3a. Public Reporting/Accountability and 3b. Quality Improvement) Rationale: The Committee had no questions or concerns on the use and usability of this measure
  
  
• Review for Related and Competing Measures:
  • The Committee considered this measure to potentially competes with 2286: Functional Change: Change in Self-Care Score (UDSMR) and was asked to vote on these measures. While the Committee agreed that these measures are competing, the Committee did not achieve consensus on whether one measure is superior. Therefore, these measures will go forward to NQF Member vote as recommended and will be brought forward to the CSAC on June 9th for resolution of the best in class decision.
  • The Committee also considered this measure to be related to 2635: Inpatient Rehabilitation Facility (IRF) Functional Outcome Measure: Discharge Self-Care Score for Medical Rehabilitation Patients (CMS), however there were no recommendations for harmonization.
  
  
  
• Endorsement Public Comments: March 2, 2015- March 31, 2015 Comments received:
  • One commenter noted that these are important measures but they need to be analyzed and improved as additional data is collected. Another commenter concurred with the Committee’s concern with the validity and reliability of measures developed using a cross-sectional study design from a demonstration project, which did not follow the same patients across venues of care and thus limiting applicability across sites.
  Committee response:
  • The Committee requested additional information to allow for more comprehensive evaluation of the consensus not reached and not recommended measures. This additional information was discussed on the post-comment committee call and the Committee had an opportunity to re-vote on the applicable measures. This measure was recommended by the Committee after reviewing the additional information and the comments.
  Developer response:
  • Thank you for your comment. As discussed during the measure review on January 22, 2015 and documented in the Person- and Family-Centered Care Phase 2 Draft Report on page 11, the Post-Acute Care Payment Reform demonstration was a prospective cohort study, not a cross-sectional study. In addition to collecting admission and discharge data using the CARE Tool during the post-acute care stay, inpatient claims data for acute care stays prior to and following the post-acute care stay were linked to the CARE admission and discharge data. The reliability and validity of the CARE function items were presented and discussed during the January 21-22, 2015 meeting, and several committee members referred to our analysis as very good. We have also submitted provider-level reliability data to the committee for review, as requested during the January 21-22, 2015 meeting. The Improving Medicare Post Acute Care Transformation (IMPACT) Act directs the Secretary to specify quality measures on which PAC providers are required to submit standardized patient assessment data and other necessary data specified by the Secretary with respect to five quality domains, one of which is functional status, cognitive function, and changes in function and cognitive function.
  • The Post-Acute Care Payment Reform Demonstration was a prospective cohort study. It was not a cross-sectional study. For the study, data were collected at admission and discharge for each patient in the study. In addition, we collected interim assessment data for patients in the cost-resource utilization segment of the study. As part of the study, we also linked the CARE admission and discharge data with acute care and post-acute care claims data in order to examine episodes of care and post discharge readmissions. (B). The items and the summed self-care and mobility scores are statistically significantly associated with several outcomes, including length of stay and discharge destination. The admission IRF self-care and IRF mobility scores were moderately correlated with length of stay with coefficients of -0.463 (p < .0001) for self-care and -0.474 (p < .001) for mobility. As expected, the summed self-care and mobility discharge scores for patients who were discharged to home were significantly different than the scores of patients discharged to a long-term care/nursing home setting. The mean (standard deviation) discharge self-care score for patients going home and to long-term care/nursing home were 34.29 (7.04) and 24.57 (9.39), respectively. For mobility, the mean (standard deviation) scores were 57.35 (15.68) and 36.57 (15.07), respectively. The patients going home had higher scores, indicating more function, as we expected. (C). The CARE function items included in the 4 IRF quality measures and 2 LTCH quality measures have undergone validity testing. In addition to the results we present in our testing documentation, the data presented above (in 3b), we examined the relationship between the current functional assessment items and the CARE items for each PAC setting. The reports describing the testing are available at: http://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Post-Acute-Care-Quality-Initiatives/CARE-Item-Set-and-B-CARE.html.
  
  
  
• Endorsement Committee Recommendation: Standing Committee Recommendation for Endorsement: Y-10; N-5; Y-16; N-2
  
  

Measure Specifications

• NQF Number (if applicable): 2636
  
  
• Description: This quality measure estimates the percentage of Skilled Nursing Facility residents who meet or exceed an expected discharge mobility score. (The endorsed specifications of the measure are: This measure estimates the percentage IRF patients who meet or exceed an expected discharge mobility score.)
  
  
• Numerator: The numerator is the number of residents in a Skilled Nursing Facility with a discharge mobility score that is equal to or higher than a calculated expected discharge mobility score. (The endorsed specifications of the measure are: The numerator is the number of patients in an IRF with a discharge mobility score that is equal to or higher than a calculated expected discharge mobility score.)
  
  
• Denominator: Skilled Nursing Facility residents included in this measure are at least 21 years of age, Medicare Fee-for-Service beneficiaries, and have complete stays.(The endorsed specifications of the measure are: IRF patients included in this measure are at least 21 years of age, Medicare beneficiaries, and have complete stays.)
  
  
• Exclusions: This quality measure has 7 exclusion criteria: 1. Residents with incomplete stays 2. Residents with the following medical conditions: coma; persistent vegetative state; complete tetraplegia; locked-in syndrome; or severe anoxic brain damage, cerebral edema, or compression of the brain 3. Residents younger than age 21 4. Residents discharged to hospice 5. Residents not covered by the Medicare Fee-for-Service program 6. Residents in swing beds in critical access hospitals 7. Residents who do not receive rehabilitation therapy services(The endorsed specifications of the measure are: This quality measure has 4 exclusion criteria:1) Patients with incomplete stays. Rationale: It can be challenging to gather accurate discharge functional status data for patients who experience incomplete stays. Patients with incomplete stays include patients who are unexpectedly discharged to an acute care setting (Short-stay Acute Hospital, Critical Access Hospital, Inpatient Psychiatric Facility, or Long-term Care Hospital) because of a medical emergency; patients who die or leave an Inpatient Rehabilitation Facility (IRF) against medical advice; and patients with a length of stay less than 3 days. 2) Patients with the following medical conditions on admission: coma, persistent vegetative state, complete tetraplegia, locked-in syndrome, or severe anoxic brain damage, cerebral edema or compression of brain.Rationale: These patients are excluded because they may have limited or less predictable mobility improvement with the selected items.3) Patients younger than age 21.Rationale: There is only limited evidence published about functional outcomes for individuals younger than 21.4) Patients discharged to hospice.Rationale: Patient goals may change during the IRF stay. 5) Patients who are not Medicare beneficiaries.)
  
  
• HHS NQS Priority: Patient and Family Engagement, Communication and Care Coordination
  
  
• HHS Data Source: CARE data
  
  
• Measure Type: Outcome
  
  
• Steward: Centers for Medicare & Medicaid Services
  
  
• Endorsement Status: Never Submitted
  
  
• Is the measure specified as an electronic clinical quality measure? No
  
  

Preliminary Analysis of Measure

• Preliminary analysis result: Encourage Continued Development
  
  
• Preliminary analysis summary
  • Contribution to program measure set:Research has shown differences in SNF residents’ functional outcomes by geographic region and race/ethnicity after adjusting for key patient demographic characteristics and admission clinical status, which supports the need to monitor SNF residents’ functional outcomes.
  • Impact on quality of care for patients:The mobility and self-care quality measures will standardize the collection of functional status data, which can improve communication when patients are transferred between providers. Most SNF patients receive care in an acute care hospital prior to the SNF stay, and many SNF patients receive care from another provider after the SNF stay. Use of standardized clinical data to describe a patient´s status across providers can facilitate communication across providers. In describing the importance of functional status, the National Committee on Vital and Health Statistics Subcommittee on Health (2001) noted, “Information on functional status is becoming increasing essential for fostering healthy people and a healthy population. Achieving optimal health and well-being for Americans requires an understanding across the life space of the effects of people’s health conditions on their ability to do basic activities and participate in life situations, in other words, their functional status.” This quality measure will inform SNF providers about opportunities to improve care in the area of function and strengthen incentives for quality improvement related to patient function.
  
  
  
• Does the measure address a program goal or objective? Yes. The measure promotes alignment and standardization of patient assessment across PAC settings (IMPACT Act).
  
  
• Is this a high-value measure? Yes. This MUC is an outcome measure.
  
  
• Does this measure fill a gap in the program measure set? Yes. Functional status and goal attainment.
  
  
• Measure development status: Early Development
  
  
• Is the measure NQF endorsed for the program's setting and level of analysis? No. Never Submitted
  
  
• Does the measure address a high-priority quality issue in the dual eligible beneficiary population? No.
  
  

Rationale for measure provided by HHS
During a Skilled Nursing Facility (SNF) stay, the goals of treatment include fostering the patient’s ability to manage his or her daily activities so that the patient can complete self-care and mobility activities as independently as possible and if feasible, return to a safe, active and productive life in a community-based setting. Previous research has found direct relationships between increased intense therapy services and improved functional outcomes in the SNF setting. Jette et. al (2005) found that higher physical and occupational therapy intensities were associated with greater odds of improving by at least 1 stage in the mobility and activities of daily living functional independence across each condition including patients with stroke, orthopedic conditions, and cardiovascular and pulmonary conditions. Similarly, a randomized control trial, of 26 SNF patients compared higher intensity rehabilitation to the standard-of-care found greater improvement for mobility activities including gait speed, longer walking distances, and a trend for improvement for self-care activities as measured by the Barthel index (Lenze et. al 2012). The mobility and self-care quality measures will standardize the collection of functional status data, which can improve communication when patients are transferred between providers. Most SNF patients receive care in an acute care hospital prior to the SNF stay, and many SNF patients receive care from another provider after the SNF stay. Use of standardized clinical data to describe a patient´s status across providers can facilitate communication across providers. In describing the importance of functional status, the National Committee on Vital and Health Statistics Subcommittee on Health (2001) noted, “Information on functional status is becoming increasing essential for fostering healthy people and a healthy population. Achieving optimal health and well-being for Americans requires an understanding across the life space of the effects of people’s health conditions on their ability to do basic activities and participate in life situations, in other words, their functional status.” This quality measure will inform SNF providers about opportunities to improve care in the area of function and strengthen incentives for quality improvement related to patient function. Jette, D. U., R. L. Warren, & C. Wirtalla. (2005). The relation between therapy intensity and outcomes of rehabilitation in skilled nursing facilities. Archives of Physical Medicine and Rehabilitation, 86 (3), 373-9. Lenze, E. J., Host, H. H., Hildebrand M. W., Morrow-Howell, N., Carpenter, B., Freedland, K. E., … Binder, E, F. (2012). Enhanced medical rehabilitation increases therapy intensity and engagement and improves functional outcomes in postacute rehabilitation of older adults: a randomized-controlled trial. Journal of the American Medical Directors Association. 13(8):708-12. National Committee on Vital and Health Statistics Subcommittee on Health. Classifying and Reporting Functional Status. 2001. Retrieved from http://www.ncvhs.hhs.gov/010617rp.pdf

Summary of NQF Endorsement Review

• Year of Most Recent Endorsement Review: 2015
  
  
• Project for Most Recent Endorsement Review: Person- and Family-Centered Care: Phase 2
  
  
• Review for Importance: The measure meets the Importance criteria (1a. Evidence, 1b. Performance Gap, 1c. High Impact) 1a. Evidence: Y-13; N-0; 1b. Performance Gap: H-3; M-8; L-2; I-0; 1c. High Priority: H-7; M-6; L-0; I-0

  Rationale: The Committee highlighted that there are talks about these quality measures becoming pay-for-performance measures; however, in IRFs there are currently requirements for pay for reporting such as a two-percent reduction in payments for failure to submit certain quality data. The Committee questioned the connection between these specific measures and pay-for-performance measures. The developer clarified that the Inpatient Rehabilitation Quality Reporting Program assigns a penalty for failure to report quality data however it is not tied to a pay-for-performance program.
  
  

• Review for Scientific Acceptability: The measure meets the Scientific Acceptability criteria (2a. Reliability - precise specifications, testing; 2b. Validity - testing, threats to validity) 2a. Reliability: H-0; M-10; L-0; I-3 2b. Validity: H-1; M-9; L-1; I-2

  Rationale:

  • The developers utilized different types of reliability including inter-rater reliability and the use of video to assess clinician assessments. Items that did not test well during the PAC demo were not included. Test-retest reliability was not performed due to the instability of the patients’ change in function.
  • The Committee expressed concerns that reliability and validity data was at the scale level and the not facility level. However, since this is an outcome measure, the Committee agreed that both reliability and validity should be considered moderate.
  • The developers confirmed that the data elements they are using in the risk adjustment model and the observed or expected calculation comes from the other assessment data and comorbidities from the claims data.
  
  
  
• Review for Feasibility: H-6; M-5; L-2; I-0 (4a. Clinical data generated during care delivery; 4b. Electronic sources; 4c.Susceptibility to inaccuracies/ unintended consequences identified 4d. Data collection strategy can be implemented)

  Rationale: The Committee questioned the length of time it takes to administer or grade the instrument. The developer noted that clinicians are assessing patients on the ability to complete the items activities listed in the measure.
  
  

• Review for Usability: H-6; M-5; L-0; I-2 (Meaningful, understandable, and useful to the intended audiences for 3a. Public Reporting/Accountability and 3b. Quality Improvement)

  Rationale: The developer noted that IRF measures are limited to Medicare only and that the Long-Term Care Hospital Quality Reporting Program was established as a Medicare program.
  
  

• Review for Related and Competing Measures: The Committee considered this measure to be related to 2321: Functional Change: Change in Mobility Score and 2634: Inpatient Rehabilitation Facility (IRF) Functional Outcome Measure: Change in Mobility Score for Medical Rehabilitation Patients. however there were no recommendations for harmonization.
  
  
• Endorsement Public Comments: March 2, 2015- March 31, 2015

  Comments received: Measures 2634 and 2636 received two similar comments. The first commenter supported the underlying concept of the measures, stating that inpatient rehabilitation facilities need to be measured on outcomes based on functional improvement. However, the commenter suggested that an alternative measure that determines how the provider improved the patient’s life (mobility) would better incentivize a change in clinical practice and associated patient-level outcomes as opposed to measure 2634 and measure 2636. Another commenter concurred with the Committee’s concern with the validity and reliability of measures developed using a cross-sectional study design from a demonstration project, which did not follow the same patients across venues of care and thus limiting applicability across sites.

  NQF response: NQF is limited to reviewing measures that are submitted for endorsement. We have added this suggestion to the measure gap list in the report. Thank you for your comment.

  Developer response: Thank you for your comment. As discussed during the measure review on January 22, 2015 and documented in the Person- and Family-Centered Care Phase 2 Draft Report on page 11, the Post-Acute Care Payment Reform demonstration was a prospective cohort study, not a cross-sectional study. In addition to collecting admission and discharge data using the CARE Tool during the post-acute care stay, inpatient claims data for acute care stays prior to and following the post-acute care stay were linked to the CARE admission and discharge data. The reliability and validity of the CARE function items were presented and discussed during the January 21-22, 2015 meeting, and several committee members referred to our analysis as very good. We have also submitted provider-level reliability data to the committee for review, as requested during the January 21-22, 2015 meeting. The Improving Medicare Post Acute Care Transformation (IMPACT) Act directs the Secretary to specify quality measures on which PAC providers are required to submit standardized patient assessment data and other necessary data specified by the Secretary with respect to five quality domains, one of which is functional status, cognitive function, and changes in function and cognitive function.
  
  

• Endorsement Committee Recommendation: Standing Committee Recommendation for Endorsement: Y-11; N-2
  
  

Measure Specifications

• NQF Number (if applicable): 2635
  
  
• Description: This quality measure estimates the percentage of Skilled Nursing Facility residents who meet or exceed an expected discharge self-care score. (The endorsed specifications of the measure are: This measure estimates the percentage of IRF patients who meet or exceed an expected discharge self-care score.)
  
  
• Numerator: The numerator is the number of residents in a Skilled Nursing Facility with a discharge self-care score that is equal to or higher than a calculated expected self-care mobility score. (The endorsed specifications of the measure are: The numerator is the number of patients in an IRF with a discharge score that is equal to or higher than the calculated expected discharge score.)
  
  
• Denominator: Skilled Nursing Facility residents included in this measure are at least 21 years of age, Medicare Fee-for-Service beneficiaries, and have complete stays.(The endorsed specifications of the measure are: Inpatient Rehabilitation Facility patients included in this measure are at least 21 years of age, Medicare beneficiaries, and are not independent on all of the self-care activities at the time of admission, and have complete stays.)
  
  
• Exclusions: This quality measure has 7 exclusion criteria: 1. Residents with incomplete stays. 2. Residents with the following medical conditions: coma; persistent vegetative state; complete tetraplegia; locked-in syndrome; or severe anoxic brain damage, cerebral edema, or compression of the brain. 3. Residents younger than age 21 4. Residents discharged to hospice. 5. Residents not covered by the Medicare Fee-for-Service program. 6. Residents in swing beds in critical access hospitals 7. Residents who do not receive rehabilitation therapy services(The endorsed specifications of the measure are: This quality measure has 5 exclusion criteria:1) Patients with incomplete stays. Rationale: It can be challenging to gather accurate discharge functional status data for patients who experience incomplete stays. Patients with incomplete stays include patients who are unexpectedly discharged to an acute care setting (Short-stay Acute Hospital, Critical Access Hospital, Inpatient Psychiatric Facility, or Long-term Care Hospital), because of a medical emergency; patients discharged to a hospice; patients discharged to another IRF; patients who die or leave an Inpatient Rehabilitation Facility (IRF) against medical advice; patients discharged directly to another IRF and patients with a length of stay less than 3 days. 2) Patients with the following medical conditions: coma; persistent vegetative state; complete tetraplegia; locked-in syndrome; or severe anoxic brain damage, cerebral edema or compression of the brain.Rationale: These patients are excluded because they may have limited or less predictable self-care improvement with the selected self-care items.3) Patients younger than age 21.Rationale: There is only limited evidence published about functional outcomes for children.4) Patients discharged to Hospice.Rationale: Patient goals may change during the IRF stay.5) Patients not covered by the Medicare program.)
  
  
• HHS NQS Priority: Patient and Family Engagement, Communication and Care Coordination
  
  
• HHS Data Source: CARE data
  
  
• Measure Type: Outcome
  
  
• Steward: Centers for Medicare & Medicaid Services
  
  
• Endorsement Status: Never Submitted
  
  
• Is the measure specified as an electronic clinical quality measure? No
  
  

Preliminary Analysis of Measure

• Preliminary analysis result: Encourage Continued Development
  
  
• Preliminary analysis summary
  • Contribution to program measure set:Research has shown differences in SNF residents’ functional outcomes by geographic region and race/ethnicity after adjusting for key patient demographic characteristics and admission clinical status, which supports the need to monitor SNF residents’ functional outcomes.
  • Impact on quality of care for patients:The mobility and self-care quality measures will standardize the collection of functional status data, which can improve communication when patients are transferred between providers. Most SNF patients receive care in an acute care hospital prior to the SNF stay, and many SNF patients receive care from another provider after the SNF stay. Use of standardized clinical data to describe a patient´s status across providers can facilitate communication across providers. In describing the importance of functional status, the National Committee on Vital and Health Statistics Subcommittee on Health (2001) noted, “Information on functional status is becoming increasing essential for fostering healthy people and a healthy population. Achieving optimal health and well-being for Americans requires an understanding across the life space of the effects of people’s health conditions on their ability to do basic activities and participate in life situations, in other words, their functional status.” This quality measure will inform SNF providers about opportunities to improve care in the area of function and strengthen incentives for quality improvement related to patient function.
  
  
  
• Does the measure address a program goal or objective? Yes. The measure promotes alignment and standardization of patient assessment across PAC settings (IMPACT Act).
  
  
• Is this a high-value measure? Yes. This MUC is an outcome measure.
  
  
• Does this measure fill a gap in the program measure set? Yes. Functional status and goal attainment.
  
  
• Measure development status: Early Development
  
  
• Is the measure fully tested for the program's setting and level of analysis? Yes. Adaptation of endorsed IRF measure; currently in field-testing for SNF population.
  
  
• Is the measure currently in use? No. The measure from which this MUC is being adapted will be implemented in October, 2016.
  
  
• Does the MUC contribute to alignment and efficient use of measurement resources (burden and cost of measurement)? Yes. This MUC supports alignment and standardization of metrics across PAC/LTC programs.
  
  
• Is the measure NQF endorsed for the program's setting and level of analysis? No. Never Submitted
  
  
• Does the measure address a high-priority quality issue in the dual eligible beneficiary population? No.
  
  

Rationale for measure provided by HHS
During a Skilled Nursing Facility (SNF) stay, the goals of treatment include fostering the patient’s ability to manage his or her daily activities so that the patient can complete self-care and mobility activities as independently as possible and if feasible, return to a safe, active and productive life in a community-based setting. Previous research has found direct relationships between increased intense therapy services and improved functional outcomes in the SNF setting. Jette et. al (2005) found that higher physical and occupational therapy intensities were associated with greater odds of improving by at least 1 stage in the mobility and activities of daily living functional independence across each condition including patients with stroke, orthopedic conditions, and cardiovascular and pulmonary conditions. Similarly, a randomized control trial, of 26 SNF patients compared higher intensity rehabilitation to the standard-of-care found greater improvement for mobility activities including gait speed, longer walking distances, and a trend for improvement for self-care activities as measured by the Barthel index (Lenze et. al 2012). The mobility and self-care quality measures will standardize the collection of functional status data, which can improve communication when patients are transferred between providers. Most SNF patients receive care in an acute care hospital prior to the SNF stay, and many SNF patients receive care from another provider after the SNF stay. Use of standardized clinical data to describe a patient´s status across providers can facilitate communication across providers. In describing the importance of functional status, the National Committee on Vital and Health Statistics Subcommittee on Health (2001) noted, “Information on functional status is becoming increasing essential for fostering healthy people and a healthy population. Achieving optimal health and well-being for Americans requires an understanding across the life space of the effects of people’s health conditions on their ability to do basic activities and participate in life situations, in other words, their functional status.” This quality measure will inform SNF providers about opportunities to improve care in the area of function and strengthen incentives for quality improvement related to patient function. Jette, D. U., R. L. Warren, & C. Wirtalla. (2005). The relation between therapy intensity and outcomes of rehabilitation in skilled nursing facilities. Archives of Physical Medicine and Rehabilitation, 86 (3), 373-9. Lenze, E. J., Host, H. H., Hildebrand M. W., Morrow-Howell, N., Carpenter, B., Freedland, K. E., … Binder, E, F. (2012). Enhanced medical rehabilitation increases therapy intensity and engagement and improves functional outcomes in postacute rehabilitation of older adults: a randomized-controlled trial. Journal of the American Medical Directors Association. 13(8):708-12. National Committee on Vital and Health Statistics Subcommittee on Health. Classifying and Reporting Functional Status. 2001. Retrieved from http://www.ncvhs.hhs.gov/010617rp.pdf

Summary of NQF Endorsement Review

• Year of Most Recent Endorsement Review: 2015
  
  
• Project for Most Recent Endorsement Review: Person- and Family-Centered Care: Phase 2
  
  
• Review for Importance: The measure meets the Importance criteria (1a. Evidence, 1b. Performance Gap, 1c. High Impact) 1a. Evidence: Y-15; N-1; 1b. Performance Gap: H-2; M-12; L-1; I-1; 1c. High Priority: H-9; M-6; L-1; I-0

  Rationale:

  • The Committee noted that the measure is proposed for use for Medicare only, and felt that this limits the use of the measure and potentially introduces duplication of efforts if using multiple tools for differing payer populations.
  • The Committee requested clarification on the intent of the measure and if it was a reflection of the care in the IRF or how the patient was prepared for integration back into the community. Specifically, they wanted to know if there is a connection between how a patient is doing at discharge and how they will do in the community. The developer indicated that information was provided in the supplemental information for the measure’s evidence. CMS further explained this is another attempt to standardize measurement and allow tracking of patients as they traverse the care continuum and between settings. The measures allow the comparison of uniform outcome measurement, whether it is self-care or mobility.
  • The Committee asked for the reasoning behind the proposal of four measures using essentially the same data. The developer explained that when testing understanding of the measures with consumers, they were led to develop both a change score concept for use by facilities, and then the percentage of patients that achieve a certain status to improve consumer understanding. They would have provided in the same measure if the NQF submissions allowed. There was a suggestion that these two pairs of measures be considered “paired” measure to promote their use together. A member from the rehabilitation community indicated he would find the information provided from both levels of measurement useful; it could be used internally for the facility for quality improvement and externally with consumers.
  • The Committee requested clarification of the 6-point measure scale. Based on input from an expert panel and comparison of current tools in use for similar purposes, the scale proposed was deemed the best fit for purpose. This became important because there is another tool in use by IRFs (the FIM) that is required for payment and uses a different scale; members indicated that facilities may find that confusing if there were different requirements for different programs. CMS indicated that a determination has not been made to convert to the function items from the CARE item Set [tool].
  
  
  
• Review for Scientific Acceptability: The measure meets the Scientific Acceptability criteria (2a. Reliability - precise specifications, testing; 2b. Validity - testing, threats to validity) 2a. Reliability: H-0; M-7; L-2; I-6 (Consensus not reached) 2b. Validity: H-1; M-7; L-1; I-6 (consensus not reached) UPDATED VOTES FOR 2a. Reliability: H-5; M-11; L-2; I-0 2b. Validity: H-3; M-14; L-0; I-1

  Rationale:

  • As raised with previous measures, the Committee indicated a strong interest in seeing scientific acceptability data at the facility level. A member noted that Crohnbach alphas provided are at the patient level. The developer indicated they could provide facility level error bars on splines for consideration.
  • The Committee asked the developer to consider if it would be more accurate to assess change in function between admission and discharge versus coming up with an expected functional level and seeing if it could be achieved. The assumption is that the comparison to an expected score would be more gameable. The developer indicated they use every bit of data they have available and the true intent of the percent of patients measure is for consumer understandability.
  • The Committee acknowledged the wealth of data provided on the reliability and validity of the CARE tool. They continued to struggle with lack of data at the facility level. The developer directed the Committee to supplemental information they provided which was submitted late and may have come in after the Committee reviewed each measure. The supplemental information included the relationship between discharge scores and discharge back to the community and between CARE scores and length of stay.
  • The Committee noted that there was some data available, specifically generalized estimation equation data, that have splines and error bars, and upon submission will be extremely helpful.
  • NQF staff clarified that this is not a unique situation and as measures become operationalized, more data becomes available and as a standing committee, that data will come back for further review. There is also the understanding that with the movement toward pay for performance, Committees want more data and NQF is trying to work those issues into the process.
  
  
  
• Review for Feasibility: H-4; M-8; L-3; I-0 (4a. Clinical data generated during care delivery; 4b. Electronic sources; 4c.Susceptibility to inaccuracies/ unintended consequences identified 4d. Data collection strategy can be implemented)

  Rationale: The Committee had no questions or concerns on the feasibility of this measure
  
  

• Review for Usability: H-3; M-7; L-3; I-2 (Meaningful, understandable, and useful to the intended audiences for 3a. Public Reporting/Accountability and 3b. Quality Improvement) Rationale: The Committee had no questions or concerns on the use and usability of this measure
  
  
• Review for Related and Competing Measures: The Committee considered this measure to be related to 2633: IRF Functional Outcome Measure: Discharge Self-Care Score for Medical Rehabilitation Patients (CMS) and 2286: Functional Change: Change in Self-Care Score (USDMR), however there were no recommendations for harmonization.
  
  
• Endorsement Public Comments: March 2, 2015- March 31, 2015

  Comments received: One commenter noted that these are important measures but they need to be analyzed and improved as additional data is collected. Another commenter concurred with the Committee’s concern with the validity and reliability of measures developed using a cross-sectional study design from a demonstration project, which did not follow the same patients across venues of care and thus limiting applicability across sites.

  Committee response: The Committee requested additional information to allow for more comprehensive evaluation of the consensus not reached and not recommended measures. This additional information was discussed on the post-comment committee call and the Committee had an opportunity to re-vote on the applicable measures. This measure was recommended by the Committee after reviewing the additional information and the comments.

  Developer response:

  • Thank you for your comment. As discussed during the measure review on January 22, 2015 and documented in the Person- and Family-Centered Care Phase 2 Draft Report on page 11, the Post-Acute Care Payment Reform demonstration was a prospective cohort study, not a cross-sectional study. In addition to collecting admission and discharge data using the CARE Tool during the post-acute care stay, inpatient claims data for acute care stays prior to and following the post-acute care stay were linked to the CARE admission and discharge data. The reliability and validity of the CARE function items were presented and discussed during the January 21-22, 2015 meeting, and several committee members referred to our analysis as very good. We have also submitted provider-level reliability data to the committee for review, as requested during the January 21-22, 2015 meeting. The Improving Medicare Post Acute Care Transformation (IMPACT) Act directs the Secretary to specify quality measures on which PAC providers are required to submit standardized patient assessment data and other necessary data specified by the Secretary with respect to five quality domains, one of which is functional status, cognitive function, and changes in function and cognitive function.
  • The Post-Acute Care Payment Reform Demonstration was a prospective cohort study. It was not a cross-sectional study. For the study, data were collected at admission and discharge for each patient in the study. In addition, we collected interim assessment data for patients in the cost-resource utilization segment of the study. As part of the study, we also linked the CARE admission and discharge data with acute care and post-acute care claims data in order to examine episodes of care and post discharge readmissions. (B). The items and the summed self-care and mobility scores are statistically significantly associated with several outcomes, including length of stay and discharge destination. The admission IRF self-care and IRF mobility scores were moderately correlated with length of stay with coefficients of -0.463 (p < .0001) for self-care and -0.474 (p < .001) for mobility. As expected, the summed self-care and mobility discharge scores for patients who were discharged to home were significantly different than the scores of patients discharged to a long-term care/nursing home setting. The mean (standard deviation) discharge self-care score for patients going home and to long-term care/nursing home were 34.29 (7.04) and 24.57 (9.39), respectively. For mobility, the mean (standard deviation) scores were 57.35 (15.68) and 36.57 (15.07), respectively. The patients going home had higher scores, indicating more function, as we expected. (C). The CARE function items included in the 4 IRF quality measures and 2 LTCH quality measures have undergone validity testing. In addition to the results we present in our testing documentation, the data presented above (in 3b), we examined the relationship between the current functional assessment items and the CARE items for each PAC setting. The reports describing the testing are available at: http://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Post-Acute-Care-Quality-Initiatives/CARE-Item-Set-and-B-CARE.html.
  
  
  
• Endorsement Committee Recommendation: Standing Committee Recommendation for Endorsement: Y-10; N-5; Y-17; N-1
  
  

Measure Specifications

• NQF Number (if applicable):
  
  
• Description: This measure describes the risk-standardized rate of Medicare fee-for-service (FFS) patients/residents/persons who are discharged to the community following a post-acute stay/episode, and do not have an unplanned (re)admission to an acute care hospital or LTCH in the 31 days following discharge to community, and remain alive during the 31 days following discharge to community.
  
  
• Numerator: This measure does not have a simple form for the numerator and denominator. The numerator is defined as the risk-adjusted estimate of the number of patients/residents/persons included in the measure who are discharged to the community, and do not have an unplanned (re)admission to an acute care hospital or LTCH on the day of discharge or in the 31 days following discharge to community, and remain alive during the 31 days following discharge to community. The numerator estimate includes risk adjustment for patient/resident/person characteristics, and a statistical estimate of the facility/agency effect beyond case mix.
  
  
• Denominator: The denominator is computed with the same model used for the numerator. It is the model developed using all non-excluded facility/agency stays/episodes in the national data. The measure includes all facility/agency stays/episodes in the measurement period that are observed in national Medicare FFS data and do not fall into an excluded category. For a particular facility/agency, the model is applied to the patient/resident/person population, but the facility/agency effect term is 0. In essence, it is the number of discharges to community that would be expected for that patient/resident/person population at the average facility/agency.
  
  
• Exclusions: i) Age under 18 years; (ii) No short-term acute care stay within 30 days prior to SNF admission; (iii) Discharges to psychiatric hospital; (iv) Discharges against medical advice; (v) Discharges to federal hospitals or disaster alternative care sites;; (vi) Patients not continuously enrolled in Part A FFS Medicare for the 12 months prior to the SNF admission date, and at least 31 days after SNF discharge date; (vii) Patients whose prior short-term acute-care stay was for non-surgical treatment of cancer; (viii) Discharges to hospice; (ix) SNF stays that end in transfer to another SNF; (x) SNF stays with claims data that are problematic (e.g., anomalous records for stays that overlap wholly or in part, or are otherwise erroneous or contradictory); (xi) Patients who received care from a provider located outside of the US, Puerto Rico, or a US territory..
  
  
• HHS NQS Priority: Making Care Safer, Patient and Family Engagement, Making Care Affordable
  
  
• HHS Data Source: Claims
  
  
• Measure Type: Outcome
  
  
• Steward: Centers for Medicare & Medicaid Services
  
  
• Endorsement Status: Never Submitted
  
  
• Is the measure specified as an electronic clinical quality measure? No
  
  

Preliminary Analysis of Measure

• Preliminary analysis result: Encourage continued development
  
  
• Preliminary analysis summary
  • Contribution to program measure set:This measure meets the priorities, needs and requirements of the SNF QRP, addresses a high-impact area of function and resource use, and meets the requirements of the IMPACT Act. This measure would be unique to the SNF QRP in terms of addressing this quality issue (discharge to community); the program does not currently include any resource use measures in this area.
  • Impact on quality of care for patients:The ultimate goals of post-acute care are avoiding institutionalization and returning patients to their previous level of independence and functioning, with discharge to community being the primary goal for the majority of post-acute patients. For many, home is a symbol of independence, privacy, and competence. Discharge to community is considered a valuable outcome to measure because it is a multifaceted measure that captures the patient’s functional status, cognitive capacity, physical ability, and availability of social support at home.
  
  
  
• Does the measure address a program goal or objective? Yes. This measure is related to patient and family engagement and communication and care coordination, two of the primary domains for measures in the program.
  
  
• Is this a high-value measure? Yes. This is a resource use measure that assesses discharge to community and is being proposed to meet the requiremetns of the IMPACT Act.
  
  
• Does this measure fill a gap in the program measure set? Yes. Currently, there are no other measures that address discharge to community in the SNF QRP. This measure also addresses care coordination and transition planning concepts.
  
  
• Measure development status: Early Development
  
  
• Is the measure fully tested for the program's setting and level of analysis? No. This measure has not yet been tested.
  
  
• Since no, could the measure be revised to use in the setting or at level of analysis under consideration? N/A. N/A
  
  
• Is the measure currently in use? No. This measure is not currently being used in a program.
  
  
• Does the MUC contribute to alignment and efficient use of measurement resources (burden and cost of measurement)? Yes. This measure would be unique to the SNF QRP in terms of measuring discharge to community; the program does not currently include any resource use measures in this area. The measure would also promote alignment of similar measure concepts across programs.
  
  
• Is the measure NQF endorsed for the program's setting and level of analysis? No. Never Submitted
  
  
• Does the measure address a high-priority quality issue in the dual eligible beneficiary population? Yes.
  
  

Rationale for measure provided by HHS
The ultimate goals of post-acute care are avoiding institutionalization and returning patients to their previous level of independence and functioning, with discharge to community being the primary goal for the majority of post-acute patients. For many, home is a symbol of independence, privacy, and competence. Discharge to community is considered a valuable outcome to measure because it is a multifaceted measure that captures the patient’s functional status, cognitive capacity, physical ability, and availability of social support at home. There is considerable variation in discharge to community rates within and across post-acute settings. Studies show geographic variation, variation across patient socioeconomic characteristics (for example, race and ethnicity), and variation by facility characteristics (for profit vs. nonprofit, freestanding vs. hospital-based, urban vs. rural). In the IRF setting, discharge to community rates vary across providers, ranging from about 60% to 75%. The 2015 MedPAC report shows that, in FY 2013, the facility-level, mean risk-adjusted discharge to community rate for IRFs within 100 days of admission was 75.8%, and the mean observed rate was 74.7%. Discharge to community rates also vary widely in the SNF setting, ranging from as low as 31% to as high as 65%. The 2015 MedPAC report shows a mean risk-adjusted discharge to community rate of 37.5% for SNFs within 100 days of admission, and mean observed rate of 40.1%. A multicenter study of 23 LTCHs reported that only 28.8% of 1,061 patients who were ventilator-dependent on admission were discharged to home or assisted living facility. A study of 66,510 Medicare beneficiaries during pre- and post-HH episodes, revealed that 64 percent of beneficiaries discharged from HH did not use any other Medicare-reimbursed acute or post-acute services in the 30 days following HH discharge. Significant numbers of patients were admitted to inpatient facilities (29 percent) and lesser numbers to skilled nursing facilities (7.6 percent), inpatient rehabilitation (1.5 percent) and home health (7.2 percent) or hospice (3.3 percent) within 30 days of HH discharge (Wolff et al., 2008).

Measure Specifications

• NQF Number (if applicable):
  
  
• Description: Percentage of stays Inpatient Rehabilitation Facility (IRF), Long Term Care Facility (LTCH), and Skilled Nursing Facility (SNF) or care episodes Home Health (HH) in which a drug regimen review was conducted at the Admission (IRF, LTCH or SNF)/ Start of Care (SOC)/ Resumption of Care (ROC) (HH) and timely follow-up with a physician occurred each time potential clinically significant medication issues were identified throughout the stay (IRF, LTCH, or SNF) or care episode (HH).
  
  
• Numerator: Number of stays or care episodes where the medical record contains documentation of a drug regimen review conducted at admission or start-of-care or resumption-of-care with all potential clinically significant medication issues identified during the course of care and followed-up with a physician or physician designee.
  
  
• Denominator: Care episodes or stays ending during the reporting period. Assessment timing is as follows: Beginning of care episode or stay: • HH – SOC or ROC • SNF – Admission • IRF - Admission • LTCH – Admission End of care episode or stay: • HH – Transfer, Discharge, or Death at Home • SNF – Discharge, or expired • IRF – Discharge, or expired • LTCH – Discharge, or expired
  
  
• Exclusions: Denominator Exclusion: NONE Numerator Exclusion: NONE
  
  
• HHS NQS Priority: Making Care Safer, Communication and Care Coordination
  
  
• HHS Data Source: Patient Assessment Data - the Minimum Data Set (MDS)
  
  
• Measure Type: Process
  
  
• Steward: Centers for Medicare & Medicaid Services
  
  
• Endorsement Status: Never Submitted
  
  
• Is the measure specified as an electronic clinical quality measure? No
  
  

Preliminary Analysis of Measure

• Preliminary analysis result: Encourage Continued Development
  
  
• Preliminary analysis summary
  • Contribution to program measure set:This measure of medication reconciliation addresses an important aspect of care for patients as they transition from one setting to another. During transitions of care, there can be many changes to the drug regimen. Identifying medication issues and resolving these with a physician or physician-designee are important steps to prevent issues such as; adverse drug reaction, ineffective drug therapy, drug interactions, duplicate therapy, dosage errors etc., any of which could lead to hospitalization or re-hospitalization. Medication reconciliation is currently one of the Joint Commission National Patient Safety Goals –which is to maintain and communicate accurate patient medication information.
  • Impact on quality of care for patients:The improvement in the patient’s medical condition, decrease in medication errors or events as well as a decrease in use of the Emergency Department (ED) and hospitalization are all outcomes that would improve the quality of care for patients.
  
  
  
• Does the measure address a program goal or objective? Yes. Addresses a high priority domain (communication/care coordination), supports alignment across programs under the IMPACT Act, and addresses program gap in measurement.
  
  
• Is this a high-value measure? Yes. This is a high value process measure with strong links to both hospitalization and re-hospitalization measures which focus on outcomes. The measure is not duplicative and captures a broad population in the SNF setting.
  
  
• Does this measure fill a gap in the program measure set? Yes. This measures identifies a high priority of care in the comunication and care coordination domain, which includes the concepts of transitions, re-hospitalization and medication reconcilitation. Currently there are no medciation management measures in the SNF QRP.
  
  
• Measure development status: Early Development
  
  
• Is the measure fully tested for the program's setting and level of analysis? No. This measure has not yet been tested. Testing is to be conducted for Post-Acute Care settings in 2015.
  
  
• Since no, could the measure be revised to use in the setting or at level of analysis under consideration? N/A. It is anticipated that this measure will be fully tested for each setting of care for which it is proposed.
  
  
• Is the measure currently in use? No. This measure has not currently being used in a program.
  
  
• Does the MUC contribute to alignment and efficient use of measurement resources (burden and cost of measurement)? Yes. This measure is proposed to meet the requirements of the IMPACT Act and is in development to ensure alignment acorss PAC settings of care, thus should promote efficient use of resources.
  
  
• Is the measure NQF endorsed for the program's setting and level of analysis? No. Never Submitted
  
  
• Does the measure address a high-priority quality issue in the dual eligible beneficiary population? No.
  
  

Rationale for measure provided by HHS
Medication review in post-acute care is generally considered to include medication reconciliation for all medications and medication review for what poses as potential clinically significant medication issues for the patient/resident. As a process measure, medication reconciliation and medication review for potential clinically significant medication issues are expected to reduce re-hospitalizations, reduce adverse events related to medications and improve health outcomes.

Measure Specifications

• NQF Number (if applicable):
  
  
• Description: The MSPB-PAC Measure for SNFs evaluates providers’ efficiency relative to the efficiency of the national median SNF provider. Specifically, the MSPB-PAC Measure assesses the cost to Medicare for services during an episode of care, which consists of a treatment period and an associated services period. The episode is triggered by an admission to a SNF stay. The treatment period begins at the trigger and ends at discharge. The associated services period begins at the trigger and ends 30 days after the end of the treatment period (i.e., discharge). These periods constitute the episode window during which beneficiaries’ Medicare services are counted toward the episode. The MSPB-PAC episode includes all services during the episode window that are attributable to the SNF provider and those rendered by other providers, except those services during the associated services period that are clinically unrelated to SNF responsibilities (e.g., planned care and routine screening).
  
  
• Numerator: The numerator is the attributed provider’s average MSPB-PAC Amount. The MSPB-PAC Amount for each SNF provider depends on two factors: i) the average of the ratio of standardized episode spending level and expected episode spending for each SNF provider; and ii) the average standardized episode spending across all SNF providers. To calculate the MSPB-PAC Amount for each SNF, one finds the average of the ratio of the standardized episode spending over the expected episode spending, and then multiplies this quantity by the average episode spending level across all SNFs.
  
  
• Denominator: The denominator for a SNF’s MSPB-PAC Measure is the weighted median MSPB-PAC Amount across all episodes for SNFs nationally.
  
  
• Exclusions: The measure excludes the following episodes: • Any episode that is triggered by an SNF stay that happens outside the 50 states or DC. • Any episode that is triggered by an SNF stay for which we see Part C crossover claims. • Any episode for which standard allowed amount of the SNF stay could not be calculated or is equal to 0. • Any episode in which a beneficiary is not enrolled in Medicare Fee-for-Service for the entirety of the lookback period plus the episode window or is enrolled in Part C for any part of the lookback plus episode window. • Any episode in which a beneficiary has a primary payer other than Medicare for any part of the lookback plus episode window. • Any episode for which the lookback period extends beyond our observation period.
  
  
• HHS NQS Priority: Making Care Affordable
  
  
• HHS Data Source: Claims
  
  
• Measure Type: Cost/Resource Use
  
  
• Steward: Centers for Medicare & Medicaid Services
  
  
• Endorsement Status: Never Submitted
  
  
• Is the measure specified as an electronic clinical quality measure? No
  
  

Preliminary Analysis of Measure

• Preliminary analysis result: Encourage for continued development
  
  
• Preliminary analysis summary
  • Contribution to program measure set:The potential value of this measure is to assess the level of health care costs in the MSPB-PAC population and increase alignment across high priority gap areas. This measure has the potential to identify the determining factors of high cost of health care and can lead to new efficiency measures that can result in reduced costs.
  • Impact on quality of care for patients:The potential impact is that patients may receive additional support, education and services to enable them to transition to the community and stay in the community without hospitalizations, ED use as well as admissions to SNFs and home health care.
  
  
  
• Does the measure address a program goal or objective? Yes. This measure meets the making care more affordable (efficiency-based measures) program goal.?
  
  
• Is this a high-value measure? Yes. This is an outcome and utilization measure (cost/resource use) that focuses on the efficient use of health care resources.
  
  
• Does this measure fill a gap in the program measure set? Yes. The measure could fit into more than one of the high leverage areas identified for performance measurement including cost/access and making care affordable.There are no other efficienty based measures currently in the SNF QRP.
  
  
• Measure development status: Early Development
  
  
• Is the measure fully tested for the program's setting and level of analysis? Testing is in progress. The information received indicates testing is occuring across PAC settings, which should include SNFs.
  
  
• Is the measure currently in use? No. This measure is not currently in use.
  
  
• Does the MUC contribute to alignment and efficient use of measurement resources (burden and cost of measurement)? Yes. This measure would contribute to alignment as it will be used across all of PAC settings except Hospice; in addition, since it will be calculated from claims data it should not pose a burden to any of the PAC settings.
  
  
• Is the measure NQF endorsed for the program's setting and level of analysis? No. Never Submitted
  
  

Rationale for measure provided by HHS
Medicare payments to PAC have grown at a consistently higher rate than other major Medicare sectors. Between 2001 and 2013, Medicare PAC spending grew at an annual rate of 6.1 percent and doubled to $59.4 billion.

Measure Specifications

• NQF Number (if applicable):
  
  
• Description: This measure reports the percentage of skilled nursing facility residents who are receiving an antipsychotic medication during a quarter but who were not receiving an antipsychotic medication at admission.
  
  
• Numerator: Skilled nursing facility residents who are receiving an antipsychotic medication during a quarter but who were not receiving an antipsychotic medication on their first assessment after admission.
  
  
• Denominator: Skilled nursing facility residents, except for those who meet the exclusion criteria.
  
  
• Exclusions: Residents are excluded from the denominator if they are diagnosed with any of the following conditions: Schizophrenia, Tourette’s Syndrome and Huntington’s Disease.
  
  
• HHS NQS Priority: Making Care Safer, Effective Prevention and Treatment
  
  
• HHS Data Source: Nursing Home Minimum Data Set (MDS) Version 3.0
  
  
• Measure Type: Process
  
  
• Steward: Centers for Medicare & Medicaid Services
  
  
• Endorsement Status: Never Submitted
  
  
• Is the measure specified as an electronic clinical quality measure? No
  
  

Preliminary Analysis of Measure

• Preliminary analysis result: Encourage Continued Development
  
  
• Preliminary analysis summary
  • Contribution to program measure set:Currently, there are no medication management measures in the SNF QRP. This measure fills a significant gap area within a vulnerable population and is expected to reduce hospitalizations, adverse events related to medications, and improve health outcomes.
  • Impact on quality of care for patients:The potential impact of this measure is to promote medication management within the older adult population taking antipsychotic medications. This measure is intended to improve health outcomes and quality of care for patients in the SNF setting.This measure is similar to a currently endorsed NQF measure: Antipsychotic Use in Persons with Dementia (#2111),a health plan Part D measure and consideration may be given to alignment/harmonization since these measures have congruent rationale.
  
  
  
• Does the measure address a program goal or objective? Yes. Addresses a high priority domain (Communication/Care Coordination), supports alignment across programs under the IMPACT Act, and addresses a program gap in measurement.
  
  
• Is this a high-value measure? Yes. This is a high value process measure that is not duplicative and captures a broad population in the SNF setting.
  
  
• Does this measure fill a gap in the program measure set? Yes. Currently there are no medication management measures in the SNF QRP.
  
  
• Measure development status: Field Testing
  
  
• Is the measure fully tested for the program's setting and level of analysis? Yes. The level of analysis is at the facility level within the nursing home setting.
  
  
• Is the measure currently in use? Yes. This measure is currently in use in SNF QRP (2012-Present), but is undergoing substantial specification changes. The focus of the measure is changing from prevalent to incident populations.
  
  
• Does a review of its performance history raise any concerns? . The measure target population includes all skilled nursing facility residents. The developers noted the high possibility of a performance gap due to increased use of antipsychotic medications in the older adult population and variation in facility level administration of medication and prescribing rates. This could lead to unintended consequences such as avoiding admission of patients with diagnoses that may put them at risk for needing antipsychotic medication.
  
  
• Does the MUC contribute to alignment and efficient use of measurement resources (burden and cost of measurement)? Yes. This measure contributes to the alignment across all care settings under the IMPACT Act, addresses two NQS priorities of making care safer by reducing harm caused in the delivery of care and promoting effective communication and coordination of care, starting with cardiovascular disease. This measure also contributes to the efficient use of measurement resource.
  
  
• Is the measure NQF endorsed for the program's setting and level of analysis? No. Never Submitted
  
  
• Does the measure address a high-priority quality issue in the dual eligible beneficiary population? No.
  
  

Rationale for measure provided by HHS
Antipsychotic medications can be potentially dangerous for the elderly, especially for those who do not have the clinical indication. Of particular concern is the off-label use of these drugs for elders with dementia or dementia-related psychoses or agitation (Jeste et al., 2008). In April 2005, the FDA issued a black box warning against prescribing atypical antipsychotic medications for elderly with dementia (Rosack, 2005). The evidence on which the black box warning was based came from a meta-analysis of data from 17 randomized trials with a total of 5,106 patients which identified an “approximately 1.6- to 1.7-fold increase in mortality in the combined studies” (Rosack, 2005). In June 2008, the FDA extended the warning to all categories of antipsychotic drugs, conventional as well as atypical (Rosack, 2005). In this warning, the FDA advised health care professionals, "Antipsychotics are not indicated for the treatment of dementia-related psychosis." Besides elevated mortality risk, clinical trials of atypical antipsychotic medications also show elevated risk for serious adverse events including falls, somnolence and abnormal gait (Rosack, 2005; FDA, 2008; Ballard & Margallo-Lana, 2004; Martin et al., 2003; Neil, Curran, and Wattis, 2003; Doody et al., 2001; Jackson-Siegal, 2004). Additionally, there is evidence of increased risk for cerebrovascular adverse events associated with certain atypical antipsychotic medications (e.g., risperidone, olanzapine, and aripiprazole) (Jeste et al., 2008). While the black box warnings applied to all antipsychotic medications, a recent study identified some differences in mortality risk by medication and dose among a large population based cohort of dually-eligible nursing home residents prescribed antipsychotic medications (Huybrechts et al., 2012). In addition to being a threat to patient safety, antipsychotic medications are also expensive to consumers and Medicare. Atypical antipsychotic drugs cost more than $13 billion in 2007, "nearly 5 percent of all U.S. drug expenditures" (Alexander et al., 2011). They are also responsible for a significant portion of expenditures for Medicare Part D (Doody et al., 2001). Furthermore, the OIG report found that 51% of Medicare atypical antipsychotic drug claims for elderly nursing home residents were erroneous, amounting to $116 million. (OIG, 2011). Use of this measure should prompt nursing facilities to re-examine their prescribing patterns which may result in practice consistent with clinical recommendations and guidelines. Reference: Alexander, G., Gallagher, S., Mascola, A., et al.: Increasing off-label use of antipsychotic medications in the United States, 1995-2008. Pharmacoepidemiol Drug Saf 20(2):177-184, 2011. Ballard, C.G., and Margallo-Lana, M.L.: The relationship between antipsychotic treatment and quality of life for patients with dementia living in residential and nursing home care facilities. J Clin Psychiatry 65 Suppl 11:23-28, 2004. Doody, R.S., Stevens, J.C., Beck, C., et al.: Practice parameter: management of dementia (an evidence-based review). Report of the Quality Standards Subcommittee of the American Academy of Neurology. Neurology 56(9):1154-1166, 2001. FDA: Information for Healthcare Professionals: Conventional Antipsychotics. FDA Alert (June 16, 2008). http://www.fda.gov/drugs/drugsafety/postmarketdrugsafetyinformationforpatientsandproviders/ucm124830.htm Huybrechts, K.F., Gerhard, T., Crystal, S., et al.: Differential risk of death in older residents in nursing homes prescribed specific antipsychotic drugs: population based cohort study. BMJ 344:e977, 2012. Jackson-Siegal, J.M., Schneider, L.S., Baskys, A., et al.: Recognizing and responding to atypical antipsychotic side effects. J Am Med Dir Assoc 5(4 Suppl):H7-10, 2004. Jeste, D.V., Blazer, D., Casey, D., et al.: ACNP White Paper: update on use of antipsychotic drugs in elderly persons with dementia. Neuropsychopharmacology 33(5):957-970, 2008. Martin, H., Slyk, M.P., Deymann, S., et al.: Safety profile assessment of risperidone and olanzapine in long-term care patients with dementia. J Am Med Dir Assoc 4(4):183-188, 2003. Neil, W., Curran, S., and Wattis, J.: Antipsychotic prescribing in older people. Age Ageing 32(5):475-483, 2003. Office of Inspector General (OIG): Medicare Atypical Antipsychotic Drug Claims For Elderly Nursing Home Residents, 2011. Rosack, J.: FDA orders new warning on atypical antipsychotics. Psychiatr News 40(9):1-50, 2005."

Measure Specifications

• NQF Number (if applicable): 676
  
  
• Description: This measure reports the percentage of skilled nursing facility residents who have reported daily pain with at least one episode of moderate to severe pain, or severe or horrible pain of any frequency in the 5 days prior to the assessment. (The endorsed specifications of the measure are: This measure reports the percentage of short-stay residents, of all ages, in a nursing facility, who have reported almost constant or frequent pain, and at least one episode of moderate to severe pain, or any severe or horrible pain, in the 5 days prior to the target assessment. This measure is based on data from the Minimum Data Set (MDS 3.0) OBRA, PPS, and/or discharge assessments. Short-stay residents are identified as residents who have had 100 or fewer days of nursing facility care. )A separate measure (NQF#0677, Percent of Residents Who Self-Report Moderate to Severe Pain (Long-Stay)) is to be used for residents who had at least 100 days of nursing facility care.
  
  
• Numerator: The numerator is the number of skilled nursing facility residents who are able to self-report with a selected Minimum Data Set (MDS) assessment during the selected quarter and who report almost constant or frequent pain AND at least one episode of moderate to severe pain in the 5 days prior to the assessment OR who report very severe/horrible pain of any frequency in the 5 days prior to the assessment. (The endorsed specifications of the measure are: The numerator is the number of short-stay residents who are able to self-report with a selected Minimum Data Set (MDS) assessment (OBRA admission, quarterly, annual or significant change/correction assessments; PPS 5-,14-, 30-, 60-, 90-day, or readmission/return assessments; or discharge assessment with or without return anticipated) during the selected quarter and who report almost constant or frequent pain AND at least one episode of moderate to severe pain in the 5 days prior to the assessment OR who report very severe/horrible pain of any frequency in the 5 days prior to the assessment.)
  
  
• Denominator: Skilled nursing facility residents, except those who meet the exclusion criteria. (The endorsed specifications of the measure are: The denominator is the total of all short-stay residents in the nursing facility who have an OBRA, PPS or discharge MDS assessment during the selected 6 month period and who do not meet the exclusion criteria.)
  
  
• Exclusions: A resident is excluded if they did not meet the pain symptom conditions for the numerator AND any of the following conditions are true: 1) The pain assessment interview was not completed (J0200 = 0, -, ^) OR 2) The pain presence item was not completed (J0300 = 09, ^) OR 3) For residents with pain or hurting at any time in the last 5 days (J0300 = 1), any of the following are true: 3.1) The pain frequency item was not completed (J0400 = [9, -, ^]); 3.2) Neither of the pain intensity items were completed (J0600A = [99, -, ^] and J0600B = [99, -, ^]); 3.3) The numeric pain intensity item indicates no pain (J06000A = [00]). (The endorsed specifications of the measure are: A resident is excluded from the denominator if they did not meet the pain symptom conditions for the numerator AND any of the following conditions are true: 1) the resident cannot self-report; 2) there are missing data in the responses to the relevant questions in the MDS assessment; OR 3) the assessment indicates that the resident had pain or hurting at any time in the last 5 days (J0300 = 1), but the numeric pain intensity item indicates no pain (J0600A = 00).Nursing facilities with fewer than 20 residents in the sample are excluded from public reporting because of small sample size.)
  
  
• HHS NQS Priority: Patient and Family Engagement, Effective Prevention and Treatment
  
  
• HHS Data Source: Nursing Home Minimum Data Set (MDS) Version 3.0
  
  
• Measure Type: Outcome
  
  
• Steward: Centers for Medicare & Medicaid Services
  
  
• Endorsement Status: Endorsed; MUC same as endorsed version:No (target population differs)
  
  
• Changes to Endorsed Measure Specifications?: The MUC list indicates the measure has not been modified from its endorsed version.
  
  
• Is the measure specified as an electronic clinical quality measure? No
  
  

Preliminary Analysis of Measure

• Preliminary analysis result: Encourage Continued Development
  
  
• Preliminary analysis summary
  • Contribution to program measure set:Pain assessment has been identified as a key and systematic component of ensuring patient engagement and goal attainment for post-acute settings; this measure will promote a standardized and continuous assessment of patient perceptions of pain and impact on quality of life.
  • Impact on quality of care for patients:This measure will promote patient engagement and specifically ensuring care/treatment is delivered to address patient preferences and goal attainment.
  
  
  
• Does the measure address a program goal or objective? Yes. This outcome measure meets the overarching PAC and SNF QRP program goals of improving outcomes and improving patient engagement.
  
  
• Is this a high-value measure? Yes. This is an outcome measure
  
  
• Does this measure fill a gap in the program measure set? Yes. There are no measures in the SNF QRP; this measure addresses an identified priority areas of improving outcomes and increasing patient engagement.
  
  
• Measure development status: Field Testing
  
  
• Is the measure fully tested for the program's setting and level of analysis? Yes. Measure is in the field-testing process and is being tested at the SNF level.
  
  
• Is the measure currently in use? No. While similar measures are in use if alternate programs, this MUC is not in use in the SNF QRP, and has a different target population as compared to the other programs.
  
  
• Does the MUC contribute to alignment and efficient use of measurement resources (burden and cost of measurement)? Yes. Although a different target population (denominator) this measure is being constructed as a modification to measures currently being collected through the MDS 3.0.
  
  
• Is the measure NQF endorsed for the program's setting and level of analysis? No. Endorsed; MUC same as endorsed version:No (target population differs)
  
  
• Does the measure address a high-priority quality issue in the dual eligible beneficiary population? No.
  
  

Rationale for measure provided by HHS
The opportunity for improving unrelieved pain in nursing home residents continues to be demonstrated by reports of less-than-optimal pain management, considerable variation in pain management, and data from interventions aimed at improving pain management in nursing homes. In 2011, a report from the Institute of Medicine stated, “evidence indicates that nursing homes undertreat pain, especially in cognitively impaired and minority residents” (Institute of Medicine, 2011). Recent reports indicate that pain management in nursing home can be improved by improving pain assessment, including use of structured assessment tools. Investigations of pain management strategies have increasingly broadened to include comprehensive approaches that are evidence based, multidisciplinary, and use behavioral approaches to educate and train staff (Cervo, et al., 2012; Savvas et al., 2014). Comprehensive interventions attempt to improve both pain assessment and pain treatment by adopting pain-assessment tools and pain-management clinical guidelines. Pain management may also be improved by nonpharmacological approaches to pain management, such as cognitive behavioral therapy, mindfulness meditation, relaxation techniques, assistive devices, physical activity and exercise, and complementary therapies. (Abdulla et al., 2013). References: 1. Abdulla, A., Adams, N., Bone, M., Elliott, A. M., Gaffin, J., Jones, D., et al. (2013). Guidance on the management of pain in older people. Age and Ageing, 42 Suppl 1, i1-57. 2. Cervo, F. A., Bruckenthal, P., Fields, S., Bright-Long, L. E., Chen, J. J., Zhang, G., et al. (2012). The role of the CNA Pain Assessment Tool (CPAT) in the pain management of nursing home residents with dementia. Geriatric Nursing (New York, NY), 33(6), 430-438. 3. Institute of Medicine. (2011). Relieving Pain in America: A Blueprint for Transforming Prevention, Care, Education, and Research. Washington DC: National Academics Press. 4. Savvas, S. M., Toye, C. M., Beattie, E. R., & Gibson, S. J. (2014). An evidence-based program to improve analgesic practice and pain outcomes in residential aged care facilities. Journal of the American Geriatrics Society, 62(8), 1583-1589.

Summary of NQF Endorsement Review

• Year of Most Recent Endorsement Review: 2011
  
  
• Project for Most Recent Endorsement Review: National Voluntary Consensus Standards for Nursing Homes
  
  
• Review for Importance: See CDP Recommendation]
  
  
• Review for Scientific Acceptability: See CDP Recommendation]
  
  
• Review for Feasibility: See CDP Recommendation]
  
  
• Review for Usability: See CDP Recommendation]
  
  
• Review for Related and Competing Measures: See CDP Recommendation]
  
  
• Endorsement Public Comments: Multiple commenters were concerned about nursing home residents who are unable to self-report pain. In response, the developer proposed changing the title of the measure to include "who self-report," and the Committee agreed that this adequately addressed theissue.
  
  
• Endorsement Committee Recommendation: For both NQF 0676 and 0677): Research indicates that at least 40 percent to 85 percent of nursing facility residents have persistent pain. The percentage may be evenhigher; research suggests pain often is not fully documented. 63,64,65,66,67,68,69These two measures were discussed concurrently with measure 009, and many of the same issues apply. The Committee voted to recommend the measure for time-limited endorsement with the following conditions: evaluate the patient’s cognitive status whenreporting pain;further examine missing data to ensure there is not an underreporting of pain in order to improve the facility’s rating;address concerns regarding the frequency of pain, e.g., does decreased frequency, but increased intensity, equal effective care;address concerns around unintendedconsequences that may occur when interpreting the measure results during reliability testing. The range of 0 to 10 is not linear and therefore does not account for potential changes in pain score. It was noted that some residents would rather have pain atlevel 4 or 5 than take opiates; andaccount for patient preference in pain management.

  During follow-up, the developer explained it plans to examine the results of this measure compared to those produced by independentmeasures solely focused on cognitive status (i.e., Brief Interview of Mental Status [BIMS] or resident ability to complete the MDS self-report pain assessment) during testing.

  The Committee voted to recommend these measures for time-limited endorsement.

  The measures meet the National Priority of Care Coordination.
  
  

Measure Specifications

• NQF Number (if applicable): 680
  
  
• Description: The measure reports the percentage of skilled nursing facility residents who are assessed and appropriately given the seasonal influenza vaccine. (The endorsed specifications of the measure are: The measure reports the percentage of residents or patients who are assessed and appropriately given the seasonal influenza vaccine.)This measure includes residents or patients 180 days of age or older on target date of assessment in the denominator. The measure is based on data from the Minimum Data Set (MDS) 3.0 assessments of nursing home residents, Inpatient Rehabilitation Facility Patient Assessment Instrument (IRF-PAI) Version 1.2 assessments for Inpatient Rehabilitation Facility (IRF) patients, and the Long-Term Care Hospital (LTCH) Continuity Assessment Record & Evaluation (CARE) Data Set Version 2.01 assessments of LTCH patients.Data are collected in each of these three settings using standardized items across the three assessment instruments. For the nursing homes/skilled nursing facilities, the measure is limited to short-stay residents, identified as residents who have had 100 or fewer days of nursing facility care. For the LTCHs, this measure will include all patients, irrespective of a patient’s length of stay. For IRFs, this measure will include all Medicare Part A and Part C patients, irrespective of a patient’s length of stay. This measure mirrors the NQF standard specifications that were developed to achieve a uniform approach to data collection across healthcare settings and populations by addressing who is included in and excluded from the target denominator population, who is included in and excluded from the numerator population, time window for measurement and time window for vaccinations. National Quality Forum. (2008, December).
  
  
• Numerator: The numerator is the number of skilled nursing facility residents, during the numerator time window, who meet any of the following criteria: (1) received the seasonal influenza vaccine during the most recently-completed influenza vaccination season; (2) were offered but declined the seasonal influenza vaccine; or (3) were ineligible due to contraindication(s). The numerator time window coincides with the most recently-completed seasonal influenza vaccination season (which begins on October 1, or when the vaccine first becomes available, and ends on March 31 of the following year). (The endorsed specifications of the measure are: The numerator is the number of residents or patients in the denominator sample who, during the numerator time window, meet any of the following criteria: (1) those who received the seasonal influenza vaccine during the most recently-completed influenza vaccination season, either in the facility/hospital or outside the facility/hospital; (2) those who were offered but declined the seasonal influenza vaccine; or (3) those who were ineligible due to contraindication(s). The numerator time window coincides with the most recently-completed seasonal influenza vaccination season (which begins on October 1, or when the vaccine first becomes available, and ends on March 31 of the following year).)Each criterion in the numerator is computed and reported separately.
  
  
• Denominator: The denominator consists of skilled nursing facility residents aged 180 days and older on target date of the assessment during the denominator time window. The denominator time window is defined as the most recently-completed influenza vaccination season, which begins on October 1 or when the vaccine first becomes available, and ends on March 31 of the following year.(The endorsed specifications of the measure are: The denominator consists of patients or short-stay residents aged 180 days of age and older on target date of assessment who were in the facility/hospital for at least one day during the denominator time window. The denominator time window is defined as the most recently-completed influenza vaccination season, which begins on October 1 and ends on March 31 of the following year. This measure is based on the NQF´s National Voluntary Standards for Influenza and Pneumococcal Immunizations.Note: The IRF-PAI data are submitted for Medicare patients only.)
  
  
• Exclusions: Residents whose age is 179 days or younger on target date of the selected assessment are excluded.(The endorsed specifications of the measure are: Residents or patients whose age is 179 days or less on target date of selected influenza vaccination assessment are excluded. Facilities with denominator counts of less than 20 residents/patients in the sample will be excluded from public reporting owing to small sample size.)
  
  
• HHS NQS Priority: Making Care Safer, Communication and Care Coordination, Effective Prevention and Treatment
  
  
• HHS Data Source: Nursing Home Minimum Data Set (MDS) Version 3.0
  
  
• Measure Type: Process
  
  
• Steward: Centers for Medicare & Medicaid Services
  
  
• Endorsement Status: Endorsed; MUC same as endorsed version: No
  
  
• Changes to Endorsed Measure Specifications?: The MUC list indicates the measure has not been modified from its endorsed version.
  
  
• Is the measure specified as an electronic clinical quality measure? No
  
  

Preliminary Analysis of Measure

• Preliminary analysis result: Encourage Continued Development
  
  
• Preliminary analysis summary
  • Contribution to program measure set:This measure addresses NQS priorities and contributes to overall PAC and SNF program goals of promoting preventive care to improve outcomes.
  • Impact on quality of care for patients:Influenza is a significant contributor to morbidity and mortality in the target population, and evidence indicates substantial room for improvement in the delivery of vaccinations to prevent the condition.
  
  
  
• Does the measure address a program goal or objective? Yes. This process measure is under consideration in order to meet program objectives to include metrics focused on making care safer and preventing illness.
  
  
• Is this a high-value measure? Yes. This is a process measure close to outcomes with a strong evidence link (decrease in influenza incidence)
  
  
• Does this measure fill a gap in the program measure set? Yes. The measure will fill gaps related to the National Quality Strategy, specifically: promoting the most effective prevention and treatment practices for the leading causes of mortality
  
  
• Measure development status: Field Testing
  
  
• Is the measure fully tested for the program's setting and level of analysis? Yes. Measure is a modification of an endorsed measure, and is being tested in the SNF resident population
  
  
• Is the measure currently in use? No. In use in other programs, different target population
  
  
• Does the MUC contribute to alignment and efficient use of measurement resources (burden and cost of measurement)? Yes. This measure is currently used In other federal programs, but with different target populations; it is being developed to harmonize with applicable related/competing measures.
  
  
• Is the measure NQF endorsed for the program's setting and level of analysis? No. Endorsed; MUC same as endorsed version: No
  
  
• Does the measure address a high-priority quality issue in the dual eligible beneficiary population? Yes.
  
  

Rationale for measure provided by HHS
Influenza and pneumonia are now reported as the fifth-leading cause of death among people aged 65 or older in the United States (CMS, 2011). As of 2011, there are over 200,000 hospitalizations from influenza, on average, every year (CMS, 2011). An average of 36,000 Americans die annually due to influenza and its complications and most are people 65 years old and over (CMS, 2011). Vaccination can be cost-effective and successful in preventing influenza. A study conducted in 2002 by Nichol and Goodman found that vaccination of healthy elderly was associated with a 36% reduction in hospitalization for pneumonia or influenza, an 18% reduction in hospitalization for all respiratory conditions, and a 40% reduction in mortality. (Nichol and Goodman, 2002) Influenza vaccination was also associated with cost savings. (Nichol and Goodman, 2002). Influenza vaccination is recommended for those over 65 years old and those with medical conditions, which describes the population of post-acute care facilities, making it an appropriate quality measure for skilled nursing facilities. By focusing on skilled nursing facility residents during the influenza season, publicly reporting this measure will increase vaccination during that time period and prevent influenza outbreaks in skilled nursing facilities. References: 1. Centers for Medicare & Medicaid Services (2011, May). Adult immunization: overview. Retrieved from https://www.cms.gov/adultImmunizations/ 2. Nichol KL, Goodman M., Cost effectiveness of influenza vaccination for healthy persons between ages 65 and 74 years. Vaccine. 2002 May 15;20(Suppl 2):S21-4.

Summary of NQF Endorsement Review

• Year of Most Recent Endorsement Review: 2011; 2012
  
  
• Project for Most Recent Endorsement Review: 1) National Voluntary Consensus Standards for Nursing Homes; 2) Ad-hoc Review: Expansion of Settings
  
  
• Review for Importance: See CDP Recommendation]
  
  
• Review for Scientific Acceptability: See CDP Recommendation]
  
  
• Review for Feasibility: See CDP Recommendation]
  
  
• Review for Usability: See CDP Recommendation]
  
  
• Review for Related and Competing Measures: See CDP Recommendation]
  
  
• Endorsement Committee Recommendation: 1) 2011 report (reviewed with NQF#681): Almost 60,000 deaths in 2004 were caused by influenza and pneumonia, and more than 85 percent of those were among the elderly.14Frail elderly are especially vulnerable and subject to complications of influenza. In the same year, approximately 123,000 death certificates identified influenza and pneumonia as a secondary cause of death. Further, the death rate from influenza and pneumonia is nearly130 times higher among persons aged 85 and older than among persons 45 to 54 years of age. The Committee unanimously agreed thesetwo measures are important and have strong evidence to support them. The Committee asked for and received assurance these measures are harmonized with other NQFvaccination measures. Further discussion focused on the definition of long-stay residents, the consequences of excluding missing data, and a specified time frame for vaccination.The Committee placed requirements on endorsement for this measure, requesting that for the long-stay measure the denominator should include only residents whose stay in the facility is longer than 100 days from the date of admission; the short-stay measure denominator should include only patients whose stay is 100 days or fewer. Second, patients with missing data in the MDS 3.0 should be counted as patients who did not receive the vaccine, rather than be excluded.Third, the numerator and denominator should be clearly harmonized with NQF’s previously endorsed vaccination measures. Finally, thetime frame for the seasonal vaccination should be altered to harmonize with the standard NQF influenza season definition. The steward agreed to meet these conditions for both measures, so the Committee voted to recommend these measures for endorsement. These measures meet the National Priority of Population Health. 2) 2012 (ad hoc review): The National Quality Forum (NQF) Board of Directors has approved an expansion of settings for two nursing home measures: #0680, Percent of Nursing Home Residents Who Were Assessed and Appropriately Given the Seasonal Influenza Vaccine (Short-Stay); and #0682, Percent of Residents Assessed and Appropriately Given the Pneumococcal Vaccine (Short-Stay).The Centers for Medicare & Medicaid Services, the measure steward, expanded the measures beyond the nursing home setting to include inpatient rehabilitation facilities and long-term acute care hospitals. The ad-hoc review request fell under the third NQF criterion for justifying a review: material changes to a currently endorsed measure.Technical experts were convened to assess the proposed measure changes and agreed to recommend expansion of both measures given the following reasons: the age range was expanded to reflect current evidence; the time window was made consistent with standard specifications and harmonized with other NQF-endorsed measures; data on the validity was provided; and the measure submission form was modified to ensure that the documentation supports the changes made.
  
  

Measure Specifications

• NQF Number (if applicable):
  
  
• Description: All-condition risk-adjusted potentially preventable hospital readmission rates
  
  
• Numerator: This measure does not have a simple form for the numerator and denominator. The numerator is defined as the risk-adjusted estimate of the number of unplanned, potentially preventable readmissions that occurred within 30 days post discharge from SNF services. The numerator, as defined, includes risk adjustment for patient characteristics and a statistical estimate of the facility effect beyond patient mix.
  
  
• Denominator: The denominator is computed with the same model used for the numerator. It is the model developed using all non-excluded stays in the national data. The measure includes all stays in the measurement period that are observed in national Medicare FFS data and do not fall into an excluded category. For a particular facility, the model is applied to the patient population, but the facility effect term is 0. In essence, it is the number of unplanned potentially preventable readmissions that would be expected for that patient population at the average facility.
  
  
• Exclusions: 1. Patients who died during the SNF stay. 2. Patients less than 18 years old. 3. Patients who were transferred to the same level of care or the hospital at the end of their SNF stay. 4. Patients who were not continuously enrolled in Part A FFS Medicare for the 12 months prior to the SNF admissions, and at least 30 days after SNF discharge. 5. Patients who did not have a short-term acute-care stay within 30 days prior to the SNF admission date. 6. Patients who leave the SNF against medical advice; 7. Patients for whom the prior short-term acute-care stay was for the nonsurgical treatment of cancer. 8. Patients who were transferred to a federal hospital from the SNF. 9. Patients who received care from a provider located outside of the US, Puerto Rico, or a US territory. 10. SNF stays with data that are problematic (e.g., anomalous records for hospital stays that overlap wholly or in part or are otherwise erroneous or contradictory).
  
  
• HHS NQS Priority: Communication and Care Coordination
  
  
• HHS Data Source: Claims: This measure is based on Medicare claims data for fee-for-service beneficiaries using SNFservices. Several variables from the inpatient claims and SNF claims will be used including admission/discharge dates and diagnostic/procedure information.
  
  
• Measure Type: Outcome
  
  
• Steward: Centers for Medicare & Medicaid Services
  
  
• Endorsement Status: Never Submitted
  
  
• Is the measure specified as an electronic clinical quality measure? No
  
  

Preliminary Analysis of Measure

• Preliminary analysis result: Encourage continued development
  
  
• Preliminary analysis summary
  • Contribution to program measure set:This is a high value measure that is not duplicative, captures a broader population of condition specific readmissions.
  • Impact on quality of care for patients:The potential impact of this measure could lead to decreased readmissions rates, decreased costs, and increases quality of care within the facility to ensure patients are properly diagnosed and discharged if needed. This measure can also support the care coordination needs of the patients upon discharge. Overall, this measure has potential to improve patient outcomes upon discharge to their communities from SNF.
  
  
  
• Does the measure address a program goal or objective? Yes. Addresses a high priority domain (Communication/Care Coordination), supports alignment across programs under the IMPACT Act, and addresses program gap in measurement.
  
  
• Is this a high-value measure? Yes. This is an outcome measure that addresses the Communication/Care Coordination high priority domain for future measure consideration under the program.
  
  
• Does this measure fill a gap in the program measure set? Yes. This measure of potentially preventable readmissions differs from the SNFRM (NQF #2510) which is a measure of all-cause unplanned readmissions. This new measure also looks at a different risk period, which is the 30 days following discharge from SNF services.This measure fills a gap by aligning across all conditions and examines potentially avoidable conditions for patient readmissions.
  
  
• Measure development status: Early Development
  
  
• Is the measure fully tested for the program's setting and level of analysis? No. Testing data not available
  
  
• Since no, could the measure be revised to use in the setting or at level of analysis under consideration? N/A. N/A
  
  
• Is the measure currently in use? No. New measure, never used in a program
  
  
• Does the MUC contribute to alignment and efficient use of measurement resources (burden and cost of measurement)? Yes. This measure is based on Medicare claims data for fee-for-service beneficiaries using SNF services. Several variables from the inpatient and SNF claims will be used including admission/discharge dates and diagnostic/procedure. Harmonized across different setting according to the IMPACT Act. Not duplicative due to specific care setting.
  
  
• Is the measure NQF endorsed for the program's setting and level of analysis? N/A. Never Submitted
  
  
• Does the measure address a high-priority quality issue in the dual eligible beneficiary population? Yes.
  
  

Rationale for measure provided by HHS
Hospital readmissions of Medicare beneficiaries discharged from a hospital to a skilled nursing facility (SNF) are prevalent and expensive, and prior studies suggest that a large proportion of readmissions are preventable (Mor et al., 2010). According to Mor et al., based an analysis of SNF data from 2006 Medicare claims merged with the Minimum Data Set (MDS), 23.5 percent of SNF stays resulted in a rehospitalization within 30 days of the initial hospital discharge. The average Medicare payment for each readmission was $10,352 per hospitalization, for a total of $4.34 billion. Of these rehospitalizations, 78 percent were deemed potentially avoidable, and applying this figure to the aggregate cost indicates that avoidable hospitalizations resulted in an excess cost of $3.39 billion (78 percent of $4.34 billion) to Medicare (Mor, Intrator, Feng, et al., 2010). Several analyses of hospital readmissions of SNF patients suggest there is opportunity for reducing hospital readmissions among SNF patients (Li et al., 2012; Mor et al., 2010), and multiple studies suggest SNF structural and process characteristics that impact readmission rates (Coleman et al., 2004; MedPAC 2011).

Measure Specifications

• NQF Number (if applicable):
  
  
• Description: All-condition risk-adjusted potentially preventable hospital readmission rates (required under PAMA)
  
  
• Numerator: This measure does not have a simple form for the numerator and denominator. The numerator is defined as the risk-adjusted estimate of the number of unplanned, potentially preventable readmissions that occurred within 30 days from discharge from the prior proximal acute hospitalization. The numerator, as defined, includes risk adjustment for patient characteristics and a statistical estimate of the facility effect beyond patient mix.
  
  
• Denominator: The denominator is computed with the same model used for the numerator. It is the model developed using all non-excluded stays in the national data. The measure includes all stays in the measurement period that are observed in national Medicare FFS data and do not fall into an excluded category. For a particular facility, the model is applied to the patient population, but the facility effect term is 0. In essence, it is the number of unplanned potentially preventable readmissions that would be expected for that patient population at the average facility.
  
  
• Exclusions: The following are the sample exclusions: 1. SNF stays where the patient had one or more intervening post-acute care (PAC) admissions which occurred either between the prior proximal hospital discharge and SNF admission or after the SNF discharge, within the 30-day risk window. Also excluded are SNF admissions where the patient had multiple SNF admissions after the prior proximal hospitalization, within the 30-day risk window. 2. SNF stays with a gap of greater than 1 day between discharge from the prior proximal hospitalization and the SNF admission. 3. SNF stays where the patient did not have at least 12 months of FFS Medicare enrollment prior to the proximal hospital discharge (measured as enrollment during the month of proximal hospital discharge and the for 11 months prior to that discharge). 4. SNF stays in which the patient did not have FFS Medicare enrollment for the entire risk period (measured as enrollment during the month of proximal hospital discharge and the month following the month of discharge). 5. SNF stays in which the principal diagnosis for the prior proximal hospitalization was for the medical treatment of cancer. Patients with cancer whose principal diagnosis from the prior proximal hospitalization was for other diagnoses or for surgical treatment of their cancer remain in the measure. 6. SNF stays where the patient was discharged from the SNF against medical advice. 7. SNF stays in which the principal primary diagnosis for the prior proximal hospitalization was for “rehabilitation care; fitting of prostheses and for the adjustment of devices” 8. SNF stays in which the prior proximal hospitalization was for pregnancy.
  
  
• HHS NQS Priority: Making Care Safer
  
  
• HHS Data Source: Administrative claims
  
  
• Measure Type: Outcome
  
  
• Steward: Centers for Medicare & Medicaid Services
  
  
• Endorsement Status: Never Submitted
  
  
• Is the measure specified as an electronic clinical quality measure? No
  
  

Preliminary Analysis of Measure

• Preliminary analysis result: Encourage Continued Development
  
  
• Preliminary analysis summary
  • Contribution to program measure set:Outcome measures are considered of high-value for program implementation. In addition, this MUC meets a PAMA legislation mandate for the implementation of an outcome measure focusing on potentially preventable readmissions.
  • Impact on quality of care for patients:Several analyses of hospital readmissions of SNF patients suggest there is opportunity for reducing hospital readmissions among SNF patients (Li et al., 2012; Mor et al., 2010), and multiple studies suggest SNF structural and process characteristics that impact readmission rates (Coleman et al., 2004; MedPAC 2011).
  
  
  
• Does the measure address a program goal or objective? Yes. The MUC meets a PAMA legislation mandate.for the implementation of an outcome measure focusing on potentially preventable readmissions.
  
  
• Is this a high-value measure? Yes. This is an outcome measure.
  
  
• Does this measure fill a gap in the program measure set? Yes. Legislation mandates that a measure focused on potentially preventable readmissions be implemented by October 2016, and replace the existing all-cause re-admission measure.
  
  
• Measure development status: Early Development
  
  
• Is the measure fully tested for the program's setting and level of analysis? Yes. MUC is being tested for the SNF population.
  
  
• Is the measure currently in use? No.
  
  
• Does the MUC contribute to alignment and efficient use of measurement resources (burden and cost of measurement)? Yes. The measure is considered to be of high-value as an outcome measure, is aligned with other MUCs proposed for PAC/LTC programs, and is not duplicative.
  
  
• Is the measure NQF endorsed for the program's setting and level of analysis? No. Never Submitted
  
  
• Does the measure address a high-priority quality issue in the dual eligible beneficiary population? Yes.
  
  

Rationale for measure provided by HHS
Hospital readmissions of Medicare beneficiaries discharged from a hospital to a SNF are prevalent and expensive, and prior studies suggest that a large proportion of readmissions are preventable (Mor et al., 2010). According to Mor et al., based an analysis of SNF data from 2006 Medicare claims merged with the Minimum Data Set (MDS), 23.5 percent of SNF stays resulted in a rehospitalization within 30 days of the initial hospital discharge. The average Medicare payment for each readmission was $10,352 per hospitalization, for a total of $4.34 billion. Of these rehospitalizations, 78 percent were deemed potentially avoidable, and applying this figure to the aggregate cost indicates that avoidable hospitalizations resulted in an excess cost of $3.39 billion (78 percent of $4.34 billion) to Medicare (Mor, Intrator, Feng, et al., 2010). Several analyses of hospital readmissions of SNF patients suggest there is opportunity for reducing hospital readmissions among SNF patients (Li et al., 2012; Mor et al., 2010), and multiple studies suggest SNF structural and process characteristics that impact readmission rates (Coleman et al., 2004; MedPAC 2011).

Program Index

• Home Health Quality Reporting Program 
  
  
• Hospice Quality Reporting Program 
  
  
• Inpatient Rehabilitation Facility Quality Reporting Program 
  
  
• Long-Term Care Hospital Quality Reporting Program 
  
  
• Skilled Nursing Facility Quality Reporting System 
  
  
• Skilled Nursing Facility Value-Based Purchasing Program 
  
  

Full Program Summaries

Program History and Structure: The Home Health Quality Reporting Program (HH QRP) was established in accordance with section 1895 (b)(3)(B)(v)(II) of the Social Security Act. Home Health Agencies (HHAs) are required by the Act to submit quality data for use in evaluating quality for Home Health agencies. Section 1895(b) (3)(B)(v)(I) of the Act also requires that HHAs that do not submit quality data to the Secretary be subject to a 2 percent reduction in the annual payment update, effective in calendar year 2007 and every subsequent year. Data sources for the HH QRP include the Outcome and Assessment Information Set (OASIS) and Medicare FFS claims. Data is publically reported on the Home Health Compare website. The HH QRP measure development and selection activities take into account established national priorities and input from multi-stakeholder groups.

Further, the Improving Medicare Post-Acute Care Transformation (IMPACT Act of 2014, amends title XVIII (Medicare) of the Social Security Act (the Act) to direct the Secretary of the Department of Health and Human Services (HHS) to require Long-term Care Hospitals (LTCH), Inpatient Rehabilitation Facilities (IRF), Skilled Nursing Facilities (SNF) and Home Health Agencies (HHA) to report data on resource use and other measures and standardized patient assessment data on quality measures and specified assessment domains via the assessment instruments that are used to submit assessment data to CMS by these post-acute care (PAC) providers; amending the Social Security Act for each of the provider types to add such requirements under the IMPACT Act. The IMPACT Act delineates the reporting of standardized assessment data on quality measures in at least the following domains: Functional status, cognitive function, and changes in function and cognitive function; skin integrity and changes in skin integrity; medication reconciliation; incidence of major falls; transfer of health information when the individual transitions from the hospital/critical access hospital to PAC provider or home, or from PAC provider to another settings. It also delineates the implementation of resource use and other measures in at least these following domains: Total estimated Medicare spending per beneficiary Discharge to the community, all condition risk adjusted potentially presentable hospital readmission rates. Further, the IMPACT Act requires the modification of such assessment instruments to achieve the standardization of such data.

High Priority Domains for Future Measure Consideration: CMS identified the following domains as high-priority for future measure consideration:

1. Patient and Family Engagement: Quality care in home health settings should addressed not only assessing for what the patient/family desires, but also to assess how well care is provided and what services are offered to meet an individual’s care preferences.
2. Patient and Family Engagement: Functional status and functional decline are important to assess for individuals who reside in a home-based setting. Individuals who receive care in home-based settings may have functional limitations and may be at risk for further decline in function due to limited mobility and ambulation. Therefore, measures to assess functional status are in development.
3. Making Care Safer: Safety for individuals in a home-based setting is an important priority for the HH QRP as persons in home health settings are at risk for major injury due to falls, new or worsened pressure ulcers, pain, and functional decline. Therefore, these concepts will be considered for future measure development.
4. Making Care Affordable: An important consideration for the HH QRP is to better assess medical costs based on PAC episodes of care. Therefore, CMS is considering developing efficiency-based measures such as a Medicare Spending per Beneficiary measure concept.
5. Communication/Care Coordination: Assessing an individual’s care transitions and rehospitalizations is important. Therefore, CMS is considering developing measures that assesses discharge to the community and potentially preventable readmissions.
6. Communication/Care Coordination: Infrastructure and processes for care coordination are important for the HH QRP. Therefore, a medication reconciliation quality measure for individuals in a home health setting is being considered for future quality measure development. Medication reconciliation conceptually highlights care transitions and resident follow-up.

Current Measures: NQF staff have compiled the program's measures in a spreadsheet organized according to concepts.

Program History and Structure: The Hospice Quality Reporting Program (HQRP) was established in accordance with section 1814(i) of the Social Security Act, as amended by section 3004(c) of the Affordable Care Act. The HQRP applies to all hospices, regardless of setting. Proposed data sources for future HQRP measures include the Hospice Item Set and the Hospice Consumer Assessment of Healthcare Providers and Systems (CAHPS) questionnaire. HQRP measure development and selection activities take into account established national priorities and input from multi-stakeholder groups. Beginning in FY 2014, Hospices that fail to submit quality data will be subject to a 2.0 percentage point reduction to their annual payment update.

High Priority Domains for Future Measure Consideration: CMS identified the following domains as high-priority for Hospice QRP future measure consideration:

1. Overall goal HQRP: Symptom Management Outcome Measures. There is a lack of tested and endorsed outcome measures for hospice across domains of hospice care, including symptom management (e.g.; physical and other symptoms). Developing and implementing outcome measures for hospice is important for providers, patients and families, and other stakeholders because symptom management is a central aspect of hospice care.
2. Communication/Care Coordination and/or Patient and Family Engagement: Patient preference for care is difficult to measure at end of life when patients may or may not be able to state their preferences, and may have changes in their preferences. However, a central tenet of hospice care is responsiveness to patient and family care preferences; as much as possible, patient preferences should be incorporated into new measure development.
3. Patient and Family Engagement: Measurement of goal attainment is naturally linked to determining patient/family preferences. Quality care in hospice should address not only establishing what the patient/family desires but also providing care and services in line with those preferences.
4. Making Care Safer: Timeliness/responsiveness of care. While timeliness of referral to hospice is not within a hospices’ control, hospice initiation of treatment once a patient has elected the hospice benefit is under the control of the hospice. Responsiveness of the hospice during times of patient or family need is an important indicator about hospice services for consumers in particular.
5. Communication/Care Coordination: Measurement of care coordination is integral to the provision of quality care and should be aligned across care settings.

Current Measures: NQF staff have compiled the program's measures in a spreadsheet organized according to concepts.

Program History and Structure: The Quality Reporting Program (QRP) for Inpatient Rehabilitation Facilities (IRFs) was established in accordance with section 1886(j) of the Social Security Act as amended by section 3004(b) of the Affordable Care Act. The IRF QRP applies to all IRF facilities that receive the IRF PPS (e.g., IRF hospitals, IRF units that are co-located with affiliated acute care facilities, and IRF units affiliated with critical access hospitals [CAHs]). Data sources for IRF QRP measures include Medicare FFS claims, the Center for Disease Control’s National Health Safety Network (CDC NHSN) data submissions, and Inpatient Rehabilitation Facility - Patient Assessment instrument (IRF-PAI) records. The IRF QRP measure development and selection activities take into account established national priorities and input from multi-stakeholder groups. Beginning in FY 2014, IRFs that fail to submit data will be subject to a 2.0 percentage point reduction of the applicable IRF Prospective Payment System (PPS) payment update. Plans for future public reporting of IRF QRP measures are under development.

Further, the Improving Medicare Post-Acute Care Transformation (IMPACT Act of 2014, amends title XVIII (Medicare) of the Social Security Act (the Act) to direct the Secretary of the Department of Health and Human Services (HHS) to require Long-term Care Hospitals (LTCH), Inpatient Rehabilitation Facilities (IRF), Skilled Nursing Facilities (SNF) and Home Health Agencies (HHA) to report data on resource use and other measures and standardized patient assessment data on quality measures and specified assessment domains via the assessment instruments that are used to submit assessment data to CMS by these post-acute care (PAC) providers; amending the Social Security Act for each of the provider types to add such requirements under the IMPACT Act. The IMPACT Act delineates the reporting of standardized assessment data on quality measures in at least the following domains: Functional status, cognitive function, and changes in function and cognitive function; skin integrity and changes in skin integrity; medication reconciliation; incidence of major falls; transfer of health information when the individual transitions from the hospital/critical access hospital to PAC provider or home, or from PAC provider to another settings. It also delineates the implementation of resource use and other measures in at least these following domains: Total estimated Medicare spending per beneficiary Discharge to the community, all condition risk adjusted potentially presentable hospital readmission rates. Further, the IMPACT Act requires the modification of such assessment instruments to achieve the standardization of such data.

High Priority Domains for Future Measure Consideration: CMS identified the following four domains as high-priority for future measure consideration:

1. Making Care Safer (subdomains: hospital-acquired infections and hospital-acquired conditions): Patient safety is an important priority domain for the IRF QRP as IRF patients are at risk for injury due to falls, new or worsened pressure ulcers and infections such as CAUTI, C. Diff. and MRSA.
2. Patient and Family Engagement: A primary focus of IRF care is restoring functional status. Metrics showing change in self-care and mobility function and discharge self-care and mobility are under development. Metric for achievement of functional status goals such as discharge to community. In addition, the experiences of patients and caregivers are important to measure and are important priority for the IRF QRP.
3. Making Care Affordable: An important consideration for the IRF QRP is to better assess medical costs based on PAC episodes of care. Therefore, CMS is considering developing efficiency-based measures such as a Medicare Spending per Beneficiary measure concept.
4. Communication/Care Coordination: Assessing patient care transitions and rehospitalizations are important. Therefore, CMS is considering developing measures that assesses discharge to the community and potentially preventable readmissions.
5. Communication/Care Coordination: Infrastructure and processes for care coordination are important for the IRF QRP. Therefore, a medication reconciliation quality measure for IRF patients is being considered for future quality measure development. Medication reconciliation conceptually highlights care transitions and resident follow-up.

Current Measures: NQF staff have compiled the program's measures in a spreadsheet organized according to concepts.

Program History and Structure: The Long-Term Care Hospital (LTCH) Quality Reporting Program (QRP) was established in accordance with section 1886(m) of the Social Security Act, as amended by Section 3004(a) of the Affordable Care Act. The LTCH QRP applies to all LTCHs facilities designated as an LTCH under the Medicare program. Data sources for LTCH QRP measures include Medicare FFS claims, the Center for Disease Control and Prevention’s National Health Safety Network (CDC’s NHSN) data submissions, and the LTCH Continuity Assessment Record and Evaluation Data Sets (LCDS). The LTCH QRP measure development and selection activities take into account established national priorities and input from multi-stakeholder groups. Beginning in FY 2014, LTCHs that fail to submit data will be subject to a 2.0 percentage point reduction of the applicable Prospective Payment System (PPS) increase factor.

Further, the Improving Medicare Post-Acute Care Transformation (IMPACT Act of 2014, amends title XVIII (Medicare) of the Social Security Act (the Act) to direct the Secretary of the Department of Health and Human Services (HHS) to require Long-term Care Hospitals (LTCH), Inpatient Rehabilitation Facilities (IRF), Skilled Nursing Facilities (SNF) and Home Health Agencies (HHA) to report data on resource use and other measures and standardized patient assessment data on quality measures and specified assessment domains via the assessment instruments that are used to submit assessment data to CMS by these post-acute care (PAC) providers; amending the Social Security Act for each of the provider types to add such requirements under the IMPACT Act. The IMPACT Act delineates the reporting of standardized assessment data on quality measures in at least the following domains: Functional status, cognitive function, and changes in function and cognitive function; skin integrity and changes in skin integrity; medication reconciliation; incidence of major falls; transfer of health information when the individual transitions from the hospital/critical access hospital to PAC provider or home, or from PAC provider to another settings. It also delineates the implementation of resource use and other measures in at least these following domains: Total estimated Medicare spending per beneficiary Discharge to the community, all condition risk adjusted potentially presentable hospital readmission rates. Further, the IMPACT Act requires the modification of such assessment instruments to achieve the standardization of such data.

High Priority Domains for Future Measure Consideration: CMS identified the following domains as high-priority for LTCH QRP future measure consideration:

1. Effective Prevention and Treatment: Having measures related to ventilator use, ventilator-associated event and ventilator weaning rate are a high priority for CMS as prolonged mechanical ventilator use is quite common in LTCHs and respiratory diagnosis with ventilator support for 96 or more hours is the most frequently occurring diagnosis.
2. Effective Prevention and Treatment (Aim: Healthy People/Healthy Communities): In discussions with LTCH providers, it was noted that mental health status is an important measure of care for LTCH patients. CMS is considering a Depression Assessment & Management quality measure.
3. Patient and Family Engagement: While rehabilitation and restoring functional status are not the primary goals of patient care in the LTCH setting, functional outcomes remain an important indicator of LTCH quality as well as key to LTCH care trajectories. Providers must be able to provide functional support to patients with impairments. Thus, metrics showing change in self-care and mobility function are under development.
4. Patient and Family Engagement: CMS would like to explore measures that will evaluate the patient’s experiences of care as this is a high priority of providers. Therefore, the HCAHPS and Care Transition quality measure (CTM)-3 is being considered.
5. Making Care Affordable: An important consideration for the LTCH QRP is to better assess medical costs based on PAC episodes of care. Therefore, CMS is considering developing efficiency-based measures such as a Medicare Spending per Beneficiary measure concept.
6. Communication/Care Coordination: Assessing patient care transitions and rehospitalizations are important. Therefore, CMS is considering developing measures that assesses discharge to the community and potentially preventable readmissions.
7. Communication/Care Coordination: Infrastructure and processes for care coordination are important for the LTCH QRP. Therefore, a medication reconciliation quality measure for LTCH patients is being considered for future quality measure development. Medication reconciliation conceptually highlights care transitions and resident follow-up.

Current Measures: NQF staff have compiled the program's measures in a spreadsheet organized according to concepts.

Program History and Structure: The Improving Medicare Post-Acute Care Transitions Act of 2014 (The IMPACT Act) added Section 1899B to the Social Security Act establishing the Skilled Nursing Facility (SNF) Quality Reporting Program (QRP). Facilities that submit data under the SNF PPS are required to participate in the SNF QRP, excluding units that are affiliated with critical access hospitals (CAHs). Data sources for SNF QRP measures include Medicare FFS claims as well as Minimum Data Set (MDS) assessment data. The SNF QRP measure development and selection activities take into account established national priorities and input from multi-stakeholder groups. Beginning in FY 2018, providers that fail to submit required quality data to CMS will have their annual updates reduced by 2.0 percentage points.

Further, the Improving Medicare Post-Acute Care Transformation (IMPACT Act of 2014, amends title XVIII (Medicare) of the Social Security Act (the Act) to direct the Secretary of the Department of Health and Human Services (HHS) to require Long-term Care Hospitals (LTCH), Inpatient Rehabilitation Facilities (IRF), Skilled Nursing Facilities (SNF) and Home Health Agencies (HHA) to report data on resource use and other measures and standardized patient assessment data on quality measures and specified assessment domains via the assessment instruments that are used to submit assessment data to CMS by these post-acute care (PAC) providers; amending the Social Security Act for each of the provider types to add such requirements under the IMPACT Act. The IMPACT Act delineates the reporting of standardized assessment data on quality measures in at least the following domains: Functional status, cognitive function, and changes in function and cognitive function; skin integrity and changes in skin integrity; medication reconciliation; incidence of major falls; transfer of health information when the individual transitions from the hospital/critical access hospital to PAC provider or home, or from PAC provider to another settings. It also delineates the implementation of resource use and other measures in at least these following domains: Total estimated Medicare spending per beneficiary Discharge to the community, all condition risk adjusted potentially presentable hospital readmission rates. Further, the IMPACT Act requires the modification of such assessment instruments to achieve the standardization of such data.

High Priority Domains for Future Measure Consideration: CMS identified the following four domains as high-priority for future measure consideration:

1. Patient and Family Engagement: Functional status and functional decline are important to assess for residents in SNF settings. Residents who receive care while in a SNF may have functional limitations and may be at risk for further decline in function due to limited mobility and ambulation. Therefore, measures to assess functional status are in development.
2. Making Care Safer: Resident safety is an important priority domain for the SNF QRP as persons in SNF settings are at risk for major injury due to falls, new or worsened pressure ulcers, pain, and functional decline. Therefore, these concepts will be considered for future measure development.
3. Making Care Affordable: An important consideration for the SNF QRP is to better assess medical costs based on PAC episodes of care. Therefore, CMS is considering developing efficiency-based measures such as a Medicare Spending per Beneficiary measure concept.
4. Communication/Care Coordination: Assessing resident care transitions and rehospitalizations are important. Therefore, CMS is considering developing measures that assesses discharge to the community and potentially preventable readmissions.
5. Communication/Care Coordination: Infrastructure and processes for care coordination are important for the SNF QRP. Therefore, a medication reconciliation quality measure for SNF residents is being considered for future quality measure development. Medication reconciliation conceptually highlights care transitions and resident follow-up.

Current Measures: NQF staff have compiled the program's measures in a spreadsheet organized according to concepts.

Program History and Structure: The Skilled Nursing Facility Value-Based Purchasing (SNF-VBP) Program was established by Section 215 (b)of the Protecting Access to Medicare Act of 2014. The facility adjusted Federal per diem rate will be reduced by 2% and an incentive payment will then be applied to facilities based upon readmission measure performance.

The legislation mandates that CMS will specify a SNF all-cause all-condition hospital readmission measure by no later than October 1, 2015. It further requires that a resource use measure that reflects resource use by measuring all-condition risk-adjusted potentially preventable hospital readmission rates for SNFs will be specified no later than October 1, 2016 and replace the all-cause all-condition measure as soon as is practicable.

High Priority Domains for Future Measure Consideration: CMS identified the following categories as high-priority for future measure consideration:

1. The sole measure requirement at this time is the specification of a potentially preventable readmission measure. CMS lacks the authority to implement additional measures beyond the two described in the statute.

Measure Requirements: CMS applies criteria for measures that may be considered for potential adoption in the SNF-VBP program. At a minimum, the following requirements must be met for selection in the SNF-VBP program:

• Must meet statutory requirements for all-condition potentially preventable hospital readmissions measure for SNFs.
• Must provide documentation sufficient to complete MUC list required data fields.
• Measures should be NQF endorsed, save where due consideration is given to endorsed measures of the same specified area or medical topic.
• May incorporate Medicare claims and/or alternative data sources will be considered dependent upon available infrastructure.

Current Measures: NQF staff have compiled the program's measures in a spreadsheet organized according to concepts.

Index of Measures (by Program)

General Comments

• General comments
  
  

• Discharge to Community-Post Acute Care (PAC) Home Health Quality Reporting Program (Required under the IMPACT Act) (Public comments received:9; MUC ID: MUC15-523)
  
  
• Drug Regimen Review Conducted with Follow-Up for Identified Issues-Post Acute Care (PAC) Home Health Quality Reporting Program (Required under the IMPACT Act) (Public comments received:7; MUC ID: MUC15-1127)
  
  
• Falls risk composite process measure (Public comments received:5; MUC ID: MUC15-207)
  
  
• Improvement in Dyspnea in Patients with a Primary Diagnosis of Congestive Heart Failure, Chronic Obstructive Pulmonary Disease and/or Asthma (Public comments received:5; MUC ID: MUC15-235)
  
  
• Medicare Spending Per Beneficiary-Post Acute Care (PAC) Home Health Quality Reporting Program (Required under the IMPACT Act) (Public comments received:5; MUC ID: MUC15-1134)
  
  
• Potentially Preventable 30-Day Post-Discharge Readmission Measure for Home Health Quality Reporting Program (Required under the IMPACT Act) (Public comments received:11; MUC ID: MUC15-234)
  
  

• Hospice and Palliative Care Composite Process Measure (Public comments received:6; MUC ID: MUC15-231)
  
  
• Hospice Visits When Death Is Imminent (Public comments received:6; MUC ID: MUC15-227)
  
  

• Discharge to Community-Post Acute Care (PAC) Inpatient Rehabilitation Facility Quality Reporting Program (Required under the IMPACT Act) (Public comments received:7; MUC ID: MUC15-408)
  
  
• Drug Regimen Review Conducted with Follow-Up for Identified Issues-Post Acute Care (PAC) Inpatient Rehabilitation Facility Quality Reporting Program (Required under the IMPACT Act) (Public comments received:5; MUC ID: MUC15-1128)
  
  
• Medicare Spending per Beneficiary-Post Acute Care (PAC) Inpatient Rehabilitation Facility Quality Reporting Program (Required under the IMPACT Act) (Public comments received:4; MUC ID: MUC15-287)
  
  
• Potentially Preventable 30-Day Post-Discharge Readmission Measure for Inpatient Rehabilitation Facility Quality Reporting Program (Required under the IMPACT Act) (Public comments received:8; MUC ID: MUC15-496)
  
  
• Potentially Preventable Within Stay Readmission Measure for Inpatient Rehabilitation Facilities (Public comments received:6; MUC ID: MUC15-497)
  
  

• Compliance with Spontaneous Breathing Trial (SBT) (including Tracheostomy Collar Trial (TCT) or Continuous Positive Airway Pressure (CPAP) Breathing Trial)) by Day 2 of the LTCH Stay (Public comments received:1; MUC ID: MUC15-400)
  
  
• Discharge to Community-Post Acute Care (PAC) Long-Term Care Hospital Quality Reporting Program (Required under the IMPACT Act) (Public comments received:6; MUC ID: MUC15-414)
  
  
• Drug Regimen Review Conducted with Follow-Up for Identified Issues-Post Acute Care (PAC) Long-Term Care Hospital Quality Reporting Program (Required under the IMPACT Act) (Public comments received:2; MUC ID: MUC15-1129)
  
  
• Medicare Spending per Beneficiary-Post Acute Care (PAC) Long-Term Care Hospital Quality Reporting Program (Required under the IMPACT Act) (Public comments received:3; MUC ID: MUC15-289)
  
  
• Percent of Patients Who Received an Antipsychotic (AP) Medication (Public comments received:1; MUC ID: MUC15-530)
  
  
• Potentially Preventable 30-Day Post-Discharge Readmission Measure for Long-Term Care Hospital Quality Reporting Program (Required under the IMPACT Act) (Public comments received:5; MUC ID: MUC15-498)
  
  
• Ventilator Weaning (Liberation) Rate (Public comments received:1; MUC ID: MUC15-398)
  
  

• Application of IRF Functional Outcome Measure: Change in Mobility Score for Medical Rehabilitation Patients (NQF #2634) (Public comments received:9; MUC ID: MUC15-527)
  
  
• Application of IRF Functional Outcome Measure: Change in Self-Care Score for Medical Rehabilitation Patients (NQF #2633) (Public comments received:10; MUC ID: MUC15-236)
  
  
• Application of IRF Functional Outcome Measure: Discharge Mobility Score for Medical Rehabilitation Patients (NQF #2636) (Public comments received:9; MUC ID: MUC15-529)
  
  
• Application of IRF Functional Outcome Measure: Discharge Self-Care Score for Medical Rehabilitation Patients (NQF #2635) (Public comments received:9; MUC ID: MUC15-528)
  
  
• Discharge to Community-Post Acute Care (PAC) Skilled Nursing Facility Quality Reporting Program (Required under the IMPACT Act) (Public comments received:9; MUC ID: MUC15-462)
  
  
• Drug Regimen Review Conducted with Follow-Up for Identified Issues-Post Acute Care (PAC) Skilled Nursing Facility Quality Reporting Program (Required under the IMPACT Act) (Public comments received:9; MUC ID: MUC15-1130)
  
  
• Medicare Spending per Beneficiary-Post Acute Care (PAC) Skilled Nursing Facility Quality Reporting Program (Required under the IMPACT Act) (Public comments received:6; MUC ID: MUC15-291)
  
  
• Percent of Skilled Nursing Facility Residents Who Newly Received an Antipsychotic Medication (Public comments received:3; MUC ID: MUC15-1133)
  
  
• Percent of Skilled Nursing Facility Residents Who Self-Report Moderate to Severe Pain (Public comments received:2; MUC ID: MUC15-1131)
  
  
• Percent of Skilled Nursing Facility Residents Who Were Assessed and Appropriately Given the Influenza Vaccine (Public comments received:4; MUC ID: MUC15-1132)
  
  
• Potentially Preventable 30-Day Post-Discharge Readmission Measure for Skilled Nursing Facility Quality Reporting Program (Required under the IMPACT Act) (Public comments received:10; MUC ID: MUC15-495)
  
  

• Skilled Nursing Facility 30-Day Potentially Preventable Readmission Measure (SNFPPR) (required by PAMA) (Public comments received:8; MUC ID: MUC15-1048)
  
  

Full Comments (Listed by Measure)

• The AAFP urges the MAP to be aware and cognizant of the efforts of the multi-stakeholder Core Quality Measures Collaborative while reviewing measures for inclusion for federal programs. This effort works to ensure alignment, harmonization, and the avoidance of competing quality measures among payers. These sets contain a variety of measure types targeting both resource use, quality, and patient satisfaction. The AAFP has a general concern that many of these measures are currently under development, or are in field testing. Because of this, there is no clear way of knowing if they will measure what they intend to measure, or if there are unintended consequence that have yet been uncovered. We propose to include only fully develop measured in this year’s rule making. (Submitted by: American Academy of Family Physicians )
  
  
• On behalf of Adventist Health System (AHS), I wish to provide the following overall comments. The timeframe of this comment opportunity is too short to provide substantive comments about the potential of these Measures under Consideration to add value, improve outcomes, create undue data collection burden or access whether a better measure exists. We urge the Measure Applications Partnership (MAP) to ensure that the measures that the Centers for Medicare and Medicaid Services (CMS) is considering for use in federal programs are fully developed, tested and vetted to ensure reliability and validity. We believe that the National Quality Forum (NQF) should not offer support or even “conditional” support for any measures that have not received NQF endorsement. In addition, the MAP should assess all Measures under Consideration for alignment with measures currently in use in federal programs. This assessment should ensure that measures are unique and well-correlated with outcomes that are meaningful to patients. Priority should be given to the most valid, reliable and meaningful measures. Any recommendation for a new measure that competes with a measure currently in use in a federal program should include a condition that the competing measure be removed from the federal program or a condition that the two measures be harmonized prior to the adoption of the new measure. (Submitted by: Adventist Health System)
  
  
• Thank you for the opportunity to review and provide comment on the Measures Under Consideration. The AAN continues to have a concern over the lack of neurology specific measures available through the different CMS programs especially given the growing population of aging adults. With the prevalence of chronic neurologic disease continuing to grow in the aging population and the relative number of available neurologists decreasing we must do something to assist treating physicians participate in applicable programs. The AAN believes that expanding the list of measures to be more inclusive of specialties will only serve to benefit of populations with neurological needs. (Submitted by: American Academy of Neurology)
  
  

• We recommend against approval of the measures at this time because they are not fully developed and ask the NQF MAP PAC/LTC workgroup to exercise their option to vote to "encourage further development" of this set of measures. The rationale for this recommendation is supported by the facts that: • many of the measures are not fully specified yet – the initial CMS comment periods for some of these measures is ongoing at this point (e.g., Discharge to Community and Potentially Preventable Hospitalization) and no specifications have been released for others (e.g., Spending per Beneficiary); • the intended uses for each measure have not been specified; • while some are existing measures, they have not been tested for specific purposes or are now being recommended for use in settings where they have not been tested (e.g., IRF measures now being applied to the SNF setting); • some have never been used previously and many have had no testing for reliability and validity. These measures remain early in their development process and do not meet NQF MAP criteria for recommendation at this time. We are concerned that the timeline for the process of review, stakeholder input and testing has been shortened unrealistically and has not allowed for adequate vetting of the measures and their potential applications. A more thorough and transparent process is required before these can and should be used in programs that have the potential to significantly affect provider behavior and practices. Please keep in mind as you discern these measures, that the SNF does not have authority to control conditions once a patient leaves our setting. It is the SNF responsibility to set up a proper discharge plan, provide proper patient/care giver education; however, patient choice to not follow recommendations is beyond our control and authority. Thank you for your consideration of these comments. (Submitted by: Benedictine Health System)
  
  
• AHCA is pleased to have the opportunity to comment on the List of Measures under Consideration for December 1, 2015”. The American Health Care Association (AHCA) represents more than 12,000 non-profit and proprietary skilled nursing centers and assisted living communities. Rather than having many of our individual members who have contacted us with comments and to more efficiently provide NQF with feedback, we have received from our various members, we have summarized their comments here. Thank you again for the opportunity to provide these comments. Please contact me at DGifford@ahca.org with any questions or for additional information. A preliminary draft yet incomplete specifications for this measure were provided by CMS for public comment, with the deadline for comment being December 1st, 2015, which is after the MUC list was published. No information on reliability or validity of the measure is available and the risk adjustment methodology is not fully specified. In addition, CMS has provided inadequate time for public review and comment of this measure. In fact, the TEP never saw the preliminary draft full specification of this measure until it was made available for public comment in late November 2105 after the TEP meeting. Therefore, for these reasons and our concerns detailed below, we believe this measure does not meet NQF MAP criteria for endorsement but rather should receive a vote of “encourage continued development” This measure (MUC15-1048) double counts readmissions with other SNF potentially preventable readmission measure (MUC15-495). Measure MUC15-1048 includes readmission that occur within a SNF stay and after a SNF stay. Thus, this measure will capture the same readmissions as in measure MUC14-495 resulting in a double counting of measures. This measure is developed for the SNF VBP. The SNF VPB statute does not require readmissions after SNF stay to be included. The statute only states that (g) Skilled nursing facility readmission measure (1) Readmission measure Not later than October 1, 2015, the Secretary shall specify a skilled nursing facility all-cause all-condition hospital readmission measure (or any successor to such a measure). (2) Resource use measure Not later than October 1, 2016, the Secretary shall specify a measure to reflect an all-condition risk-adjusted potentially preventable hospital readmission rate for skilled nursing facilities. https://www.govtrack.us/congress/bills/113/hr4302/text/enr#link=II_215_a_~T1&nearest=H77F0E089948D48C8A638611B926D2ABF In addition, this measure does not harmonized with the within stay IRF measure MUC15-497, which only counts readmissions that occur during the IRF stay. Thus, this measure needs to be revised to only count readmission during the SNF stay. We have additional concerns with this measure that mirror our comments on measure MUC15-495 related to: 1. The name of the measures should reflect the limited population to which they apply – fee-for service (FFS) Medicare beneficiaries; particularly since in many states 40% or more of Medicare beneficiaries are enrolled in MA plans and for SNFs over half of SNF admissions and discharges are not enrolled in FFS Medicare. 2. The specifications are based on ICD-9 but all providers as of October 2015 are required to use ICD-10 and no cross walk with ICD-10 is provided. 3. The numerator definition and methodology of calculating a “predicted actual” is extremely confusing, which makes the data less likely to be used and is of questionable benefit. Data showing how this approach is superior to using an actual rate divided by the expected rate needs to be provided. 4. Many of the potentially preventable admissions contains diagnoses that do not meet the proposed definition of potentially preventable readmissions; therefore, these should be dropped. 5. The process for developing the list of potentially preventable admissions used existing literature but would have benefited from a more formal process such as the RAND appropriateness rating system (e.g. modified Delphi approach) rather than ad hoc TEP and consultant experts. 6. Risk adjustment does not included a. SES characteristics, an issue identified in the last round of NQF readmission measure reviews as a requirement. b. Functional status, one of the strongest predictors of hospitalizations. All the PAC settings are now required to utilize standard functional status assessment – Section GG from the CARE tool. c. The risk adjustment variables are not specifically specified with respect to data source and categories but are given as examples with just a descriptor. For example is age continuous or categorical variable and if categorical how and from what source? d. Prior utilization variables indicate they “vary by measure” but how they vary by measure is not provided. 7. Exclusions are confusing and need some modifications a. The denominator exclusions appear confusing and overlapping so that it is unclear if discharges from IRF or LTCH to SNF or HH are included in the IRF and LTCH denominator. Similarly, if SNF discharges to HH are included. b. We agree that AMA discharges from PAC provider should be excluded but so should hospital discharges that are AMA but end up in PAC provider. 8. The list of related to other NQF endorsed measures is incomplete and only compares to CMS or RTI endorsed measures. All other NQF endorsed measures should be included. 9. We agree with the expansion of unplanned readmission diagnoses from YALE’s list, but will this also be applied to the SNFRM and hospital readmission measures? (Submitted by: American Health Care Association )
  
  
• this measure is redundant and overlapping with MUC15-495 adn seems in conflict with the IRF measure MUC15-497. the list of preventable diagnoses includes several that are NOT preventable by the SNF and do not reflect the care provided. This is an incomplete measure, as written and we would recommend additional work (Submitted by: leadingage)
  
  
• While written comments were not provided, the commenter indicated their support for this measure in this program (Submitted by: Smith & Nephew )
  
  
• Avoiding potentially preventable readmissions is an important component of improving resident satisfaction and avoiding unnecessary healthcare spending. The key to this measure being useful and appropriate for inclusion is the ability to consistently and effectively distinguish between preventable and unpreventable hospital visits. SHE was not able to find information about how these determinations weremade. Could such a measure lead to resident selection that would serve as a barrier to care skilled nursing facility care for persons thought to be at higher risk of “unplanned” hospital visits (e.g., those with more severe underlying illness)? (Submitted by: The Society for Healthcare Epidemiology of America)
  
  
• AMDA is pleased to have the opportunity to comment on the List of Measures under Consideration for December 1, 2015”. AMDA - The Society for Post-Acute and Long-Term Care Medicine (AMDA) represents over 5,500 physician medical directors, attending physicians, and other practitioners who work in post-acute and long-term care settings. Thank you again for the opportunity to provide these comments. Please contact me at abardakh@amda.com with any questions or for additional information. A preliminary draft yet incomplete specifications for this measure were provided by CMS for public comment, with the deadline for comment being December 1st, 2015, which is after the MUC list was published. No information on reliability or validity of the measure is available and the risk adjustment methodology is not fully specified. In addition, CMS has provided inadequate time for public review and comment of this measure. In fact, the TEP never saw the preliminary draft full specification of this measure until it was made available for public comment in late November 2105 after the TEP meeting. Therefore, for these reasons and our concerns detailed below, we believe this measure does not meet NQF MAP criteria for endorsement but rather should receive a vote of “encourage continued development” This measure (MUC15-1048) double counts readmissions with other SNF potentially preventable readmission measure (MUC15-495). Measure MUC15-1048 includes readmission that occur within a SNF stay and after a SNF stay. Thus, this measure will capture the same readmissions as in measure MUC14-495 resulting in a double counting of measures. This measure is developed for the SNF VBP, but the definition of a “potentially preventable” readmission is not stated. The SNF VPB statute does not require readmissions after SNF stay to be included. The statute states that: (g) Skilled nursing facility readmission measure (1) Readmission measure Not later than October 1, 2015, the Secretary shall specify a skilled nursing facility all-cause all-condition hospital readmission measure (or any successor to such a measure). (2) Resource use measure Not later than October 1, 2016, the Secretary shall specify a measure to reflect an all-condition risk-adjusted potentially preventable hospital readmission rate for skilled nursing facilities. https://www.govtrack.us/congress/bills/113/hr4302/text/enr#link=II_215_a_~T1&nearest=H77F0E089948D48C8A638611B926D2ABF In addition, this measure does not harmonized with the within stay IRF measure MUC15-497, which only counts readmissions that occur during the IRF stay. Thus, this measure needs to be revised to only count readmission during the SNF stay. We have additional concerns with this measure that mirror our comments on measure MUC15-495 related to: 1. The name of the measures should reflect the limited population to which they apply – fee-for service (FFS) Medicare beneficiaries; particularly since in many states 40% or more of Medicare beneficiaries are enrolled in MA plans and for SNFs over half of SNF admissions and discharges are not enrolled in FFS Medicare. 2. The specifications are based on ICD-9 but all providers as of October 2015 are required to use ICD-10 and no cross walk with ICD-10 is provided. 3. The numerator definition and methodology of calculating a “predicted actual” is extremely confusing, which makes the data less likely to be used and is of questionable benefit. Data showing how this approach is superior to using an actual rate divided by the expected rate needs to be provided. 4. Many of the potentially preventable admissions contains diagnoses that do not meet the proposed definition of potentially preventable readmissions; therefore, these should be dropped. 5. The process for developing the list of potentially preventable admissions used existing literature but would have benefited from a more formal process such as the RAND appropriateness rating system (e.g. modified Delphi approach) rather than ad hoc TEP and consultant experts. 6. Risk adjustment does not included a. SES characteristics, an issue identified in the last round of NQF readmission measure reviews as a requirement. b. Functional status, one of the strongest predictors of hospitalizations. All the PAC settings are now required to utilize standard functional status assessment – Section GG from the CARE tool. c. The risk adjustment variables are not specifically specified with respect to data source and categories but are given as examples with just a descriptor. For example is age continuous or categorical variable and if categorical how and from what source? d. Prior utilization variables indicate they “vary by measure” but how they vary by measure is not provided. 7. Exclusions are confusing and need some modifications a. The denominator exclusions appear confusing and overlapping so that it is unclear if discharges from IRF or LTCH to SNF or HH are included in the IRF and LTCH denominator. Similarly, if SNF discharges to HH are included. b. We agree that AMA discharges from PAC provider should be excluded but so should hospital discharges that are AMA but end up in PAC provider. 8. The list of related to other NQF endorsed measures is incomplete and only compares to CMS or RTI endorsed measures. All other NQF endorsed measures should be included. 9. We agree with the expansion of unplanned readmission diagnoses from YALE’s list, but will this also be applied to the SNFRM and hospital readmission measures? (Submitted by: AMDA-The Society for Post-Acute and Long-Term Care Medicine)
  
  
• Initially, we would urge you to vote to require additional development on this proposed set of measures. We are concerned that these measures are not fully developed and require additional testing. In support of this position, we would submit that: • Many of the measures are not fully specified and public comments are still pending; in fact, Golden Living submitted comments last week on Potentially Preventable Hospital Rehospitalization. • No specifications have been released for certain measures, for example, the measure on Spending per Beneficiary. • Intended uses for each measure have not been specified. • While some of the proposed measures are existing measures, they have not been tested for specific purposes, or are now being submitted for use in settings where they have not been tested, for example IRF measures would be applied to SNF settings. • Some measures have never been used and many have not been tested for reliability and validity. We are concerned that many of these measures are only in the development phase and do not meet NQF MAP criteria for recommendation at this time. We are also concerned about the challenging timeframe for review and stakeholder input that has been applied to these measures. We urge you to allow adequate time for vetting, comment, and testing of these measures and their application. In addition to urging additional development of all the measures, we would submit the following comments on specific measures. • NQF #2633 is primarily intended for use by inpatient rehab facilities and while it shares some elements also included in Section GG on the MDS for implementation October 2016, there are additional elements in this measure not included on the MDS. These include, in the Self Care Area – wash upper body, shower/bathe self, upper body dressing, lower body dressing, putting on/taking off footwear; and in the Mobility Area – roll left and right, care transfer, walk 10 feet, walking 10 feet on uneven surface, 1 step (curb), 4 steps, 12 steps, and picking up object. Since the IMPACT Act mandates reporting of “standardized data set” from all PAC settings, we would urge you to require providers in all settings to report the same items. • Discharge to Community-Post Acute Care (PAC) Skilled Nursing Facility Quality Reporting Program – We generally believe that this measure insures a consistent, standardized approach to calculation of the measure for all SNFs. • Percent of Skilled Nursing Facility Residents Who Self-Report Moderate to Severe Pain – While we generally support this measure, we urge the development of a companion measure to report the percentage of short-stay patients who demonstrate a change in reporting pain during the stay. This would measure the effectiveness of the facility’s effort to address and manage the pain. Thank you for your attention to the concerns of Golden Living. Candace Bartlett National Senior Director Regulatory Affairs Golden Living 8930 Valley Creek Dr. Arlington, TN 38002 (901) 504-8313 robin.bartlett@goldenliving.com (Submitted by: Golden Living)
  
  
• At least one NRHI member has expressed concerns regarding this measure. Specifically, clinical staff at HealthInsight have concerns with inclusion of this measure due to the validity of and complicated methodologies used for "risk adjustment for patient characteristics and a statistical estimate of the facility effect beyond patient mix." This methodology is frequently cited as the basis for determining potentially avoidable hospital facility admissions. This methodology is not well-established. (Submitted by: Network for Regional Healthcare Improvement)
  
  

• In general, NAHC supports the Drug Regimen Review measure as proposed. However, to ensure accurate measure comparison and data exchange among the post-acute care providers, we urge the developers to establish clear definitions for the terms “clinically significant” medication issues and “one calendar day”. (Submitted by: National Association for Home Care & Hospice (NAHC))
  
  
• • MUC15-1127: Drug Regimen Review Conducted with Follow-Up for Identified Issues Post Acute Care (PAC) Home Health Quality Reporting Program (Required under the IMPACT Act)- The numerator contains components of a drug regimen review, as well as a follow up with the physician or physician designee. The numerator does not specifically state there is a process of medication reconciliation. Medication reconciliation is mentioned in the rationale for this measure. There is no mention of medications being prescribed by specialists that may not be relevant to the acute care stay. We recommend continued development of the measure under consideration because we feel there should be an emphasis on medication reconciliation, drug/drug interactions, food/drug interactions and compliance with medications. The way this measure under consideration is designed, it minimally assesses the process of reviewing medication and reporting this review to the physician. (Submitted by: LeadingAge)
  
  
• While we recognize the inclusion of a Medication Reconciliation measure as part of the IMPACT Act requirements, we are highly concerned that the measure under consideration is a process measure that will not evaluate the quality of care being provided, but instead will measure the effectiveness/efficiency of the data collection of a check-box or Yes/No value. Furthermore, the requirements of participation in Medicare payment systems for many of the PAC venues already include considerations for Medication Reconciliation and/or Drug Regimen Review, suggesting that this process is already in place for PAC sites, and that the quality measure results should show near 100% compliance for all facilities. We would also note that in CMS' public comment summary for this measure, created on November 6, 2015, the "Preliminary Recommendations" state the plan for CMS and the measure developers to "Continue measure testing and development" as well as planning on "Conducting pilot testing to assess feasibility, reliability, and validity of assessment data". This suggests that the measure has not yet been finalized, and also may not NQF requirements for measure endorsement. (Submitted by: UDSMR)
  
  
• The Alliance supports the use of this measure, but is concerned about the lack of clarity involving the definition of key terms and administrative burden associated with the measure, and the need for adequate testing before such a measure is applied broadly. The Alliance is concerned that there are not clear definitions for the term “potential clinically significant” medication issues, as well as for what constitutes “significant drug interactions,” and “significant side effects,” “any potential adverse effects.” Greater clarity is also needed to understand what constitutes ineffective drug therapy. Moreover, the Alliance is concerned about burden associated with this measure. The follow up time with the physician or physician designee must be clearly defined, but also adaptive to the risk and urgency of follow-up with a physician. We are also concerned about whether agencies will be penalized if physicians are not responsive to home health agency follow-up. Unfortunately, physicians often are not conscientious about follow up with home health agencies that are reaching out regarding their patients. Finally, the Alliance urges testing of this measure and reconsideration before it is finalized. Testing and validation should be no less than six months with an opportunity to modify the measure prior to finalizing it. A similar approach was used for many of the OASIS-based measures that CMS uses for home health agencies. (Submitted by: Alliance for Home Health Quality and Innovation)
  
  
• We support the inclusion of this cross-setting measure in the home health quality reporting program and believe that cross-setting comparisons are appropriate. (Submitted by: Association of Home & Hospice Care of NC/SC Home Care & Hospice Association)
  
  
• At least one NRHI member has expressed concerns regarding this measure. Specifically, clinical staff at HealthInsight have indicated a lack of clarity surrounding the methodologies that can be used to determine "follow up with a physician" and "physician designee." (Submitted by: Network for Regional Healthcare Improvement)
  
  
• This measure is currently in the early stage of development, has never been submitted to NQF and thus is not NQF-endorsed (Submitted by: AdvaMed)
  
  

• While we recognize the inclusion of a Medication Reconciliation measure as part of the IMPACT Act requirements, we are highly concerned that the measure under consideration is a process measure that will not evaluate the quality of care being provided, but instead will measure the effectiveness/efficiency of the data collection of a check-box or Yes/No value. Furthermore, the requirements of participation in Medicare payment systems for many of the PAC venues already include considerations for Medication Reconciliation and/or Drug Regimen Review, suggesting that this process is already in place for PAC sites, and that the quality measure results should show near 100% compliance for all facilities. We would also note that in CMS' public comment summary for this measure, created on November 6, 2015, the "Preliminary Recommendations" state the plan for CMS and the measure developers to "Continue measure testing and development" as well as planning on "Conducting pilot testing to assess feasibility, reliability, and validity of assessment data". This suggests that the measure has not yet been finalized, and also may not NQF requirements for measure endorsement. (Submitted by: UDSMR)
  
  
• This measure is currently under development. We have concerns that, as drafted, the information form on which CMS sought public comment describes a process that is inconsistent with hospital practices. The form describes how a PAC setting should reconcile medications and then contact the physician within one day. In the hospital, it is the physician who takes the initiative to review and reconcile medications that the patient was taking prior to admission, and then writes orders for those to be continued, deleted, or changed. We recommended to CMS changes that reflect standard hospital practice. Furthermore, the information form states that the data associated with this measure will be collected via the Inpatient Rehabilitation Facility Patient Assessment Instrument (IRF-PAI). While the IRF-PAI is an excellent instrument for this type of data collection, significant changes were made to it in the Fiscal Year (FY) 2016 Inpatient Rehabilitation Facility Prospective Payment System (IRF PPS) final rule expanding it in length from eight to eighteen pages. Therefore, any additions to the IRF-PAI should be considered carefully in terms of any additional administrative burden, even if such additions are limited to a few items or questions. Additionally, changes to the IRF-PAI can require staff training, the hiring of additional staff, and changes to electronic health records and billing and documentation software all of which take time and can be costly. These changes should be considered when developing the timeline for modifying the IRF-PAI to give the field adequate time to make the required changes. Finally, having a mechanism in place to validate the data associated with this measure to ensure it is entered and completed is critically important. As proposed, a provider could easily check a box “yes” to indicate the requirements of the measure were met. We believe the documentation or other form of verification that this measure was completed accurately is important and we will work with CMS and NQF to develop such a mechanism that minimize the administrative burden. Until these modifications are made we do not believe the measure fulfills its promise to improve outcomes for patients. Also, it is still in the development phase within CMS. Therefore, it may be premature to endorse this measure for the purpose of the IRF QRP. (Submitted by: American Medical Rehabilitation Providers Association)
  
  
• The proposed measure “Percentage of care episodes or stays in which a drug regimen review was conducted at the Admission /SOC /ROC, and timely follow-up with a physician occurred each time potentially significant medication issues were identified throughout the care episode or stay” is intended to be collected by 3 items: • Item 1 asks providers if a complete drug regimen review conducted at admission identified any “potentially clinically significant issues” for a patient. • Item 2 asks providers if they contacted a physician within one calendar day and completed the recommended course of action for the “identified clinically significant medication issues” found at admission, if any. • Item 3 asks providers if they contacted a physician and completed the recommended actions within one calendar day each time clinically significant medication issues were identified following admission. While the IMPACT Act designated “medication reconciliation” as one of several domains for which cross-setting quality measures are required, the proposed measure instead utilizes “drug regimen review” to calculate the measure numerator. Although there is an overlap between “medication reconciliation” and “drug regimen review,” drug regimen review is a fundamentally distinct process from medication reconciliation. According to the definition of “drug regimen review” as stated by the Home Health (HH) Conditions of Participation (CoPs) at §484.55c, the intent is to “review medications a patient is currently using in order to identify potential adverse effects or drug reactions.” This most notably includes noncompliance with drug therapy, significant side effects, and ineffective drug therapy – all of which are typically infeasible for an admitting facility to assess during a medication reconciliation process at admission. Such drug regimen review is distinct from the medication reconciliation act of “comparing the medications a patient is taking (and should be taking) with newly ordered medications in order to identify and resolve discrepancies, including omissions, duplications, contraindications, unclear information, and changes.” Furthermore, the proposed drug regimen review tracks medication issues at any point during the care episode or stay (dividing the episode of care into “admission” and “after admission”) even though medication reconciliation occurs only during transitions of care, i.e. admission, transfer, and discharge. According to The Joint Commission, medication reconciliation occurs at transition points of care where new medications are ordered or existing orders are rewritten. The Institute for Healthcare Improvement also considers the medication reconciliation process to occur at transition points only: “Each time a patient moves from one setting to another where orders change or must be renewed, clinicians should review previous medication orders alongside new orders and plans for care, and reconcile any differences. If this process does not occur in a standardized manner that is designed to ensure complete reconciliation, medication errors may lead to adverse events and harm.” A measure that tracks variables associated with drug regimen review throughout the patient’s stay (as proposed by the measure) is fundamentally different from a measure that tracks variables associated with the medication reconciliation processes that occur only when the patient moves from one setting to another. Additionally, by lumping discharge into the overly-general “after admission” timeframe, the proposed measure also fails to adequately capture the discharge process, even though CMS’ proposed Drug Regimen Review Measure Justification Form lists the importance of discharge as a critical part of medication reconciliation: “Hospital discharge is one high risk time point with evidence that there are high levels of discrepancy. In fact, there is evidence that 50% of patients experience a clinically important medication error after hospital discharge.” A thorough and complete medication reconciliation as part of the discharge planning process is critical for a patient’s successful transition to her/his discharge destination, whether to home or community or to another inpatient care setting. It is troubling that the proposed medication reconciliation measure fails to distinguish this important aspect of a patient’s episode of care in a post-acute setting. We encourage CMS to propose a “medication reconciliation” measure that focuses on medication review at transition points of care and remove elements related to drug regimen review, as they fundamentally differ in scope and definition. Our comments related to the existing measure as proposed follow: COMMENTS RELATED TO ITEMS 1 AND 2 Drug regimen review activities that occur prior to the medication order The term “medication reconciliation” as defined by The Joint Commission NPSG 03.06.01, and used in the proposed measure, involves comparing “the medication information the patient brought to the hospital [either from the discharging facility or from home] with the medications ordered for the patient by the hospital in order to identify and resolve discrepancies.” Based on this definition, medication reconciliation only includes medications that have been ordered for the patient, but not medications that were prevented from being ordered by a drug regimen process. For example, hospitals that use an electronic medical record (“EMR”) typically utilize computer physician order entry (“CPOE”) that has safeguards in place to prevent a physician from ordering medications which may have potential adverse effects and drug reactions for the particular patient. Since these medications were never ordered, they would not be part of the medication reconciliation and drug regimen review as defined by the proposed quality measure. We ask that CMS clarify whether the measure is intended to include or exclude instances where a drug was reviewed for potential adverse effects and drug reactions prior to the medication actually being ordered (thus eliminating potential adverse effects and drug reactions). For example, patients admitted to inpatient rehabilitation hospitals are typically on anticoagulation therapy. If a physician attempts to submit an order for the anticoagulant Coumadin in the EMR, a CPOE may trigger an alert to the physician that this medication duplicates the existing order for Lovenox, another anticoagulant. The physician could then cancel the prescription and the medication would never be ordered, thus avoiding a potential “duplicate drug therapy” result. Because Coumadin was never actually “ordered,” the process engineered by the CPOE’s alert system would not be captured as an act of medication reconciliation in the proposed quality measure. It is unclear whether such processes were intended to be captured as part of the drug regimen review since, in this illustrative example, the potentially clinically significant medication issues were avoided by virtue of a drug order never being executed in the first place. Further, the physician can accept or override this type of alert in a CPOE system depending on patient-specific circumstances and his/her own clinical judgment. If they choose to override the alert, the medication would be ordered and therefore become subject to a medication reconciliation process and drug regimen review. While an EMR allows these changes and alerts to be identified and tracked before a medication is actually ordered, this would be far more difficult to capture in a non-electronic system, if these decisions occurred prior to the ordering process. Referring to the previous anticoagulation example, if a physician considered writing a prescription for Coumadin but caught this potential “duplicate drug therapy” issue prior to writing the actual order, no one would ever know. Definition of “Clinically significant medication issue” We encourage the measure developers to specifically define, and NQF to ask, what constitutes a “potential clinically significant medication issue.” Without an explicit and detailed definition of this term, it will likely vary widely across providers and care settings, which will in turn cause serious integrity issues for the resulting quality measure data. It is also important to specify what would not constitute a “potential clinically significant medication issue.” For HealthSouth’s recommendations of “potential clinically significant medication issue,” please see Appendix A. Subjectivity of “Potential” The proposed measure asks clinicians to predict if a medication issue (potential adverse effect or drug reaction) found during a drug regimen review would be “potentially clinically significant.” While some medication issues might undoubtedly cause clinically significant issues (however such issues are defined), others would be left to subjective judgment about whether the issue would or would not potentially cause clinically significant patient effects. This type of subjectivity is concerning when developing a measure that will be utilized by a wide variety of clinicians across post-acute care settings, and one that is intended to achieve standardized data points for subsequent comparison between such settings. As such, this high level of subjectivity would be a significant concern to reliability and validity of the proposed quality measure. Item 1 – Answers “0/1/9” It is unclear how a hospital would answer Item 1 if they found medication issues that were considered “issues,” but not deemed to be “potentially clinically significant.” It is implied that 0 “No, no issues found” and 1 “Yes, issues found” apply only to “potentially clinically significant medication issues” (particularly given the logic that an answer of “1” requires Item 2 to be completed), not any “issue” at all. Therefore, if a hospital finds a medication issue they deem not clinically significant (by the defined term), there is no available answer option. Accordingly, we recommend that, for answers 0 and 1, the language be clarified from “issues” to “potential clinically significant medication issues” and an additional selection be added for “Yes, medication issues were identified but were not deemed potentially clinically significant.” This additional answer choice would also presumably skip Item 2, like current selections “0” and “9”. Item 2 - Medication Follow-Up Under Item 2, it is unclear whether the facility/agency has one calendar day to “complete prescribed/recommended actions in response to the identified clinically significant medication issue” or whether that applies only to contacting the physician (or physician designee). If the proposed measure is intended to capture whether the physician was contacted and the recommendation action all occurred in one calendar day, this could be reworded to clarify. Notably, in Item 3, providers are given one calendar day for both actions. Admission Clarification The assessment timing for Item 1 and Item 2 are identified as “admission” for SNF, IRF, LTCH, and SOC/ROC for HH. However, “admission” remains undefined. Admission orders at a receiving IRF (and likely SNF and LTCH) typically follow the discharging orders from the prior level of care until a full reconciliation of the medical record and medication history can occur. This includes an evaluation by the pharmacist, a discussion with the patient and/or family, and a review by a physician. While medication review and reconciliation are initiated immediately upon admission it can take up to 48 hours for a full “admission medication reconciliation” to be completed. Therefore, we suggest the “admission” timeframe be 2 calendar days – the day of admission and the following calendar day. COMMENTS RELATED TO ITEM 3 The proposed quality measure information begins with a definition of medication reconciliation and drug regimen review: Medication reconciliation – the process of comparing the medications a patient is taking (and should be taking) with newly ordered medications in order to identify and resolve discrepancies. (Reference: The Joint Commission, National Patient Safety Goals). Drug regimen review – a review of all medications the patient is currently using in order to identify any potential adverse effects and drug reactions, including ineffective drug therapy, significant side effects, significant drug interactions, duplicate drug therapy, and noncompliance with drug therapy. (Reference: Home Health Conditions of Participation Home Health §484.55c). Medication reconciliation and drug regimen review are intended to identify and avoid potential medication issues, but the wording of Item 3 omits the word “potential.” We ask CMS to clarify whether Item 3 is intended to measure potential clinically significant medication issues or clinically significant medication issues that have already occurred. This is a critically important distinction. Reconciling medication instructions before the administration of medication is a structured, deliberate clinical activity. But handling a clinically significant issue, such as an adverse drug reaction, is highly variable and may call for anything from the application of a reversal agent to readmission to a general acute hospital. Treating these two types of medication-related events under one measure conflates two processes that are fundamentally different from one another. While we agree that measuring the timeliness of notifying the physician after a clinically significant medication issue is extremely important to patient safety, we believe it is outside the scope of “medication reconciliation” – which is intended to identify potential issues before medication is administered. Measuring physician notification after a clinically significant medication event happens is a separate and distinct quality measure, and one that should almost never take a full calendar day to occur. ADDITIONAL COMMENTS Unintended Consequences The purpose of medication reconciliation is to find and correct errors. No measure, either now or in the future, should be designed so as to discourage clinicians from reporting and correcting errors in patient medication. Numerator/Denominator Exclusions Patients who were admitted and unexpectedly discharged prior to a medication reconciliation process being completed should be excluded from this measure. It is assumed this would be indicated on the respective assessment instruments with a dash (“-“) as currently documented for other quality measures. Care Setting Differences The drug regimen review that may occur in a home health setting would be significantly different from an inpatient setting where the patient is receiving 24-hour care. If clinically significant medication issues occur in the inpatient setting, they are handled more immediately than in home health settings. Larger Emphasis on Medication Reconciliation at Discharge We encourage CMS to consider an aspect of “medication reconciliation” specifically at the discharge timeframe. A thorough and complete medication reconciliation and drug regimen review at this point in the patient’s care is part of the discharge planning process, as it is critical for a patient’s successful transition to her/his discharge destination, whether to home or community or to another inpatient care setting. (Submitted by: HealthSouth)
  
  
• At least one NRHI member has expressed concerns regarding this measure. Specifically, clinical staff at HealthInsight have indicated a lack of clarity surrounding the methodologies that can be used to determine "follow up with a physician" and "physician designee." (Submitted by: Network for Regional Healthcare Improvement)
  
  
• This measure is currently in the early stage of development, has never been submitted to NQF and thus is not NQF-endorsed (Submitted by: AdvaMed)
  
  

• While we recognize the inclusion of a Medication Reconciliation measure as part of the IMPACT Act requirements, we are highly concerned that the measure under consideration is a process measure that will not evaluate the quality of care being provided, but instead will measure the effectiveness/efficiency of the data collection of a check-box or Yes/No value. Furthermore, the requirements of participation in Medicare payment systems for many of the PAC venues already include considerations for Medication Reconciliation and/or Drug Regimen Review, suggesting that this process is already in place for PAC sites, and that the quality measure results should show near 100% compliance for all facilities. We would also note that in CMS' public comment summary for this measure, created on November 6, 2015, the "Preliminary Recommendations" state the plan for CMS and the measure developers to "Continue measure testing and development" as well as planning on "Conducting pilot testing to assess feasibility, reliability, and validity of assessment data". This suggests that the measure has not yet been finalized, and also may not NQF requirements for measure endorsement. (Submitted by: UDSMR)
  
  
• This measure is currently in the early stage of development, has never been submitted to NQF and thus is not NQF-endorsed (Submitted by: AdvaMed)
  
  

• This measure does not distinguish between Drug Regimen Review and Medication Reconciliation. Within the measure description, it does not define "potentially significant medication issues". That is too broad of a statement and comparisons between providers will not be accurate. Thank you for your consideration of these comments. (Submitted by: Benedictine Health System)
  
  
• This measure is currently in early development and measure specifications are in preliminary stages of clarification. The proposed application could create inaccuracy in measurement of patient condition and payment. The Numerator requires “documentation of drug regimen” as opposed to the IMPACT Act of 2014 requirement for “Medication Reconciliation”. The terms need to be better defined to assure compliance with the IMPACT Act provisions. Once this has been done, exclusions need to be defined for clinical reasonableness as well as payment logic. This measure needs extensive work to assure accuracy and to meet the purpose for which it was intended. Because there is potential for accuracy this measure should continue to be developed. (Submitted by: Genesis Health Care)
  
  
• While we recognize the inclusion of a Medication Reconciliation measure as part of the IMPACT Act requirements, we are highly concerned that the measure under consideration is a process measure that will not evaluate the quality of care being provided, but instead will measure the effectiveness/efficiency of the data collection of a check-box or Yes/No value. Furthermore, the requirements of participation in Medicare payment systems for many of the PAC venues already include considerations for Medication Reconciliation and/or Drug Regimen Review, suggesting that this process is already in place for PAC sites, and that the quality measure results should show near 100% compliance for all facilities. We would also note that in CMS' public comment summary for this measure, created on November 6, 2015, the "Preliminary Recommendations" state the plan for CMS and the measure developers to "Continue measure testing and development" as well as planning on "Conducting pilot testing to assess feasibility, reliability, and validity of assessment data". This suggests that the measure has not yet been finalized, and also may not NQF requirements for measure endorsement. (Submitted by: UDSMR)
  
  
• AHCA is pleased to have the opportunity to comment on the List of Measures under Consideration for December 1, 2015”. The American Health Care Association (AHCA) represents more than 12,000 non-profit and proprietary skilled nursing centers and assisted living communities. Rather than having many of our individual members who have contacted us with comments and to more efficiently provide NQF with feedback, we have received from our various members, we have summarized their comments here. Thank you again for the opportunity to provide these comments. Please contact me at DGifford@ahca.org with any questions or for additional information. This measure has not been fully specified. In fact, the TEP discussion raised some serious concerns about the measure which it is unclear if they have been addressed. The data source for this measure has not been specified nor has any of the data elements needed for this measure been tested for reliability or validity. In fact, CMS has just called to find 4 SNFs to pilot this measure. This measure also does not appear to be consistent with the legislative intent of the IMPACT act to develop a medication reconciliation measure. This a drug regime review for appropriateness of medications not a measure of reconciliation. Therefore, we believe this measure does not meet NQF MAP criteria for endorsement but rather should receive a vote of “encourage continued development” Based on information available, we have the following specific concerns with this measure. 1. The measure plans to use data elements from PAC assessments that have not yet been specified and therefore, the data does not currently exist., In fact, CMS just called for SNFs to pilot test a data collection tool in 4 SNFs. 2. The measure uses a definition of Drug Regimen Review which is defined from Conditions of Participation for Home Health. This definition is not consistent with other PAC settings. In particular, the definition in this measure is inadequate in capturing the scope of a drug/medication regimen review and does not address involvement of interdisciplinary team members. 3. The measure description includes measurement of follow up on potentially significant medication issues but does not define what constitutes a potentially significant medication issue. Lack of clear definition for this critical part of the measure will result in inconsistent data collection and measurement. 4. It is not clearly stated if this measure will only apply to short stay or also to long stay residents/patients in SNFs. 5. There is no information on reliability or validity for this measure. 6. This measure has not been tested nor has an examination of feasibility for implementing this measure across PAC settings been completed. A recent notice on this measure stated “RTI is conducting a pilot test to investigate data collection methods and the feasibility of implementing a cross-setting DRR quality measure”. During a recent inquiry, we were advised only 4 SNF, 4 LTCH and 4 IRF will be participating in this pilot. This is only 0.0002 percent of SNFs in the country which is an inadequate representation. (Submitted by: American Health Care Association )
  
  
• This measure is somewhat confusing. How does this relate to the already regualtory requirement of DDR, as defined by the CMS Requirement of Participation? Furthermore, the time lines are unclear, as is the term "significant change". it also doesn't describe what the scope of this review is to be. We believe this measure needs further consideration and testing. (Submitted by: leadingage)
  
  
• AMDA is pleased to have the opportunity to comment on the List of Measures under Consideration for December 1, 2015”. AMDA - The Society for Post-Acute and Long-Term Care Medicine (AMDA) represents over 5,500 physician medical directors, attending physicians, and other practitioners who work in post-acute and long-term care settings. Thank you again for the opportunity to provide these comments. Please contact me at abardakh@amda.com with any questions or for additional information. This measure has not been fully specified. In fact, the TEP discussion raised some serious concerns about the measure which it is unclear if they have been addressed. The data source for this measure has not been specified nor has any of the data elements needed for this measure been tested for reliability or validity. In fact, CMS has just called to find 4 SNFs to pilot this measure. This measure also does not appear to be consistent with the legislative intent of the IMPACT act to develop a medication reconciliation measure. This a drug regime review for appropriateness of medications not a measure of reconciliation. Therefore, we believe this measure does not meet NQF MAP criteria for endorsement but rather should receive a vote of “encourage continued development” Based on information available, we have the following specific concerns with this measure. 1. The measure plans to use data elements from PAC assessments that have not yet been specified and therefore, the data does not currently exist., In fact, CMS just called for SNFs to pilot test a data collection tool in 4 SNFs. 2. The measure uses a definition of Drug Regimen Review which is defined from Conditions of Participation for Home Health. This definition is not consistent with other PAC settings. In particular, the definition in this measure is inadequate in capturing the scope of a drug/medication regimen review and does not address involvement of interdisciplinary team members. 3. The measure description includes measurement of follow up on potentially significant medication issues but does not define what constitutes a potentially significant medication issue. Lack of clear definition for this critical part of the measure will result in inconsistent data collection and measurement. 4. It is not clearly stated if this measure will only apply to short stay or also to long stay residents/patients in SNFs. 5. There is no information on reliability or validity for this measure. 6. This measure has not been tested nor has an examination of feasibility for implementing this measure across PAC settings been completed. A recent notice on this measure stated “RTI is conducting a pilot test to investigate data collection methods and the feasibility of implementing a cross-setting DRR quality measure”. During a recent inquiry, we were advised only 4 SNF, 4 LTCH and 4 IRF will be participating in this pilot. This is only 0.0002 percent of SNFs in the country which is an inadequate representation. (Submitted by: AMDA-The Society for Post-Acute and Long-Term Care Medicine)
  
  
• At least one NRHI member has expressed concerns regarding this measure. Specifically, clinical staff at HealthInsight have indicated a lack of clarity surrounding the methodologies that can be used to determine "follow up with a physician" and "physician designee." (Submitted by: Network for Regional Healthcare Improvement)
  
  
• This measure is currently in the early stage of development, has never been submitted to NQF and thus is not NQF-endorsed (Submitted by: AdvaMed)
  
  
• To Whom It May Concern: The American Society of Consultant Pharmacists is the only international professional society devoted to optimal medication management and improved health outcomes for all older persons. ASCP's members manage and improve drug therapy and improve the quality of life of geriatric patients and other individuals residing in a variety of environments, including nursing facilities, sub-acute care and assisted living facilities, psychiatric hospitals, hospice programs, and home and community-based care. ASCP has a long history of advocating for the medical best interests of people who reside in long- term care facilities and those enrolled in hospice programs. We are pleased to have the opportunity to comment on the “List of Measures under Consideration for December 1, 2015.” Our specific comments relate to Measures Under Consideration (MUC) 15-1130: Drug Regimen Review. ASCP supports quality measures and believes strongly, based on experience and available data, that appropriate drug regimen review (DRR), combined with timely medication reconciliation, are among the most effective methods by which to ensure the health and safety of LTC residents. However, implementation of IMPACT Act quality measures has been a hurried process that has not allowed stakeholder groups to adequately review the measures, draft and submit comments. NQF, CMS, and CMS contractors ABT Associates and RTI have all been involved in the measures related to the IMPACT Act. There has been confusion on the meaning of terminology, especially “medication reconciliation” versus DRR. Although ASCP supports DRR and medication reconciliation quality measures, we ask the National Quality Forum (NQF) to “encourage continued development” in order to gather further input from providers with clinical expertise. The definition of Medication Reconciliation is not clear in any of the proposed measures that we reviewed. ASCP envisions medication reconciliation as an important quality indicator involving participation from members across disciplines and institutions. The process of reconciliation and the objective criteria required to ensure compliance, as well as the responsibility of providers across care settings should be clearly stated. The proposed measure should be tested for feasibility and refinement using a statistically valid sample in order to arrive at a measure that has meaning and can be replicated over the 16,000 skilled nursing facilities around the country. We appreciate NQF’s work on this important issue and would like to note that the process for vetting these quality measures has not met CMS’ normal standard of thoroughness and consideration. NQF provided notice after the comment period had begun and many of the important constituents within the LTC community have not had the opportunity to engage the process. ASCP also proposes that NQF further refine this measure. This is an excellent opportunity to add an important quality measure, but the measure should be as clear and consistent as possible before it is finalized. We look forward to working on this important process. Sincerely Sincerely, Frank Grosso, RPh Executive Director and CEO American Society of Consultant Pharmacist (Submitted by: American Society of Consultant Pharmacist )
  
  

• AHCA is pleased to have the opportunity to comment on the List of Measures under Consideration for December 1, 2015”. The American Health Care Association (AHCA) represents more than 12,000 non-profit and proprietary skilled nursing centers and assisted living communities. Rather than having many of our individual members who have contacted us with comments and to more efficiently provide NQF with feedback, we have received from our various members, we have summarized their comments here. Thank you again for the opportunity to provide these comments. Please contact me at DGifford@ahca.org with any questions or for additional information. There is no reference or link to the measure specifications for this proposed measure about pain (MUC15-1131) but we assume it is the measure currently in use by CMS and publicly reported on Nursing Home Compare and if so, support its inclusion. (Submitted by: American Health Care Association )
  
  
• AMDA is pleased to have the opportunity to comment on the List of Measures under Consideration for December 1, 2015”. AMDA - The Society for Post-Acute and Long-Term Care Medicine (AMDA) represents over 5,500 physician medical directors, attending physicians, and other practitioners who work in post-acute and long-term care settings. Thank you again for the opportunity to provide these comments. Please contact me at Abardakh@amda.com with any questions or for additional information. There is no reference or link to the measure specifications for this proposed measure about pain measure (MUC15-1131) but we assume it is the measure currently in use by CMS and publicly reported on Nursing Home Compare and if so, support its inclusion. (Submitted by: AMDA-The Society for Post-Acute and Long-Term Care Medicine)
  
  

• Inclusion of these measures will build on existing immunization-related measures currently used in the Physician Quality Reporting System to help ensure that Medicare beneficiaries benefit from these important vaccinations. (Submitted by: Trust for America's Health)
  
  
• AHCA is pleased to have the opportunity to comment on the List of Measures under Consideration for December 1, 2015”. The American Health Care Association (AHCA) represents more than 12,000 non-profit and proprietary skilled nursing centers and assisted living communities. Rather than having many of our individual members who have contacted us with comments and to more efficiently provide NQF with feedback, we have received from our various members, we have summarized their comments here. Thank you again for the opportunity to provide these comments. Please contact me at DGifford@ahca.org with any questions or for additional information. There is no reference or link to the measure specifications for this proposed measure about influenza vaccine use (#1132) but we assume it is the NQF endorsed measure currently in use by CMS and publicly reported on Nursing Home Compare and if so, support its inclusion. (Submitted by: American Health Care Association )
  
  
• ARN opposes the Percentage of Residents or Patients Who Were Assessed and Appropriately Given the Seasonal Influenza Vaccine measure. While ARN is committed to ensuring that our patients receive the highest quality health care possible, we are concerned that overutilization of quality measures – while well intentioned – may have the unintended consequence of impeding health care quality. The purpose of rehabilitative care is to promote functional recovery and achievement of goals by patients so that they are able to function to their maximum potential in the least restrictive environment. The proposed measure is not related to the specific rehabilitative care provided to the patient and/or resident of a SNF and thus should not be included in the SNF Quality Reporting Program (QRP). Numerous opportunities exist in other health care settings to receive immunizations, prior to an individual’s admission to the SNF. (Submitted by: Association of Rehabilitation Nurses)
  
  
• AMDA is pleased to have the opportunity to comment on the List of Measures under Consideration for December 1, 2015”. AMDA - The Society for Post-Acute and Long-Term Care Medicine (AMDA) represents over 5,500 physician medical directors, attending physicians, and other practitioners who work in post-acute and long-term care settings. Thank you again for the opportunity to provide these comments. Please contact me at abardakh@amda.com with any questions or for additional information. There is no reference or link to the measure specifications for this proposed measure about influenza vaccine use (#1132) but we assume it is the NQF endorsed measure currently in use by CMS and publicly reported on Nursing Home Compare and if so, support its inclusion. (Submitted by: AMDA-The Society for Post-Acute and Long-Term Care Medicine)
  
  

• AHCA is pleased to have the opportunity to comment on the List of Measures under Consideration for December 1, 2015”. The American Health Care Association (AHCA) represents more than 12,000 non-profit and proprietary skilled nursing centers and assisted living communities. Rather than having many of our individual members who have contacted us with comments and to more efficiently provide NQF with feedback, we have received from our various members, we have summarized their comments here. Thank you again for the opportunity to provide these comments. Please contact me at DGifford@ahca.org with any questions or for additional information. We have two primary concerns with this measure: 1. There is no reference or link to the measure specifications for this proposed measure about antipsychotic use (#1133) but we assume it is the measure currently in use by CMS and publicly reported on Nursing Home Compare. We believe for the NQF MAP process to adequately assess any measure for voting, the full specifications need to be made available. 2. The measure does not exclude individuals with bipolar disorder, a FDA approved diagnosis for antipsychotics. The rationale for use of this measure is the black box warning for the off-label use in elderly, particularly dementia. However, the measure does not exclude bipolar disorder. A coalition of long term providers and health care professionals and physician groups have all signed a letter supporting the addition of bipolar as an exclusion for this measure. Therefore, according to NQF MAP criteria, this measure should at best receive a vote of “support with conditions” that it must undergo NQF endorsement and that bipolar disorder be added as an exclusion. (Submitted by: American Health Care Association )
  
  
• AMDA is pleased to have the opportunity to comment on the List of Measures under Consideration for December 1, 2015”. AMDA - The Society for Post-Acute and Long-Term Care Medicine (AMDA) represents over 5,500 physician medical directors, attending physicians, and other practitioners who work in post-acute and long-term care settings. Thank you again for the opportunity to provide these comments. Please contact me at abardakh@amda.com with any questions or for additional information. We have two primary concerns with this measure: 1. There is no reference or link to the measure specifications for this proposed measure about antipsychotic use (#1133) but we assume it is the measure currently in use by CMS and publicly reported on Nursing Home Compare. We believe for the NQF MAP process to adequately assess any measure for voting, the full specifications need to be made available. 2. The measure does not exclude individuals with bipolar disorder, a FDA approved diagnosis for antipsychotics. The rationale for use of this measure is the black box warning for the off-label use in elderly, particularly dementia. However, the measure does not exclude bipolar disorder. A coalition of long term providers and health care professionals and physician groups have all signed a letter supporting the addition of bipolar as an exclusion for this measure. Therefore, according to NQF MAP criteria, this measure should at best receive a vote of “support with conditions” that it must undergo NQF endorsement and that bipolar disorder be added as an exclusion. (Submitted by: AMDA-The Society for Post-Acute and Long-Term Care Medicine)
  
  
• This measure is currently being field tested, has never been submitted to NQF and thus is not NQF-endorsed (Submitted by: AdvaMed)
  
  

• • MUC15-1134: Medicare Spending Per Beneficiary Post Acute Care (PAC) Home Health Quality Reporting Program (Required under the IMPACT Act)- An accurate determination of Medicare spending should take into account Medicare Part A, B and D expenditures. Average episode spending is in the numerator and the weighted median MSPB_PAC amount across all the home health agencies nationally is in the denominator. It is important to take into account the considerable variation in number of visits and number of episodes from state to state. A median across all the home health agencies nationally may not accurately reflect the cost of care in certain states. This measure does not meet NQF MAP criteria for endorsement and requires continued development. (Submitted by: LeadingAge)
  
  
• As an IMPACT Act measure that is stated to be "standardized and interoperable", we are highly concerned that this measure may allow for comparison between PAC venues without making adjustments for patient severity or differences inherent within the various payment systems. For example, some of the PAC venues are paid on a per diem basis, while other PAC venues have payments that are episodic. How are these differences accounted for in the measurement calculations. Additionally, each PAC venue has different regulatory requirements impacting their ability to control costs, which are not adequately accounted for in this measure. We also note that at this time some of the PAC venues are in the midst of drastic changes to their payment systems. Will the currently developed calculations of this measure be impacted by changes to underlying payment systems? And finally, the IMPACT Act also requests a report from MedPAC on the feasibility of a unified PAC payment system. This work is currently underway, and may provide further clarity to differences in payments between PAC sites. (Submitted by: UDSMR)
  
  
• The Alliance appreciates the importance of developing measures to better understand Medicare cost and resource use. However, when viewed in isolation, the Alliance is concerned that cost information alone is a confusing measure because it does not necessarily correlate with quality of care. The Alliance believes that spending alone is not an indicator of quality, nor is it an indicator of efficiency. The measure will be most useful when paired with quality outcome measures. If outcome measures are not linked to this cost measure, there may be an incentive for providers not to refer patients for reasonable and necessary services, including post-acute care services that can be used to reduce rehospitalization rates and improve patient experience. Ensuring that adequate quality outcome measures are coupled with measures of the cost of care is critical to discouraging underuse. Moreover, there is a need for the development of measures of patient access to care. Reforms that are aimed at improving efficient, cost-effective delivery of care are needed, as are measures that will help to encourage efficiency. However, patient access should not be compromised as a means to lower cost. Unfortunately, this measure alone cannot be used to assess whether patients have access to quality care. The Alliance is not aware of any measures at present that would address access to care. We would support development of such measures. Finally, the Alliance urges testing of this measure and reconsideration before it is finalized. Testing and validation should be no less than six months with an opportunity to modify the measure prior to finalizing it. A similar approach was used for many of the OASIS-based measures that CMS uses for home health agencies. (Submitted by: Alliance for Home Health Quality and Innovation)
  
  
• The amount of Medicare spending per beneficiary (MSPB) does not have a direct correlation with the quality of care provided by a home health agency. The IMPACT Act requires the establishment of a measure on resource use. Although MSPB may provide some useful information on resource utilization of Medicare providers, it should not be used as the sole resource use measure or to draw assumptions on quality of care on its own. In addition, using MSPB as a quality measure could negatively impact access to home health care for certain high risk patient populations as providers are incented to avoid “high cost” patients in order to achieve favorable scores on the measure.. Further, with HHPPS, efficient use of resources in home health is better measured by visits per episode for like patients, since the amount of Medicare spending for the 60 day episode of care would be the same regardless of the resources used beyond four visit LUPAs. The only variance would be for any Medicare spending in the 30 day post discharge period for this measure. Finally, we strongly believe that NQF should emphasize that MSPB, to the extent that it has any ultimate relevance in patient assessment or payment models, should never be used as the sole measure of value or quality. (Submitted by: National Association for Home Care & Hospice (NAHC))
  
  
• This measure is currently in the early stage of development, has never been submitted to NQF and thus is not NQF-endorsed (Submitted by: AdvaMed)
  
  

• • MUC15-207: Falls risk composite process measure- Completion of a falls assessment, the incorporation of the results of the falls assessment in the care plan, and the implementation of the care plan are noted in the numerator. Most of this information can be obtained from OASIS (M1910) that refers to a multi-factor falls risk assessment that includes components of other measures, such as transferring, multiple medications, mental impairment and falls history. The denominator is the total number of patients discharged by the home health agencies. The implementation of the care plan noted in the numerator would have to be determined by multi-factors in the patient’s medical record. It is important to note that this measure assesses the home health agencies ability to document in the care plan the falls risk. The measure falls short of determining if the home health agency is actually implementing an effective plan of care to reduce the risk of falls. We believe this measure does not meet NQF National Quality Strategies Priorities: Subcriterion 2.1 Safer care, and should be a measure that requires continued development. It is a process measure that does not reflect outcomes. (Submitted by: LeadingAge)
  
  
• Based on the limited information regarding the specifications for the measure, NAHC generally supports the measure. (Submitted by: National Association for Home Care & Hospice (NAHC))
  
  
• The Alliance supports this measure, but is concerned about the overlap of this measure with other existing measures. Currently, there are several home health measures relating to falls risk. A measure on multi-risk fall risk assessment conducted for all patients who can ambulate is in Home Health Compare. Two other measures relating to falls (relating to having the care plan reflect the falls risk assessment, and implementation of the care plan) also appear in the home health CASPER reports. This measure appears to combine aspects of the three existing measures into a single composite process measure. As such, it will capture a subset of the patients included in the existing falls-related process measures. The Alliance is concerned about the overlap in the measures. The question is whether some measures will be retired as appropriate. The growing number of overlapping measures has the potential to be confusing for the general public and even those who are using the measures as payers and policy-makers. The ideal is a streamlined, meaningful measure set and the Alliance encourages CMS to make this a goal. Finally, the Alliance urges testing of this measure and reconsideration before it is finalized. Testing and validation should be no less than six months with an opportunity to modify the measure prior to finalizing it. A similar approach was used for many of the OASIS-based measures that CMS uses for home health agencies. (Submitted by: Alliance for Home Health Quality and Innovation)
  
  
• This measure is in early stage of development, has never been submitted to NQF and thus is not NQF-endorsed (Submitted by: AdvaMed)
  
  
• We support the development of a falls risk composite measure that includes assessment, care plan development, and implementation. We believe this type of measure is appropriate for home health and across settings. (Submitted by: Association of Home & Hospice Care of NC/SC Home Care & Hospice Association)
  
  

• • MUC15- 227: Hospice Visits When Death Is Imminent- This measure under consideration does not take into account a number of factors that will impact outcomes. First, rural and frontier hospice agencies serve a large geographic area resulting in long travel times for hospice staff. Many times visits are supplemented by telephone calls from the hospice staff. Second, there is no differentiation or note in the exclusion that addresses if the hospice patient is a resident of a nursing home. We believe that these two factors should be included in the measure, and for that reason we recommend this measure requires continued development. (Submitted by: LeadingAge)
  
  
• Hospices do not currently collect and report all the visit data that is included in this measure and only some of it is captured on the Medicare hospice claims. The visits for nurses (registered nurse, licensed professional nurse or nurse practitioner if acting in the role of a nurse), hospice aides, physicians (or nurse practitioner or physician assistant if acting as the attending physician), therapists (physical therapist, occupational therapist or speech language therapist), and medical social workers are reported on hospice claims but chaplains or spiritual counselors and volunteers are not. This creates a burden for hospices as they do not currently have the infrastructure to capture and report this data. Regardless, any measure utilized in the HQRP must be properly vetted to ensure that it is, in fact, related to quality of hospice care to reduce the incidence of inappropriate conclusions that might be drawn from the measure. We believe visit measures are practice indicators that do not take into account the condition and other characteristics of the population served by a hospice nor the quality of care delivered during the visit. A visit measure should not be used in the HQRP unless its direct relationship to hospice quality has been proven, and that connection can be readily drawn by the public. (Submitted by: National Association for Home Care & Hospice (NAHC))
  
  
• We support the inclusion of all staff included in the patient's plan of care as part of the calculation of the quality measure. (Submitted by: Association of Home & Hospice Care of NC/SC Home Care & Hospice Association)
  
  
• This measure is currently in the early stage of development, has never been submitted to NQF and thus is not NQF-endorsed (Submitted by: AdvaMed)
  
  
• Hospices do not currently collect and report all the visit data that is included in this measure. Some is captured on the Medicare hospice claims. Specifically, visits for chaplains or spiritual counselors and volunteers are not currently collected and reported by hospices. Moreover, a visit measure does just that – measure the quantity of visits. If this measure were implemented, it would create a burden for those programs not having the capability/infrastructure to collect and report this visit data. For instance, in many hospices, volunteers document on paper and their visits are not captured electronically for collection and reporting. The same is true of chaplain visits in some hospices. Regardless, any claims-based or visit measure utilized in the HQRP must be properly vetted to ensure that it is, in fact, related to quality of hospice care. We believe claims-based measures are practice indicators that do not take into account the condition and other characteristics of the population served by a hospice. A claims-based measure should not be used in the HQRP unless its direct relationship to hospice quality has been proven, and that connection can be readily drawn by the public. (Submitted by: The Home Care & Hospice Association of NJ)
  
  
• Hospices do not currently collect and report all the visit data that is included in this measure. Some is captured on the Medicare hospice claims. Specifically, visits for chaplains or spiritual counselors and volunteers are not currently collected and reported by hospices. Moreover, a visit measure does just that – measure the quantity of visits. If this measure were implemented, it would create a burden for those programs not having the capability/infrastructure to collect and report this visit data. For instance, in many hospices, volunteers document on paper and their visits are not captured electronically for collection and reporting. The same is true of chaplain visits in some hospices. Regardless, any claims-based or visit measure utilized in the HQRP must be properly vetted to ensure that it is, in fact, related to quality of hospice care. We believe claims-based measures are practice indicators that do not take into account the condition and other characteristics of the population served by a hospice. A claims-based measure should not be used in the HQRP unless its direct relationship to hospice quality has been proven, and that connection can be readily drawn by the public. (Submitted by: Ohio Council for Home Care & Hospice)
  
  

• • MUC15-231: Hospice and Palliative Care Composite Process Measure- This is a composite measure based on select measures from 7 NQF endorsed measures: NQF #1641(treatment preferences), NQF #1647 (beliefs/values addressed), NQF #1634, NQF #1637 (Pain screening), NQF #1639 (Dyspnea screening), NQF #1638 Dyspnea Treatment), NQF #1617 (Patients Treated with an Opioid who are Given a Bowel Regiment). All of these components of this composite measure are already included in the Hospice Item set. We support this composite process measure. (Submitted by: LeadingAge)
  
  
• We support the composite quality measure from the existing HIS quality measures. (Submitted by: Association of Home & Hospice Care of NC/SC Home Care & Hospice Association)
  
  
• This measure uses existing and readily available data and is more closely aligned with hospice processes directly impacting quality of care. (Submitted by: National Association for Home Care & Hospice (NAHC))
  
  
• This measure is currently in the early stage of development, has never been submitted to NQF and thus is not NQF-endorsed (Submitted by: AdvaMed)
  
  
• This measure uses existing and readily available data and more closely aligned with hospice processes directly impacting quality of care. (Submitted by: The Home Care & Hospice Association of New Jersey)
  
  
• This measure uses existing and readily available data and more closely aligned with hospice processes directly impacting quality of care. (Submitted by: Ohio Council for Home Care & Hospice)
  
  

• NAHC is concerned that the measure could unfairly hold HHAs accountable for readmissions for which they have little or no control in preventing. Based on the specifications known for this measure NAHC offers the following recommendations: • Limit the number of PPR conditions to those conditions where evidence supports that a readmission could be prevented for 30 days by a home health treatment plan. • Crosswalk the selected conditions with ICD-10 diagnoses codes and ensure the diagnoses are tested and validated for inclusion in the measure prior to implementation. • Ensure the addition of the PPR measure does not conflict with the current home health re-hospitalization and acute care hospitalization measures. • Include in the risk adjustment model for home heath prior to implementation: o Socioeconomic status o Caregiver support o Prior PAC setting stay and the length of stay (Submitted by: National Association for Home Care & Hospice (NAHC))
  
  
• APTA supports the goal of improving the quality of health care. Physical therapists are committed to providing high-quality, timely care and to the promotion of evidence-based and patient-centered practice. Furthermore, APTA feels that it is essential that we move towards a common set of quality measures across the across the continuum of care. APTA supports the implementation of readmissions measures across the care settings, as approximately 20% of all Medicare patients are readmitted within 30 days of an acute care discharge and readmissions account for an estimated $17 billion in health care spending. APTA is pleased to see that the draft specifications for these measures align with existing methodologies of other readmissions measures. We believe that potentially preventable readmissions measures will focus providers on those patients who are expected to have successful transitions to the community follow in discharge from the respective post-acute care settings. However, the APTA does have some concerns regarding the proposed measure methodology. These concerns are discussed below. APTA believes that a patient’s level of function does impact the potential for readmissions. Recent evidence indicates that patient function is associated with increased risk of 30-day all-cause hospital readmissions and may be an important factor in preventing readmissions for Medicare seniors that is not currently accounted for in measure methodologies1. APTA is pleased to see “activity of daily living” scores included in the risk adjustment methodology, however, we note that this is only included for those patients in the home health setting. We would encourage measure developers to include this as a risk adjustment variable in all readmissions measures. APTA appreciates that CMS has strict deadlines for the implementation of measures under the IMPACT act, however, as these measures will be new to the respective post-acute care settings we encourage that settings have the ability to review this data as early as possible in order to understand the data and, more importantly, so that the respective setting have time to implement strategies to decrease readmissions where necessary. As many of these settings do not always receive feedback on the readmissions of their patients post-discharge, this data will be new to many facilities. Additionally, skilled nursing facilities and inpatient rehabilitation facilities will be trying to manage two measures, one within stay, and one post-discharge. APTA recognizes that the overall goal of IMPACT is for PAC providers (HH, IRF, SNF and LTCH) to collect and report standardized and interoperable patient assessment data, quality and resource use measures. We acknowledge that during the initial IMPACT implementation years that there will be a transition period which will include the addition of new measures into all of the post-acute care settings. We believe that achieving a standardized and interoperable patient assessment data set and stable quality measures as quickly as possible will allow for better cross-setting comparisons as well as the evolution of better quality measures with uniform risk standardization, thus achieving the true aim of IMPACT. 1 Greysen SR, Cenzer IS, Auerbach AD, Covinsky KE. Functional Impairment and Hospital Readmission in Medicare Seniors. JAMA Intern Med. 2015;175(4):559-565. (Submitted by: American Physical Therapy Association)
  
  
• • MUC15-234: Potentially Preventable 30- Day Post-Discharge Readmission Measure for Home Health Quality Reporting Program (Required under the IMPACT Act). There are a couple of concerns with this measure. First, there is not enough information on what CMS would determine to be a potentially preventable readmission. MedPAC has estimated that 76 percent of 30-day readmissions for Medicare beneficiaries overall were due to five potentially preventable conditions (heart failure, electrolyte imbalance, respiratory infection, sepsis, and urinary tract infection (MedPAC 2007). In the exclusions, treatment of cancer and primary psychiatric diseases are listed. The numerator includes risk adjustment for patient characteristics. There is not enough information available to assess if the exceptions, the risk adjustment and the determination of what disease or co-morbidities should be considered potentially preventable. The uncertainty of defining a potentially preventable readmission leads us to recommend that this measure require continued development. (Submitted by: LeadingAge)
  
  
• We would like to note the following concerns regarding the development of these measures: 1. Measure development for the IMPACT Act appears to create 4 separate measures that are site specific with their own set of criteria and risk adjustment factors. The IMPACT Act aims to create quality measurement within Post-Acute Care (PAC) that is "standardized and interoperable". By developing measures which differ in their calculation, such as site-specific inclusion/exclusion criteria or risk-adjustment factors, PAC sites will be subjected to comparisons of quality that do not differ based upon the quality of care provided, but rather the differences inherent in the measure calculations. We strongly recommend that CMS and the measure developers produce a measure that is calculated in a "standardized and interoperable" manner with inclusion/exclusion criteria and risk-adjustment factors the are applied consistently across all PAC sites. 2. The measures developed for the IMPACT Act hold PAC providers responsible for circumstances that occur outside of their control. While PAC providers are responsible for providing care that allows patients safe transitions to their next setting or home, holding these providers accountable fora time period in which they are not furnishing care should not differentiate the quality of care they actually provided. In Appendix A, Table A1 lists various conditions that define potentially preventable hospital readmission for 30-days post-PAC discharge. In this list, conditions such as Asthma, Congestive Heart Failure, and Dehydration are listed as a reason for a potentially preventable readmission. So if the patient fails to care for themselves post-discharge (against the advice and discharge instructions provided by a PAC provider) and requires readmission to Acute Care for one of these conditions, the PAC provider is penalized even though the patient is no longer being cared for by the PAC provider. We strongly recommend that CMS and the measure developers produce a measure that represents the quality of care provided by PAC providers while the patient is in their care. (Submitted by: UDSMR)
  
  
• This measure will encourage post-acute care providers to optimize care for patients before discharge. (Submitted by: Smith & Nephew )
  
  
• The Alliance is concerned about the evidence used to support this measure. The evidence regarding post-acute care potentially preventable readmissions is limited (see p. 5 of the measure specifications document that CMS shared on its website). The diagnosis codes identified as potentially preventable in the measure specifications are based on the ambulatory care sensitive conditions that the Agency for Healthcare Research and Quality (AHRQ) has developed. AHRQ’s list identifies conditions for which hospitalizations should be preventable if such conditions are well managed in ambulatory care settings. However, the list is not specifically targeted at conditions for which readmissions should be preventable. In other words, it is not clear whether after a hospitalization such conditions are ones for which readmissions should be considered preventable. Hospitalization significantly changes the condition of a patient and may in itself make the patient more likely to experience health risks that make the patient more likely to be readmitted. We are concerned that there is little evidence regarding the ability to prevent a subsequent post-acute care readmission for the ambulatory care sensitive conditions that are the basis of the list of diagnosis codes in the measure specifications. The Alliance recommends close analysis of the evidence base for this measure, and that modifications be made accordingly. In addition, patients that have used other post-acute care settings before using home health care tend to have higher severity and are more likely to be at risk for readmission. The measure as described in the specifications would not distinguish among patients that have been to only one post-acute care setting (home health) or three or more different post-acute care settings. The Alliance recommends considering this factor in the risk adjustment for the measure. In addition, the Alliance urges testing of this measure and reconsideration before it is finalized. Both this measure and the discharge to community measure are based on risk-adjusted estimates. Testing and validating this measure will be critical. Testing and validation should be no less than six months with an opportunity to modify the measure prior to finalizing it. A similar approach was used for many of the OASIS-based measures that CMS uses for home health agencies. If finalized, the potentially preventable readmission measure will be the third measure for home health care that involves readmissions. There is already a measure for acute care hospitalization (during the 60-day home health episode), as well as a measure for readmissions from home health care within 30 days of discharge from the acute care hospital. There is overlap among these multiple measures that each capture readmissions. The Alliance recommends that CMS provide context for how it anticipates using or applying each measure. Increasingly, there are different applications for measures and it is unclear as yet how CMS plans to use each one. The Alliance’s full comments on this measure’s specifications can be found at: http://ahhqi.org/images/uploads/Alliance_Comments_on_Potentially_Preventable_Readmissions_111615.pdf (Submitted by: Alliance for Home Health Quality and Innovation)
  
  
• Our home health agencies have long strived to keep patients independent at home and prevent costly, unnecessary hospitalizations. We support the goal of the IMPACT act to align quality measures across post-acute providers and to promote patient-centeredness in quality efforts. We support the designation of a single hospital admission/readmission measure as the target for home health agencies that would be available on their CASPER reports, publically reported on Home Health Compare, and considered for Medicare value-based purchasing initiatives. The availability of multiple quality measures derived from OASIS and Medicare claims is confusing for agencies as the calculations of these measures are complex and it is time consuming for agencies to understand and to be able to educate their staff and referral sources on them. For the draft measure developed by RTI we support appropriate risk-adjustment of this quality measure to take into account the unique characteristics of the agency’s patient population. We support the inclusion prior emergency department use in the calculation specifically for home health agencies as we believe that it is an appropriate indication of patient stability. We support the further study of risk adjusters specifically dual eligibility as we believe that it is a good proxy for other social demographic determinants. We support the inclusion of this measure in the CASPER reports to provide agencies with data on their performance. (Submitted by: Association of Home & Hospice Care of NC/SC Home Care & Hospice Association)
  
  
• We recommend that CMS ensure the data collection time frame is the same for all PAC settings. Currently, the time frame for the initial data collection for the project varies from one year (SNFs), two years (IRFs and LTCHs), and three years (HHAs), we urge the Agency to implement one standard reporting time frame across PAC providers. While reporting may be based on either one year of data, two years of data, etc., we strongly recommend CMS and its subcontractors address the inconsistent reporting periods. (Submitted by: Association of Rehabilitation Nurses)
  
  
• This set of potentially preventable readmission (PPR) measures for post-acute care (PAC) estimates the risk-standardized rate of unplanned, potentially preventable readmissions for patients (Medicare fee-for-service [FFS] beneficiaries) who receive services in one of the following post-acute care provider types: skilled nursing facilities (SNFs), inpatient rehabilitation facilities (IRFs), long-term care hospitals (LTCH), and home health agencies (HHA). These outcome measures reflect readmission rates for patients who are readmitted to a short-stay acute-care hospital or an LTCH with a principal diagnosis considered to be unplanned and potentially preventable. Six PPR PAC measures are being developed: • Four of these measures assess PPR within a 30-day window following discharge from PAC—one measure for each PAC setting (i.e., SNF, IRF, LTCH, and HH)—and are being developed to meet the requirements of the IMPACT Act. • An additional SNF measure (SNF PPR), which is being developed to meet the PAMA requirements, assesses PPR during the 30-day period following a hospital discharge to a SNF setting. • An additional IRF measure assesses PPR during the IRF stay (referred to as the within-stay window) which is being developed for use in the IRF Quality Reporting Program. Of note with respect to potentially preventable readmissions and a patient’s occupational therapy needs, several recent studies consider whether returning to the community from a recent hospitalization with unmet activities of daily living (ADL) need was associated with probability of readmission. The findings from these studies indicate that this indeed may be a considerable risk factor. The studies reveal that any older patients are discharged from the hospital with ADL disability. Those who report unmet need for new ADL disabilities after they return home from the hospital are particularly vulnerable to readmission. This area is not typically addressed in a thorough manner through current discharge practices. This needs to change. Patients' functional needs after discharge should be carefully evaluated and addressed. Factors such as enabling self-management and ensuring appropriate medication management and ADLs, such as cooking and eating are addressed, can have a direct effect on readmissions The profession of occupational therapy is built on delivering patient-centered care, seeking to keep the patient at the highest functional level in the least restrictive setting and to reduce caregiver burden and health care system resource utilization. Occupational therapy directly addresses the enablement of successful performance of ADLs. This focus, experience and research base in occupational therapy must be fully tapped to address this component of readmission prevention. A further issue is that Self-management is a key element in successful care, and occupational therapists are experts in motivation, task analysis, and psychosocial contexts, which all contribute to enabling positive outcomes. In order to successfully re-establish or establish new routines and habits to meet health care needs, such as medication management, proper sleep hygiene, and following other medical directives, is within the scope and proven effectiveness of occupational therapy. AOTA would encourage CMS to examine more fully, perhaps through pilot testing, the value of occupational therapy evaluation and intervention participation as part of discharge planning. This could identify more clearly specific ADL limitations prior to and after discharge and assure proper interventions are provided that address fully and completely the range of ADL and other activity restrictions and capacities to enable optimum recovery from the condition as well as optimum participation of the client in their own care. AOTA has reviewed the readmissions measures released for comment and makes the following general comments: • The PPR measures Numerator and Denominator definition language are consistent from previous materials. • The Readmissions Measures are based on two years of claims data. Use of multiple years is acceptable in the measure review process. • The exclusion criteria for the various measures are associated with the Yale/New Haven readmission measures. Co-developed by the Centers for Medicare & Medicaid Services (CMS) and researchers at Yale University, estimates the risk-standardized rate of unplanned, all-cause readmissions to a hospital for any eligible condition within 30 days of hospital discharge for patients aged 18 and older. The CMS/Yale measure is specified for evaluating hospital performance. AOTA would prefer for the exclusion criteria to be specified for evaluating performance in post-acute care settings. (Submitted by: American Occupational Therapy Association)
  
  
• At least one NRHI member has expressed concerns regarding this measure. Specifically, clinical staff at HealthInsight have concerns with inclusion of this measure due to the validity of and complicated methodologies used for "risk adjustment for patient characteristics and a statistical estimate of the facility effect beyond patient mix." This methodology is frequently cited as the basis for determining potentially avoidable hospital facility admissions. This methodology is not well-established. (Submitted by: Network for Regional Healthcare Improvement)
  
  
• At least one NRHI member has expressed concerns regarding this measure. Specifically, clinical staff at HealthInsight have concerns with inclusion of this measure due to the validity of and complicated methodologies used for "risk adjustment for patient characteristics and a statistical estimate of the facility effect beyond patient mix." This methodology is frequently cited as the basis for determining potentially avoidable hospital facility admissions. This methodology is not well-established. (Submitted by: Network for Regional Healthcare Improvement)
  
  

• NAHC supports the measure for improvement in dyspnea targeted for patients with a primary diagnosis of CHF, COPD and/or Asthma. (Submitted by: National Association for Home Care & Hospice (NAHC))
  
  
• • MUC15-235: Improvement in Dyspnea in Patients with a Primary Diagnosis of Congestive Heart Failure, Chronic Obstructive Pulmonary Disease and/or Asthma – In the 2015 National Impact Assessment of the Centers for Medicare & Medicaid Services (CMS) Quality Measures Report, it is noted that Improvement in Dyspnea rates decreased, so there is room for improvement in this area. We believe that there needs to be more detailed assessment of why there is less of an improvement in dyspnea. The exclusions for this measure and the specifications in the numerator and denominator do not address co-morbidities and if the patient smokes. The omission of key factors that would impact dyspnea leads us to recommend that this measure require continued development. (Submitted by: LeadingAge)
  
  
• The Alliance is supportive of this measure, but is concerned about overlap with existing measures and the absence of measures involving stabilization of function. At present, there is a home health compare measure for improvement in dyspnea for all home health patients. This will provide more specific measurement for CHF, asthma and COPD patients. This measure will capture a subset of the patients included in the existing improvement in dyspnea measure. The Alliance is concerned about the overlap in the measures and these measures raise the issue of whether some measures will be retired where new measures are becoming more focused on subsets of patients. The growing number of overlapping measures has the potential to be confusing for the general public and even those who are using the measures as payers and policy-makers. The ideal is a streamlined, meaningful measure set and the Alliance encourages CMS to make this a goal. In addition, the Alliance continues to be concerned about the exclusive focus of measurement on improvement, to the exclusion of measurement of stabilization. Many home health patients have multiple chronic conditions and two or more ADL limitations. For some, stabilization of function is a legitimate goal of treatment. The Alliance believes this is an important measure gap that must be addressed in the future. Finally, the Alliance urges testing of this measure and reconsideration before it is finalized. Testing and validation should be no less than six months with an opportunity to modify the measure prior to finalizing it. A similar approach was used for many of the OASIS-based measures that CMS uses for home health agencies. (Submitted by: Alliance for Home Health Quality and Innovation)
  
  
• This measure is currently in the early stage of development, has never been submitted to NQF and thus is not NQF-endorsed (Submitted by: AdvaMed)
  
  
• We support the inclusion of a quality measure that specifically addresses dyspnea for these diagnoses. We believe that a stabilization measure is a more appropriate target for the quality reporting program. Not all patients have the ability to improve. We believe that stabilization is in line with the goal of home health to maintain or improve a patient's ability. (Submitted by: Association of Home & Hospice Care of NC/SC Home Care & Hospice Association)
  
  

• This measure is currently in early development and measure specifications are in preliminary stages of clarification. The proposed application could definitely create inaccuracy in measurement of patient condition when used in other PAC settings. Terms have not been defined within the exclusions (i.e. "incomplete stays"). No testing of the measure in non-IRF settings has been done. These could create inaccurate data capture throughout PAC venues. The Technical Expert Panel indicated that this measure needs extensive work to assure accuracy and to meet the purpose for which it was intended. Because there is potential for accuracy this measure should continue to be developed. (Submitted by: Geneis Health Care)
  
  
• AHCA is pleased to have the opportunity to comment on the List of Measures under Consideration for December 1, 2015”. The American Health Care Association (AHCA) represents more than 12,000 non-profit and proprietary skilled nursing centers and assisted living communities. Rather than having many of our individual members who have contacted us with comments and to more efficiently provide NQF with feedback, we have received from our various members, we have summarized their comments here. Thank you again for the opportunity to provide these comments. Please contact me at DGifford@ahca.org with any questions or for additional information. 1. While this measure has been endorsed by NQF, it was developed, tested and validated in the IRF setting for IRF patients. It was not developed for SNF setting. We do not think it is appropriate to take a measure from one setting and assume it is reliable and valid in another setting. If this logic is to be followed through, there are several other NQF endorsed functional measure that rely on the CARE tool item set, which is what is used for this measure. We believe, the MAP process must review and consider all NQF endorsed measures that are similar. In particular, NQF measure 2613 and 2286 which rely on the CARE tool items. 2. In addition, the proposed measure relies on CARE tool items that were not added to section GG in the IRF, SNF or LTCH setting. As such, the specifications of this measure do not match the available data in the PAC assessment tools for SNF, IRF, and LTCH setting. 3. AHCA recommends that, before a measure developed in one setting is applied to another, that cross-setting measure testing be performed. Such testing is critically important because of the clinical sorting that goes on between the settings, with benefit rules, Medicare enrollment requirements, and the general suitability of the settings resulting in certain kinds of patients tending to go to SNFs, and other patients tending to go to IRFs. For example, the IRF benefit rule that patients can receive three or more hours of therapy each day means that certain groups of patients who would have difficulty walking will tend to sort into the SNF setting rather than the IRF setting. This phenomena will also occur on self-care domains. For example, differences in the proportion and types of stroke patients between SNF and IRF make application of an IRF measure inaccurate in the SNF setting. Because CMS has not tested whether the distribution of risk adjusted measure scores in the SNF setting are the same or widely different from those in the IRF setting, we are left with a serious concern that the measure’s distribution will be shifted up in the IRF setting and down in the SNF setting, without anything to do with the quality of care being provided in each setting. Last, we recognize that it is entirely possible that the measures perform adequately after cross-setting measure testing; however, if this testing is not performed before the measures are rolled out into live public use, CMS risks promoting a risk adjustment model that does not work as well for the SNFs compared to the IRF Quality Reporting Programs. 4. The specifications are based on ICD-9 but all providers as of October 2015 are required to use ICD-10 and no cross walk with ICD-10 is provided, yet the measure will be used during time periods when only ICD-10 data is available. Therefore, for these reasons, we believe this measure does not meet NQF MAP criteria for endorsement but rather should receive a vote of “encourage continued development” only after NQF MAP process has reviewed the other NQF endorsed functional measures to determine which the best to be recommended. (Submitted by: American Health Care Association )
  
  
• This measure was validated and tested for an in-patient rehab population. it cannot be assumed that these populations are alike, or that this is a valid measure for SNF settings. Goal setting, that is driven by patient-centered goals, is likely very different in the SNF from an IRF. This would also need to be risk adjusted for co-morbidities and pre-morbid function (Submitted by: leadingage)
  
  
• First, the IRF endorsement of this measure was conditional, requiring updates when data is available (after October 2016) to display the reliability and validity of the measure, among other things. We strongly urge the MAP PAC/LTC committee to consider whether or not to approve and/or implement a measure that has not been shown to meet NQF requirements for reliability and validity. Second, this measure should be identified as a competing measure with NQF #2613 - CARE: Improvement in Self Care, developed by AHCA and endorsed earlier this year by NQF. These measures are similar in that they utilize functional items creates as part of the CARE tool development project or PAC-PRD, but have differences in the functional items included, inclusion/exclusion criteria, and risk-adjustment methodologies. Third, a "best-in-class" determination related to the functional improvement or change in self-care is required and necessary. This measure has already been found to compete with a UDSMR developed measure (NQF #2286), and NQF staff chose not to make a "best-in-class" decision until such a time as additional measure testing was available. With the additional consideration for the AHCA measure that is already NQF endorsed noted previously, there exist 3 different ways for measuring the same functional construct. The IMPACT Act requirements indicate that measures for PAC settings should be "standardized and interoperable", yet implementation of the IMPACT Act is burdening PAC providers with the data collection of functional items for payment and similar but different functional items for quality. PAC providers should not bear the burden of data collection for measures that may not be identified as "best-in-class". (Submitted by: UDSMR)
  
  
• This measure is currently in the early stage of development, has never been submitted to NQF and thus is not NQF-endorsed (Submitted by: AdvaMed)
  
  
• This measure was developed for the IRF and as such may not be usable in the SNF without further development and testing. The admission criteria for a patient entering an IRF is different than a SNF, therefore IRF patients may be different than SNF patients. Therefore, measuring self-care could use different items for a SNF patient than measuring self-care for an IRF patient. This measure requires testing and validation before it could be used for SNF patients. (Submitted by: National Association for the Support of Long Term Care)
  
  
• ARN is supportive of the IRF functional measures; however, there are several key differences between the existing and new proposed function items that may result in variation in the resident assessment results including: (1) data collection and associated data collection instructions; (2) rating scales used to score a resident’s level of independence; and (3) item definitions. Utilizing multiple data collection systems and instruments as well as a different rating scale and item definitions is both cumbersome and burdensome. Additionally, functional measures are not a “one size fits all,” and measures should account for the benefits of the quality of life domain for patients with these extreme conditions. ARN encourages the Agency to revise the measures to reflect the following attributes: a low collection burden for providers and beneficiaries; comprehensibility for beneficiaries; a high level of significance to patients and providers; and data that is routinely captured. (Submitted by: Association of Rehabilitation Nurses)
  
  
• AMDA is pleased to have the opportunity to comment on the List of Measures under Consideration for December 1, 2015”. AMDA - The Society for Post-Acute and Long-Term Care Medicine (AMDA) represents over 5,500 physician medical directors, attending physicians, and other practitioners who work in post-acute and long-term care settings. Thank you again for the opportunity to provide these comments. Please contact me at abardakh@amda.com with any questions or for additional information. 1. While this measure has been endorsed by NQF, it was developed, tested and validated in the IRF setting for IRF patients. It was not developed for SNF setting. We do not think it is appropriate to take a measure from one setting and assume it is reliable and valid in another setting. If this logic is to be followed through, there are several other NQF endorsed functional measure that rely on the CARE tool item set, which is what is used for this measure. We believe, the MAP process must review and consider all NQF endorsed measures that are similar. In particular, NQF measure 2613 and 2686 which rely on the CARE tool items. 2. In addition, the proposed measure relies on CARE tool items that were not added to section GG in the IRF, SNF or LTCH setting. As such, the specifications of this measure does not match the available data in the PAC assessment tools for SNF, IRF, and LTCH setting. 3. AMDA recommends that, before a measure developed in one setting is applied to another, that cross-setting measure testing be performed. Such testing is critically important because of the clinical sorting that goes on between the settings, with benefit rules, Medicare enrollment requirements, and the general suitability of the settings resulting in certain kinds of patients tending to go to SNFs, and other patients tending to go to IRFs. For example, the IRF benefit rule that patients can receive three or more hours of therapy each day means that certain groups of patients who would have difficulty walking, and/or significant deconditioning, functional quadriplegia, dysphagia, or gastrostomy, will tend to sort into the SNF setting rather than the IRF setting. This phenomena will also occur on self-care domains. For example, differences in the proportion and types of stroke patients between SNF and IRF make application of an IRF measure inaccurate in the SNF setting. Because CMS has not tested whether the distribution of risk adjusted measure scores in the SNF setting are the same or widely different from those in the IRF setting, we are left with a serious concern that the measure’s distribution will be shifted up in the IRF setting and down in the SNF setting, without anything to do with the quality of care being provided in each setting. Last, we recognize that it is entirely possible that the measures perform adequately after cross-setting measure testing; however, if this testing is not performed before the measures are rolled out into live public use, CMS risks promoting a risk adjustment model that does not work as well for the SNFs compared to the IRF Quality Reporting Programs. 4. The specifications are based on ICD-9 but all providers as of October 2015 are required to use ICD-10 and no cross walk with ICD-10 is provided, yet the measure will be used during time periods when only ICD-10 data is available. Therefore, for these reasons, we believe this measure does not meet NQF MAP criteria for endorsement but rather should receive a vote of “encourage continued development” only after NQF MAP process has reviewed the other NQF endorsed functional measure to determine which the best to be recommended is. (Submitted by: AMDA-The Society for Post-Acute and Long-Term Care Medicine)
  
  
• the measure does not appear to consider functional outcomes to include the performance skills and cognitive capacities of self-care such as sequencing, problem-solving, temporal appropriateness (e.g., whether to dress for day or bed), memory, and activity planning. Nor does the measure consider performance of activities of daily living, including the broader instrumental activities of daily living (IADLs), which significantly impact a patient’s ability to function and live independently in the community. (Submitted by: American Occupational Therapy Association)
  
  
• At least one NRHI member has expressed concerns regarding this measure. Specifically, clinical staff at HealthInsight have concerns with designation of functional status scores such as "self-care score" and "mobility score" that will rate the performance of facilities. (Submitted by: Network for Regional Healthcare Improvement)
  
  

• As an IMPACT Act measure that is stated to be "standardized and interoperable", we are highly concerned that this measure may allow for comparison between PAC venues without making adjustments for patient severity or differences inherent within the various payment systems. For example, some of the PAC venues are paid on a per diem basis, while other PAC venues have payments that are episodic. How are these differences accounted for in the measurement calculations. Additionally, each PAC venue has different regulatory requirements impacting their ability to control costs, which are not adequately accounted for in this measure. We also note that at this time some of the PAC venues are in the midst of drastic changes to their payment systems. Will the currently developed calculations of this measure be impacted by changes to underlying payment systems? And finally, the IMPACT Act also requests a report from MedPAC on the feasibility of a unified PAC payment system. This work is currently underway, and may provide further clarity to differences in payments between PAC sites. (Submitted by: UDSMR)
  
  
• AMRPA has been analyzing resource use measures such as a Medicare Spending Per Beneficiary (MSPB) measure. We have done extensive work internally to prepare for this type of measure including development of the episode and risk-adjustment factors. We are concerned that an MSPB measure will not improve the quality of care patients receive and may actually have an adverse effect on outcomes. At this time an MSPB measure is not ready for inclusion in the Inpatient Rehabilitation Facility Quality Reporting Program (IRF QRP) as the development by the Centers for Medicare and Medicaid Services (CMS) is in its infancy, as noted by its Measure Under Consideration (MUC) development status of “Early Development.” The technical expert panel (TEP) convened by CMS to develop the measure met only recently and we assume its work is ongoing. Additionally, the details available through the MUC list are inadequate to provide relevant feedback that may assist in making appropriate modifications to the measure. Our comments highlight our concerns with the measure as outlined in the MUC List as well our recommendations for further development of the measure. We support the following elements of this measure as outlined in the MUC materials. First, we agree that there should be a separate MSPB measure for each post-acute care (PAC) provider type. It is critically important, given the difference among the PAC settings, payment systems, and the types of patients they treat, that like providers be compared (e.g., inpatient rehabilitation facilities (IRFs) compared to IRFs). Second, we also support the use of a national benchmark for payment purposes because of the regional variation in the number of IRFs. Third, we support the episode definition for the IRF MSPB measure as the IRF admission (index admission), the IRF stay, and 30 days following discharge from the IRF. As is done with the IPPS MSPB measure used in the Value-based Purchasing (VBP) program, the IRF MSPB measure should also be risk-adjusted. However, the severity should be based on the IRF prospective payment system (PPS) Case Mix Groups (CMGs), and not on inpatient prospective payment system (IPPS) diagnostic related groups (DRGs) because DRGs are not sensitive enough to capture the variation in IRF patients. Risk-adjustment should also factor in socioeconomic and sociodemographic characteristics as outlined in the National Quality Forum (NQF) report entitled Risk Adjustment for Socioeconomic Status or Other Sociodemographic Factors: Technical Report. These characteristics would include factors such as the availability of caregiver and community supports and race. We have also considered various exclusion and inclusion criteria for an IRF MSPB measure. We believe all patients and diagnoses should be included, but limit the measure to Medicare fee for service (FFS) patients thereby excluding patients under Medicaid and Medicare Advantage and those not enrolled in Medicare FFS for at least 90 days prior to the MSPB episode of care. All Part A and Part B services should also be included with the exception of planned readmissions and the associated costs, as well as other planned and/or routine treatments that are clinically unrelated to the IRF responsibility (e.g., chemotherapy or dialysis). Furthermore, we believe that any Medicare payments attributed to an IRF should remove the effects of or otherwise control for facility-specific payments paid under the IRF PPS (e.g., wage, low-income, rural, and teaching adjustments) as well as high-cost outlier payments received by the IRF. These payments are unrelated to any treatment services or associated services considered as cost/resource use. However, a measure must be a) risk-adjusted as noted above, and b) adjusted to address whether there are resources so that the patient can be discharged home (e.g., family, or other caregiver). Not acknowledging the availability of resources at home can lead to a bar to admissions (e.g., the patient would require discharge to other settings and therefore more payments). Claims data could be used but if a claim was denied by a Medicare contractor it should not be included in the MSPB calculation. In addition to claims data, we recommend the measure include items that examine functional and cognitive status and patient comorbidities. To ensure the highest reliability of this information, such data should be collected from PAC provider's assessment instruments and should rely on functional measurements with which the field has the most experience and for which the largest evidence base exists for its validity, reliability, and predictive value, such as the functional independence measure (FIM) scale. We understand that the IRF MSPB as outlined in the MUC List is limited to resource use and does not seem to incorporate measures of providers’ performance on quality outcomes. However, we also recognize that future CMS and/or NQF initiatives may combine resource use/MPSB with providers’ quality performance such as in a VBP program. While AMRPA supports a national benchmark for expenditures as stated earlier, AMRPA does not support holding providers to a national benchmark for quality performance such as patient outcomes measures. There are regional variations in patient outcomes beyond a provider’s control. For example, urban settings have higher return to community rates than rural regions. If a regional benchmark (however regional is defined) were to be used, it would need to include a sample size of at least 30 IRFs for statistical purposes and power analysis. As noted, we support the episode as outlined in the description and measure specifications. However, there are too few details included in the MUC List materials to effectively determine if the measure is appropriately developed. For example, the risk-adjustment factors, if any, are not identified. In addition, several keys terms need to be defined. 1. The measure description states, “The MSPB-PAC episode includes all services during the episode window that are attributable to the IRF provider and those rendered by other providers, except those services during the associated services period that are clinically unrelated to IRF responsibilities (e.g., planned care and routine screening).” However, planned care and routine services are not defined. This makes it challenging at best to determine if the episode window and services included are appropriate. A clear definition of planned care and routine screenings is critically important because patients with cancer, a transplant, or undergoing dialysis, for example, might have high costs that should not be attributed to the IRF because they are part of an ongoing treatment plan developed for the patient based on their particular clinical needs. 2. Similarly, specifications regarding planned and unplanned readmissions should be defined and open for stakeholder feedback to determine if the episode window and included services are appropriate. 3. The numerator description includes several terms that need to be defined before we can comment completely. They are standardized episode spending, expected episode spending, and average standardized episode spending across all providers. 4. The denominator is defined as “Measure is the weighted median MSPB-PAC Amount across all episodes for IRFs nationally,” but insufficient information is provided about the weighting methodology. Given the limited details provided and the fact that this measure is still under development, AMRPA cannot support the MSPB measure for inclusion in the IRF QRP at this time. Until additional details are provided, it would be premature for the NQF MAP PAC/LTC Workgroup to provide support for this measure. (Submitted by: American Medical Rehabilitation Providers Association)
  
  
• This measure is currently in the early stage of development, has never been submitted to NQF and thus is not NQF-endorsed (Submitted by: AdvaMed)
  
  
• We believe that cost information alone cannot capture the quality of care provided by IRFs. Accordingly, we think that this measure should only be used in close conjunction with other quality outcome measures. As such, we believe that NQF’s consideration of this measure should include special attention to its “intended use” within the Medicare program. Furthermore, while we support the development of resource use measures generally, we believe that such measures, like the IRF MSPB measure, should also have adequate risk adjustment in order to account for unique patient mix and related clinical characteristics that can exist within the IRF patient population. (Submitted by: HealthSouth)
  
  

• As an IMPACT Act measure that is stated to be "standardized and interoperable", we are highly concerned that this measure may allow for comparison between PAC venues without making adjustments for patient severity or differences inherent within the various payment systems. For example, some of the PAC venues are paid on a per diem basis, while other PAC venues have payments that are episodic. How are these differences accounted for in the measurement calculations. Additionally, each PAC venue has different regulatory requirements impacting their ability to control costs, which are not adequately accounted for in this measure. We also note that at this time some of the PAC venues are in the midst of drastic changes to their payment systems. Will the currently developed calculations of this measure be impacted by changes to underlying payment systems? And finally, the IMPACT Act also requests a report from MedPAC on the feasibility of a unified PAC payment system. This work is currently underway, and may provide further clarity to differences in payments between PAC sites. (Submitted by: UDSMR)
  
  
• This measure is currently in the early stage of development, has never been submitted to NQF and thus is not NQF-endorsed (Submitted by: AdvaMed)
  
  
• Cases in an LTCH can be paid according to two different payment structures, standard MS-LTC-DRG or the site neutral rate, depending on whether the case meets certain criteria. Recognizing the differences in these payment structures, the MSPB-PAC LTCH QRP Measure is constructed differently for standard and site neutral LTCH cases. Specifically, the associated services period for the standard LTCH cases begins at admission to an LTCH and ends 30 days after discharge whereas the associated services period for the site neutral cases begins at discharge from LTCH and ends 30 days after discharge. Issues Related to Standard and Site Neutral LTCH Cases Although the measure is constructed differently for each LTCH case, both standard and site neutral cases are pooled in estimating the hierarchical model that drives the predicted and expected spending in the numerator definition. Since the associated services period for standard LTCH cases also includes the LTCH stay, the resource use for the two types of cases may be systematically different even after controlling for patient characteristics observable in the Medicare claims data. As a result, hospitals that have a higher percentage of standard LTCH cases compared to the average hospital may have higher average ratio of standardized to expected episode spending than the average hospital. Therefore, the MSPB-PAC LTCH QRP measure may reflect differences in the prevalence of standard vs. site neutral cases across LTCHs instead of differences in resource use efficiency. Issues Related to Risk Adjustment The risk adjustment variables should include factors that control for differences in patient acuity across settings, including the use of assessment data being collected under the IMPACT Act. We recommend adjusting for factors that play a significant role in determining resource use and outcomes in the LTCH setting, such as prolonged mechanical ventilation, multiple organ failure, and three or more days in an intensive care unit. For consistency across PAC providers, these risk adjustors should be from a prior short term acute care hospital stay, generating the need to limit the measure sample to those who have had a short term acute care hospital stay within a given period prior to PAC stay. We also recommend adjustment based on socioeconomic factors. While Medicare claims data are more readily available than other data sources, they may not capture finer distinctions across patients that affect patients’ outcomes. Therefore, a process to include assessment data in the MSPB-PAC measure calculations, once available, needs to be established and followed. The measure should not be used to compare across setting or at least not used to compare across setting until the assessment data have been incorporated, tested, and validated. (Submitted by: National Association of Long Term Hospitals)
  
  

• This measure is currently in early development and measure specifications are in preliminary stages of clarification. The proposed application and episode extension appears to create inaccurate spending within successive post-acute settings without any indication of the true cost of care per beneficiary. The Technical Expert Panel clearly indicated that this measure needs extensive work to assure accuracy and to meet the purpose for which it was intended. Because there is obvious potential for accuracy in the proposed measure significant continued development should be encouraged. (Submitted by: Genesis Health Care)
  
  
• AHCA is pleased to have the opportunity to comment on the List of Measures under Consideration for December 1, 2015”. The American Health Care Association (AHCA) represents more than 12,000 non-profit and proprietary skilled nursing centers and assisted living communities. Rather than having many of our individual members who have contacted us with comments and to more efficiently provide NQF with feedback, we have received from our various members, we have summarized their comments here. Thank you again for the opportunity to provide these comments. Please contact me at DGifford@ahca.org with any questions or for additional information. This measure has not been specified, even in draft form. In fact, the TEP discussion raised some serious concerns about the measure. While no one has seen any specifications for the measure, including TEP members, the material shared with the TEP indicates that the measure will double count costs between PAC providers (and in the case of SNF and HH, will double count certain beneficiary costs within a provider). For example, the TEP panelists were presented with a measure model ‘concept’ that, instead of measuring the cost of each beneficiary as the beneficiary navigates through their post-acute clinical path, and attributing the true per-beneficiary costs to the provider that was the beneficiary’s PAC entry point, the model instead attributes beneficiary costs from the point of entry onto each PAC setting, as well as healthcare costs for a specified time after discharge from each setting. This results in the double-counting of per-beneficiary costs for the time overlap when a beneficiary transitions from one PAC provider to another. In the specific case of SNF, the model presented to the TEP would double-count certain beneficiary costs within a single 100-day benefit period in cases where the beneficiary was hospitalized for 8 or more days and readmitted to the SNF during the same SNF benefit period. Similarly for HH, the model presented would double-count certain beneficiary costs for a single HH provider for beneficiaries with chronic complex needs receives continuous care for the same admission over several consecutive HH 60-day ‘episode’ payment blocs. The rationale for this does not appear to make sense and also is contrary to the legislative intent of the IMPACT act to develop a set of common PAC measures that allows comparability between PAC settings. It also could create beneficiary access issues for patients with chronic complex needs in settings where per-beneficiary costs within a single course of treatment are double counted as providers would be dis-incentivized to admit patients with such characteristics. This measure requires more thoughtful deliberation. The CMS measure development contractor convened the TEP on October 30-31 without providing several panelists with necessary review materials in advance, and then asked the TEP for input on over 5 pages of 30 questions about how to develop and specify the measure on November 18th with comments due on November 25th. This was only two days before the publication of the MUC list so it is quite apparent that the TEP panel’s comments were not considered adequately prior to the submission of this proposed measure. Therefore, we believe this measure does not meet NQF MAP criteria for endorsement but rather should receive a vote of “encourage continued development”. (Submitted by: American Health Care Association )
  
  
• As an IMPACT Act measure that is stated to be "standardized and interoperable", we are highly concerned that this measure may allow for comparison between PAC venues without making adjustments for patient severity or differences inherent within the various payment systems. For example, some of the PAC venues are paid on a per diem basis, while other PAC venues have payments that are episodic. How are these differences accounted for in the measurement calculations. Additionally, each PAC venue has different regulatory requirements impacting their ability to control costs, which are not adequately accounted for in this measure. We also note that at this time some of the PAC venues are in the midst of drastic changes to their payment systems. Will the currently developed calculations of this measure be impacted by changes to underlying payment systems? And finally, the IMPACT Act also requests a report from MedPAC on the feasibility of a unified PAC payment system. This work is currently underway, and may provide further clarity to differences in payments between PAC sites. (Submitted by: UDSMR)
  
  
• This measure is currently in the early stage of development, has never been submitted to NQF and thus is not NQF-endorsed (Submitted by: AdvaMed)
  
  
• AMDA is pleased to have the opportunity to comment on the List of Measures under Consideration for December 1, 2015”. AMDA - The Society for Post-Acute and Long-Term Care Medicine (AMDA) represents over 5,500 physician medical directors, attending physicians, and other practitioners who work in post-acute and long-term care settings. Thank you again for the opportunity to provide these comments. Please contact me at abardakh@amda.com with any questions or for additional information. This measure has not been specified, even in draft form. In fact, the TEP discussion raised some serious concerns about the measure. While no one has seen any specifications for the measure, including TEP members, the material shared with the TEP indicates that the measure will double count costs between PAC providers (and in the case of SNF and HH, will double count certain beneficiary costs within a provider). For example, the TEP panelists were presented with a measure model ‘concept’ that, instead of measuring the cost of each beneficiary as the beneficiary navigates through their post-acute clinical path, and attributing the true per-beneficiary costs to the provider that was the beneficiary’s PAC entry point, the model instead attributes beneficiary costs from the point of entry onto each PAC setting, as well as healthcare costs for a specified time after discharge from each setting. This results in the double-counting of per-beneficiary costs for the time overlap when a beneficiary transitions from one PAC provider to another. In the specific case of SNF, the model presented to the TEP would double-count certain beneficiary costs within a single 100-day benefit period in cases where the beneficiary was hospitalized for 8 or more days and readmitted to the SNF during the same SNF benefit period. Also, costs may be higher in the SNF for patients admitted for prolonged courses of IV antibiotics, wound care, hemodialysis, and ventilator management. Similarly for HH, the model presented would double-count certain beneficiary costs for a single HH provider for beneficiaries with chronic complex needs receives continuous care for the same admission over several consecutive HH 60-day ‘episode’ payment blocs. The rationale for this does not appear to make sense and also is contrary to the legislative intent of the IMPACT act to develop a set of common PAC measures that allows comparability between PAC settings. It also could create beneficiary access issues for patients with chronic complex needs in settings where per-beneficiary costs within a single course of treatment are double counted as providers would be dis-incentivized to admit patients with such characteristics. This measure requires more thoughtful deliberation. The CMS measure development contractor convened the TEP on October 30-31 without providing several panelists with necessary review materials in advance, and then asked the TEP for input on over 5 pages of 30 questions about how to develop and specify the measure on November 18th with comments due on November 25th. This was only two days before the publication of the MUC list so it is quite apparent that the TEP panel’s comments were not considered adequately prior to the submission of this proposed measure. Therefore, we believe this measure does not meet NQF MAP criteria for endorsement but rather should receive a vote of “encourage continued development”. (Submitted by: AMDA-The Society for Post-Acute and Long-Term Care Medicine)
  
  
• This measure is complicated and should be fully vetted before it continues. it does not seem to meet the intent of the IMPACT Act that provides for comparing one post acute setting with another because some costs in the measure are counted twice. This measure needs more detail in order to review it and needs comprehensive testing. (Submitted by: National Association for the Suppor of Long Term Care)
  
  

• This measure is currently in the early stage of development, has never been submitted to NQF and thus is not NQF-endorsed (Submitted by: AdvaMed)
  
  

• This measure is currently in the early stage of development, has never been submitted to NQF and thus is not NQF-endorsed (Submitted by: AdvaMed)
  
  

• We have the following concerns related to this measure: 1. The inclusion of unplanned readmissions in the discharge to community measure development and calculation. Our concerns related to the inclusion of unplanned readmissions in the discharge to community measure development and calculation can be defined as follows: a. Duplication/commingling of quality metrics With the inclusion of unplanned readmissions in the discharge to community measure development and calculation, there is the potential for there to be a strong correlation between the performances of the two measures. In other words, those PAC providers that perform poorly with the unplanned readmission measure are more likely to perform poorly on this measure, while those PAC providers that perform well with the unplanned readmission measure have a better opportunity to perform well on this measure. Take for instance the scenario of two facilities with similar patient populations (age, sex, etc.), where provider A discharges 75% to a community setting with a 5% readmission rate while provider B discharges 72% to a community setting with a 2% readmission rate. Which facility performs better at discharging patients to a community setting? Is the discharge to community measure truly representative of the ability of the post-acute care provider to provide services that get the patient back to a community setting, or is too much emphasis being placed upon the ability of the patient to stay in a community setting? b. Inclusion of a factor that is outside of post-acute care provider control While a post-acute care provider can provide services that impact the ability of a patient to be discharged back to a community setting, whether or not a patient returns to an acute care facility in the 30 days following that PAC discharge can be completely unrelated to the services furnished by the PAC provider. For instance, a patient is discharged home from a post-acute care provider but is re-admitted to acute care 10 days later for pneumonia (which is listed as a diagnosis category that cannot be considered planned). Should the post-acute care provider’s discharge to community percentage be penalized if the readmission is completely unrelated to the services provided within post-acute care? With the concerns noted above, we would recommend that the measure developers remove the unplanned readmission component from the discharge to community measurement, and provide side-by-side comparison of these two separate and unique quality measures. 2. Potential for site-specific risk adjustment for a “cross-setting” measure. While we note that the risk adjustment variables are “under consideration”, we are very concerned that the variables noted in the draft specification designate certain variables as being applicable to unique or specific post-acute care providers. The IMPACT Act requires the specification of quality measures and resource use metrics that are standardized and interoperable across PAC settings. In creating a measure where consideration is being given to risk adjustment variables that differ by post-acute care setting, is CMS truly meeting the definition of “standardized and interoperable”? As an example, “Activities of Daily Living” (ADLs) are noted as one of the Clinical conditions variables, yet is noted for “the HHA setting only”. If the measure developer can show that ADLs can be used to differentiate the performance within the HHA population, couldn’t ADLs also be used to differentiate the performance of all post-acute care settings and produce a measurement value that is “standardized and interoperable”? We urge CMS and the measure developers to introduce risk-adjustment variables that are “standardized and interoperable” in order to meet the IMPACT Act requirements. (Submitted by: UDSMR)
  
  
• This measure is being developed by CMS. Many patients desire to return to their homes and communities after an acute hospital stay and subsequent post-acute care services. At this time IRFs track discharge destination for patients via the Inpatient Rehabilitation Facility Patient Assessment Instrument (IRF PAI) though this measure has the potential to change the way in which they track and report it. IRFs make every effort to return patients to their communities but at times discharge to community is not feasible particularly if a patient is very ill or injured or if the necessary family, caregiver, or community support is not available. Therefore, a discharge to community measure should be developed carefully, with appropriate risk adjustment, to ensure IRFs are not unfairly penalized for not returning such patients to the community. If this measure is not developed properly it may create barriers to access for patients who do not appear to be able to be discharged to their community. We recommend that CMS enhance the risk adjustment methodology to account for socioeconomic and sociodemographic factors as well as the patient’s functional status. As drafted this measure defines discharge to community as home with or without home health services. We believe this definition is too narrow and does not sufficiently account for patients who may not be safely discharged home. CMS intends to use patient discharge status codes as listed in Table 1 of the draft measure specifications. In our comments to the Agency we suggested that this list should be applicable to SNFs only. We believe the low rate of discharge to community in this sector and the high incidence of transfer to the residential portion of its own or another SNF’s nursing home should not be counted as a discharge to the community. Table 1 also fails to collect the level of detail associated with discharge to home as it currently is collected on the IRF PAI. At this time there are two discharge status codes on the IRF PAI associated with discharge to home which should be added for the purposes of this measure; 01: Home (private apartment, board/care, assisted living, group home, transitional living); and 06: Home under the care of an organized home health service organization. Due to our concerns with this measure as it is currently drafted, CMS and NQF should consider measures such as the patient’s return to the highest practicable level of function or independence or the achievement of his or her goals at discharge. We believe such measures are more likely to improve patient outcomes and add value to the measure set already in use in the IRF QRP. As noted, CMS is still in the measure development process. The Agency has requested comments which are due December 8. Given that this measure is in the CMS development process, it may be premature to endorse this measure for the purpose of the IRF QRP. (Submitted by: American Medical Rehabilitation Providers Association)
  
  
• The ultimate goals of post-acute care are avoiding institutionalization and returning patients to their previous level of independence and functioning, with discharge to community being the primary goal for the majority of post-acute patients. For many, home is a symbol of independence, privacy, and competence. Discharge to community is considered a valuable outcome to measure because it is a multifaceted measure that captures the patient’s functional status, cognitive capacity, physical ability, and availability of social support at home. There is considerable variation in discharge to community rates within and across post-acute settings. Studies show geographic variation, variation across patient socioeconomic characteristics (for example, race and ethnicity), and variation by facility characteristics (for profit vs. nonprofit, freestanding vs. hospital-based, urban vs. rural). In the IRF setting, discharge to community rates vary across providers, ranging from about 60% to 75%. The 2015 MedPAC report shows that, in FY 2013, the facility-level, mean risk-adjusted discharge to community rate for IRFs within 100 days of admission was 75.8%, and the mean observed rate was 74.7%. Discharge to community rates also vary widely in the SNF setting, ranging from as low as 31% to as high as 65%. The 2015 MedPAC report shows a mean risk-adjusted discharge to community rate of 37.5% for SNFs within 100 days of admission, and mean observed rate of 40.1%. A multicenter study of 23 LTCHs reported that only 28.8% of 1,061 patients who were ventilator-dependent on admission were discharged to home or assisted living facility. A study of 66,510 Medicare beneficiaries during pre- and post-HH episodes, revealed that 64 percent of beneficiaries discharged from HH did not use any other Medicare-reimbursed acute or post-acute services in the 30 days following HH discharge. Significant numbers of patients were admitted to inpatient facilities (29percent) and lesser numbers to skilled nursing facilities (7.6 percent), inpatient rehabilitation (1.5 percent) and home health (7.2 percent) or hospice (3.3 percent) within 30 days of HH discharge (Wolff et al., 2008). II. Feedback Overall, there is not enough detail on the measure specifications to provide adequate feedback. However, despite the lack of detail, a response is included below on the discharge to community measure for LTCH, IRF, SNF, and HH. Description Patients discharged to the community following PAC stay/episode and do not have an unplanned admission to an acute care hospital or LTCH in 31 days after community discharge and remain alive • Although not explicitly stated, it appears that this measure is only holding the last PAC setting responsible for the discharge? Given what is known about multiple sites of PAC use either from LTCH/IRF patients are often discharged to SNF before community discharge, as it is currently worded this measure will penalize the SNF for the more difficult patients that could not be discharged from the prior PAC setting. • Research has shown that PAC patients re-enter the healthcare system through multiple doors after the PAC community discharge why are these not captured in the measure as a failed transition I recommend that the numerator should not include individuals who were admitted to a SNF after PAC community discharge during the 30-day measure window. Numerator & Denominator There is not enough detail to provide feedback on the risk adjustment of the numerator. Without details on the variables included, the origin of those variables, and methods of risk adjustment it is difficult to provide feedback. Recommend providing further detail. Recommend including functional status, functional cognition, presence of social support (e.g., does the individual live alone), and comorbid conditions as risk adjustors Exclusion There is no discussion of excluding long-stay nursing home residents in this measure. Long-stay residents that have an acute medical event, are admitted to an acute care hospital, discharged to PAC for care, and then transitioned back to long-term care should be excluded from the measure. These individuals are not returning to the community, while the nursing home is their home, the effort and processes required for transitioning a long-stay nursing home resident from PAC back to custodial care is very different that transitioning a community- residing older adult back to their residence in the community after a PAC stay. Recommend excluding long stay nursing home residents from the measure. (Submitted by: American Occupational Therapy Association)
  
  
• ARN is supportive of a measure that works to identify unplanned (re)admissions; however, we have concerns with identifying unplanned (re)admissions based on the planned readmissions algorithm used in NQF measure #2510: SNF 30-Day All-Cause Readmission Measure (SNFRM); NQF #2502: All-Cause Unplanned Readmission Measure for 30 Days Post Discharge from IRFs; NQF #2512: All-Cause Unplanned Readmission Measure for 30 Days Post Discharge from LTCHs; and NQF #2380: Re-hospitalization During the First 30 Days of Home Health. The exclusion criteria included within NQF #2510 for SNF stays where the patient had one or more intervening PAC admissions to an IRF that occurred either between the prior proximal hospital discharge and SNF admission or after the SNF discharge within the 30-day risk window fails to allow for a medically complex patient that is treated in an IRF and readmitted to the SNF within 30 days for a condition that may initially have been treated as a comorbidity. We disagree with the rationale provided for exclusion, for while the measure assesses readmission rates while accounting for patient demographics, principal diagnosis in the prior hospitalization, comorbidities, and other patient factors, often, this may not be the reason for admission to a SNF. ARN believes the measure should include the principal diagnosis during the prior proximal hospitalization, comorbidities based on the secondary medical diagnoses listed on the patient’s prior proximal hospital claim and diagnoses from prior hospitalizations that occurred in the previous 365 days, length of stay during the patient’s prior proximal hospitalization, length of stay in the intensive care unit (ICU), body system specific surgical indicators, End-Stage Renal Disease (ESRD) status, whether the patient was disabled, and the number of prior hospitalizations in the previous 365 days. It also would be beneficial to understand the comorbidities being evaluated in the risk-adjustment model. ARN urges the Centers for Medicare and Medicaid Services (CMS) to develop of a list of comorbidities, comparable to the IRF Prospective Payment System (PPS) list of comorbidities and we encourage CMS to categorize an intervening admission to an IRF as a proximal hospitalization. Additionally, ARN has serious concerns with CMS’s proposal to require PAC providers to utilize Medicare claims data to calculate their 30-day readmission rates. Using claims data to calculate readmission rates is difficult for health care providers, as claims data are cumbersome to use and access. Employing a 30-day readmission rate measure will not provide meaningful insight or have an impact on quality improvement efforts if the PAC settings do not have unrestricted access to the data. Further, ARN believes that patients who have been discharged to the community and expire within the post-discharge window should not be included within the quality measure, given the variation in patient characteristics across the four settings. For example, as compared to all Medicare beneficiaries, the SNF and LTCH patient populations represent the most disabled, elderly, and frail beneficiaries. The Medicare Payment Advisory Commission’s (MedPAC) March 2015 Report to Congress found that compared with other beneficiaries, “SNF users are older, frailer, and disproportionately female, disabled, living in an institution, and dually eligible for both Medicare and Medicaid.” Moreover, as compared with all Medicare beneficiaries, those admitted to LTCHs are “disproportionately disabled (under age 65), over age 85, or diagnosed with [ESRD]. They are also more likely to be African American.” ARN urges CMS to exclude patients that pass away within the post-discharge window after being discharged to the community from the discharge to community quality measure, as the types of patients treated in each setting greatly varies and can lead to an inaccurate reflection of the quality of care. ARN is pleased the CMS has proposed discharge measure exclusions; however, we have concerns with the proposed exclusion of post-acute stays that end in transfer to the same level of care, and specifically, CMS’s proposal to include only the final post-acute provider in the discharge to community measure. The Agency’s proposed exclusion criteria fails to consider when a patient’s “home” is a custodial nursing facility and the patient’s post-acute episode involves a discharge back to his or her “home.” In such circumstances, including the final post-acute provider in the discharge to community measure when a patient is discharged to the originating level of care, but in essence, is returning home, may distort the findings of the quality measure. We encourage CMS to design a quality measure that is capable of capturing the difference between a patient’s return to his or her home and a patient’s post-acute episode that involves transfer to the same level of care. (Submitted by: Association of Rehabilitation Nurses)
  
  
• This measure is currently in the early stage of development, has never been submitted to NQF and thus is not NQF-endorsed (Submitted by: AdvaMed)
  
  
• ARN is supportive of a measure that works to identify unplanned (re)admissions; however, we have concerns with identifying unplanned (re)admissions based on the planned readmissions algorithm used in NQF measure #2510: SNF 30-Day All-Cause Readmission Measure (SNFRM); NQF #2502: All-Cause Unplanned Readmission Measure for 30 Days Post Discharge from IRFs; NQF #2512: All-Cause Unplanned Readmission Measure for 30 Days Post Discharge from LTCHs; and NQF #2380: Re-hospitalization During the First 30 Days of Home Health. The exclusion criteria included within NQF #2510 for SNF stays where the patient had one or more intervening PAC admissions to an IRF that occurred either between the prior proximal hospital discharge and SNF admission or after the SNF discharge within the 30-day risk window fails to allow for a medically complex patient that is treated in an IRF and readmitted to the SNF within 30 days for a condition that may initially have been treated as a comorbidity. We disagree with the rationale provided for exclusion, for while the measure assesses readmission rates while accounting for patient demographics, principal diagnosis in the prior hospitalization, comorbidities, and other patient factors, often, this may not be the reason for admission to a SNF. ARN believes the measure should include the principal diagnosis during the prior proximal hospitalization, comorbidities based on the secondary medical diagnoses listed on the patient’s prior proximal hospital claim and diagnoses from prior hospitalizations that occurred in the previous 365 days, length of stay during the patient’s prior proximal hospitalization, length of stay in the intensive care unit (ICU), body system specific surgical indicators, End-Stage Renal Disease (ESRD) status, whether the patient was disabled, and the number of prior hospitalizations in the previous 365 days. It also would be beneficial to understand the comorbidities being evaluated in the risk-adjustment model. ARN urges the Centers for Medicare and Medicaid Services (CMS) to develop of a list of comorbidities, comparable to the IRF Prospective Payment System (PPS) list of comorbidities and we encourage CMS to categorize an intervening admission to an IRF as a proximal hospitalization. Additionally, ARN has serious concerns with CMS’s proposal to require PAC providers to utilize Medicare claims data to calculate their 30-day readmission rates. Using claims data to calculate readmission rates is difficult for health care providers, as claims data are cumbersome to use and access. Employing a 30-day readmission rate measure will not provide meaningful insight or have an impact on quality improvement efforts if the PAC settings do not have unrestricted access to the data. Further, ARN believes that patients who have been discharged to the community and expire within the post-discharge window should not be included within the quality measure, given the variation in patient characteristics across the four settings. For example, as compared to all Medicare beneficiaries, the SNF and LTCH patient populations represent the most disabled, elderly, and frail beneficiaries. The Medicare Payment Advisory Commission’s (MedPAC) March 2015 Report to Congress found that compared with other beneficiaries, “SNF users are older, frailer, and disproportionately female, disabled, living in an institution, and dually eligible for both Medicare and Medicaid.” Moreover, as compared with all Medicare beneficiaries, those admitted to LTCHs are “disproportionately disabled (under age 65), over age 85, or diagnosed with [ESRD]. They are also more likely to be African American.” ARN urges CMS to exclude patients that pass away within the post-discharge window after being discharged to the community from the discharge to community quality measure, as the types of patients treated in each setting greatly varies and can lead to an inaccurate reflection of the quality of care. ARN is pleased the CMS has proposed discharge measure exclusions; however, we have concerns with the proposed exclusion of post-acute stays that end in transfer to the same level of care, and specifically, CMS’s proposal to include only the final post-acute provider in the discharge to community measure. The Agency’s proposed exclusion criteria fails to consider when a patient’s “home” is a custodial nursing facility and the patient’s post-acute episode involves a discharge back to his or her “home.” In such circumstances, including the final post-acute provider in the discharge to community measure when a patient is discharged to the originating level of care, but in essence, is returning home, may distort the findings of the quality measure. We encourage CMS to design a quality measure that is capable of capturing the difference between a patient’s return to his or her home and a patient’s post-acute episode that involves transfer to the same level of care. (Submitted by: Association of Rehabilitation Nurses)
  
  
• We believe that the PAC industry should embrace measures, such as DTC, that place an emphasis on the ability for patients to return and stay at home – a central tenant of post-acute care. We have several comments that will serve as constructive additions to the development of these measures. We hope that NQF and measure developers will analyze and consider these comments and how they could improve the DTC measure development for IRFs. A. THE DISCHARGE TO COMMUNITY MEASURE SHOULD BE DISTINGUISHED FROM THE 31-DAY POST-DISCHARGE READMISSIONS MEASURES The measure specification document states that this DTC measure tracks “patients/residents/persons who are discharged to the community, and do not have an unplanned (re)admission to an acute care hospital or LTCH on the day of discharge or in the 31 days following discharge to community, and remain alive during the 31 days following discharge to community.” Based on this description, this DTC measure, although not called a “readmissions” measure, functions as a readmission measure, similar to those post-acute readmissions measures already in place and being developed under the IMPACT Act (All-cause Unplanned Readmissions (NQF #2502), and the Potentially Preventable Readmissions (“PPR”) measures under development), but instead focused on a subset of a provider’s patient population – Medicare patients that are discharged to community. If those existing (and soon-to-be implemented) readmission measures already track the number of people who are readmitted to a hospital within 30 days after being discharged from the PAC setting either under an all-cause basis or a potentially preventable basis, and this DTC measure tracks those patients that are discharged to community and do not have a “(re)admission to an acute care hospital or LTCH” within the 31 day timespan, then all of the patients who have an “unfavorable outcome” under the DTC measure will also and already be captured in at least one, if not both, of these existing readmissions measures. Instead of using the DTC measure to track and eventually publicly report on such readmissions, which are already tracked by the other readmission measures soon to be in place, it would be more meaningful to track the number of patients who go directly to the community (numerator) vs. the number of patients expected to go directly to the community (denominator), regardless of subsequent readmissions. The proposed DTC measure design is already built around comparing the actual number of successful discharges to community vs. expected successful discharges, but our comment here is directed at the definition of what it means to be “discharged to community.” In other words, if “discharged to community” means “31 days at home without a readmission,” (as proposed by measure developers) then this measure effectively becomes a readmission measure that selectively focuses on those patients who are discharged home – a population that is already covered both by the All-cause Unplanned Readmissions measure (NQF #2502) and the proposed 30-day Post-discharge Potentially Preventable Readmission measure (under development at RTI). If, on the other hand, “discharged to community” means “patients who are discharged home without an intervening facility stay” (regardless of a subsequent readmission), that definition does not unnecessarily overlap with other readmission measures, thus giving providers and the public a more useful measure that is distinct from existing measures. B. MEASURE SHOULD REMAIN LIMITED TO MEDICARE FEE-FOR-SERVICE POPULATION According to the measure description (specifically Measure Exclusion 7), this measure is only applicable to Medicare fee-for-service (“FFS”) populations, and not other payer populations, including Medicare Advantage (“MA”). We support this exclusive focus on FFS because other payers, including private insurance companies that contract with the Medicare program to manage MA plans, often dictate what services IRFs and other PAC providers can offer to their discharged patients, thereby limiting the ability of such providers to maximize their efforts to keep people healthy at home following discharge. This has the potential to negatively impact the ability to optimize performance on this DTC measure (and other readmission measures). We note that, while CMS and RTI have made the laudable decision to focus only on Medicare FFS for this DTC measure, we think this approach should extend to other measures developed under the IMPACT Act as well. If some IMPACT Act measures focus solely on Medicare FFS while others cover all patients and all payers, a significant disconnect in the underlying data will arise. Despite the logic behind this approach, CMS’ response to the Drug Regimen Review measure public comments, released last week, indicated support for the opposite view. In response to comments on that Drug Regimen Review measure recommending that all PAC settings assess the same populations with the IMPACT measures and that the denominators be limited to Medicare fee-for-service enrollees only (since this is the population that is subject to changes in payment policy under the IMPACT Act), CMS replied saying that it felt that quality improvements are an appropriate goal for all patients, “regardless of payer source.” While no one would disagree with this broad statement of quality improvement across payer populations, this view does not account for the specific fact that, if a CMS quality measure that is ultimately tied to Medicare reimbursement measures non-Medicare populations for which PAC providers are limited (by non-Medicare payers) in post-discharge service offerings, such a scenario potentially thwarts a providers ability to control and improve performance on that measure, and could have significant payment or regulatory implications. C. RISK ADJUSTMENT FOR SOCIODEMOGRAPHIC AND FUNCTIONAL STATUS The draft specifications indicate that the DTC measure will be risk-adjusted for multiple variables, including age, sex, dialysis status, and prior acute care utilization in the past year. We generally support this broad range of risk adjustment, but also believe that the measure developers should consider including risk adjustment for additional sociodemographic factors and functional status as well, such as geographic area (urban/rural) or percentage of low income patients. Recent academic literature has added evidence to the notion that readmissions back to hospitals are driven by more directly by patient status factors and not by the quality of care delivered. Since this DTC measure, as currently proposed and as argued above, is (in its current form) effectively a readmissions measure focused on the population of patients discharged directly to community, we feel that this evidence is also applicable to this measure, not just the formal readmission measures. Furthermore, The National Quality Forum (“NQF”) in April began a two-year trial program of a temporary policy change that would allow risk adjustment of performance measures for socioeconomic and other demographic factors, a departure from earlier quality measurement positions that viewed sociodemographic risk adjustment as inappropriate. With evidence that sociodemographic risk is real and impact readmission rates, we recommend that measure developers and NQF consider including it in the list of risk adjustment factors for these measures. We also believe adjusting for functional status would benefit the accuracy of the DTC measure. Functional status has a direct correlation with a patient’s ability to remain healthy at home after PAC services have ended, meaning that PAC providers who treat more functionally impaired patients will likely have a higher readmission rate, and a lower DTC rate. Accordingly, we recommend that NQF and the measure developers consider how to apply an additional risk-adjustment factor for functional status in the overall risk adjustment methodology. D. RISK OF CONFUSION IN MULTIPLE READMISSION MEASURES As part of the existing IRF PPS Quality Reporting Program (“IRF QRP”), IRFs already report All-cause Unplanned Readmission (NQF #2502) and are scheduled to have the PPR measure publicly reported beginning next year, not to mention “within stay” readmissions measures. Other PAC types will also require public reporting of a general all-cause readmissions measure. We are concerned that, as required by the IMPACT Act and various QRP provisions, the eventual public reporting of all of these measures (which are essentially a subset of the more general all-cause measures) will result in substantial confusion amongst members of the public. For example, if both readmissions measures and the DTC measure are all publicly reported without thorough explanations, PAC providers will have one all-cause unplanned readmission rate, a separate PPR rate, a separate PPR “within stay” rate, and a DTC percentage that is measured in the opposite direction (looking at how many patients stayed home instead of how many patients were readmitted). As a side comment, we note that it may be more administratively simple if CMS instead measured DTC as a readmissions measure for patients discharged to home. That would harmonize this version of a readmissions rate with the other measures. Nevertheless, we question whether the crucial distinctions between these different, but related, readmissions indicators will be readily apparent to members of the public who take the time to assess different PAC providers based on readmission rates. Similarly, providers may have trouble accurately understanding the purpose and ultimate use of additional readmissions indicators. IRF providers already receive annual PEPPER reports and also report on the All-cause measure (NQF #2502). With the addition of a “within stay” readmissions measure, a post-discharge PPR measure, and this DTC measure, we think it would be highly valuable if CMS were to take deliberate steps to clearly communicate the intended use of and distinctions between each of these readmission tools, making sure to note and how they will relate to one another. Without such clarifying communication, providers may find themselves adrift in various readmissions data without a clear idea of how it is all being used by the Agency. Lastly, in order to standardize readmission measures and avoid unnecessary confusion, we ask that CMS and RTI standardize the language and definitions used in conjunction with establishing observation windows for all readmissions measures. For example, the IRF All-Cause Unplanned Readmissions measure (NQF #2502) is measured beginning two days after the day of discharge for 30 days; this DTC measure begins one day after the day of discharge for 31 days. These two measures both end on the same day, but it could help to avoid confusion if these discharge and observation window definitions and timeframes were standardized. (Submitted by: HealthSouth)
  
  

• APTA supports the goal of improving the quality of health care. Physical therapists are committed to providing high-quality, timely care and to the promotion of evidence-based and patient-centered practice. Furthermore, APTA feels that it is essential that we move towards a common set of quality measures across the continuum of care. APTA supports the implementation of a discharge to community measure across the care settings. We believe that successful transitions to the community following discharge from the respective post-acute care settings will decrease potentially preventable readmissions. However, the APTA does have some concerns regarding the proposed measure methodology. These concerns are discussed below. APTA believes that a patient’s level of function does impact a patients ability to transition successfully back to the community. Recent evidence indicates that patient function is associated with increased risk of 30-day all-cause hospital readmissions and may be an important factor in preventing readmissions for Medicare seniors that is not currently accounted for in measure methodologies1. APTA was pleased to see “activity of daily living” scores in the home health setting included in the risk adjustment methodology for the readmissions measures, and we recently commented, encouraging the use of patient function in the risk adjustment methodology for the post-acute care setting readmissions measures. We believe that readmissions and discharge to community are closely related measures and that patient function may also be an important risk adjustment variable for discharge to community. APTA appreciates that CMS has strict deadlines for the implementation of measures under the IMPACT act, however, as these measures will be new to the respective post-acute care settings, we encourage that settings have the ability to review this data as early as possible in order to understand and, more importantly, so that the respective setting have time to implement strategies to decrease readmissions where necessary. As many of these settings do not always receive feedback on the readmissions of their patients post-discharge, this data will be new to many facilities. APTA recognizes that the overall goal of IMPACT is for PAC providers (HH, IRF, SNF and LTCH) to collect and report standardized and interoperable patient assessment data, quality and resource use measures. We acknowledge that during the initial IMPACT implementation years that there will be a transition period which will include the addition of new measures into all of the post-acute care settings. We believe that achieving a standardized and interoperable patient assessment data set and stable quality measures as quickly as possible will allow for better cross-setting comparisons as well as the evolution of better quality measures with uniform risk standardization, thus achieving the true aim of IMPACT. 1 Greysen SR, Cenzer IS, Auerbach AD, Covinsky KE. Functional Impairment and Hospital Readmission in Medicare Seniors. JAMA Intern Med. 2015;175(4):559-565. (Submitted by: American Physical Therapy Association )
  
  
• We have the following concerns related to this measure: 1. The inclusion of unplanned readmissions in the discharge to community measure development and calculation. Our concerns related to the inclusion of unplanned readmissions in the discharge to community measure development and calculation can be defined as follows: a. Duplication/commingling of quality metrics With the inclusion of unplanned readmissions in the discharge to community measure development and calculation, there is the potential for there to be a strong correlation between the performances of the two measures. In other words, those PAC providers that perform poorly with the unplanned readmission measure are more likely to perform poorly on this measure, while those PAC providers that perform well with the unplanned readmission measure have a better opportunity to perform well on this measure. Take for instance the scenario of two facilities with similar patient populations (age, sex, etc.), where provider A discharges 75% to a community setting with a 5% readmission rate while provider B discharges 72% to a community setting with a 2% readmission rate. Which facility performs better at discharging patients to a community setting? Is the discharge to community measure truly representative of the ability of the post-acute care provider to provide services that get the patient back to a community setting, or is too much emphasis being placed upon the ability of the patient to stay in a community setting? b. Inclusion of a factor that is outside of post-acute care provider control While a post-acute care provider can provide services that impact the ability of a patient to be discharged back to a community setting, whether or not a patient returns to an acute care facility in the 30 days following that PAC discharge can be completely unrelated to the services furnished by the PAC provider. For instance, a patient is discharged home from a post-acute care provider but is re-admitted to acute care 10 days later for pneumonia (which is listed as a diagnosis category that cannot be considered planned). Should the post-acute care provider’s discharge to community percentage be penalized if the readmission is completely unrelated to the services provided within post-acute care? With the concerns noted above, we would recommend that the measure developers remove the unplanned readmission component from the discharge to community measurement, and provide side-by-side comparison of these two separate and unique quality measures. 2. Potential for site-specific risk adjustment for a “cross-setting” measure. While we note that the risk adjustment variables are “under consideration”, we are very concerned that the variables noted in the draft specification designate certain variables as being applicable to unique or specific post-acute care providers. The IMPACT Act requires the specification of quality measures and resource use metrics that are standardized and interoperable across PAC settings. In creating a measure where consideration is being given to risk adjustment variables that differ by post-acute care setting, is CMS truly meeting the definition of “standardized and interoperable”? As an example, “Activities of Daily Living” (ADLs) are noted as one of the Clinical conditions variables, yet is noted for “the HHA setting only”. If the measure developer can show that ADLs can be used to differentiate the performance within the HHA population, couldn’t ADLs also be used to differentiate the performance of all post-acute care settings and produce a measurement value that is “standardized and interoperable”? We urge CMS and the measure developers to introduce risk-adjustment variables that are “standardized and interoperable” in order to meet the IMPACT Act requirements. (Submitted by: UDSMR)
  
  
• The ultimate goals of post-acute care are avoiding institutionalization and returning patients to their previous level of independence and functioning, with discharge to community being the primary goal for the majority of post-acute patients. For many, home is a symbol of independence, privacy, and competence. Discharge to community is considered a valuable outcome to measure because it is a multifaceted measure that captures the patient’s functional status, cognitive capacity, physical ability, and availability of social support at home. There is considerable variation in discharge to community rates within and across post-acute settings. Studies show geographic variation, variation across patient socioeconomic characteristics (for example, race and ethnicity), and variation by facility characteristics (for profit vs. nonprofit, freestanding vs. hospital-based, urban vs. rural). In the IRF setting, discharge to community rates vary across providers, ranging from about 60% to 75%. The 2015 MedPAC report shows that, in FY 2013, the facility-level, mean risk-adjusted discharge to community rate for IRFs within 100 days of admission was 75.8%, and the mean observed rate was 74.7%. Discharge to community rates also vary widely in the SNF setting, ranging from as low as 31% to as high as 65%. The 2015 MedPAC report shows a mean risk-adjusted discharge to community rate of 37.5% for SNFs within 100 days of admission, and mean observed rate of 40.1%. A multicenter study of 23 LTCHs reported that only 28.8% of 1,061 patients who were ventilator-dependent on admission were discharged to home or assisted living facility. A study of 66,510 Medicare beneficiaries during pre- and post-HH episodes, revealed that 64 percent of beneficiaries discharged from HH did not use any other Medicare-reimbursed acute or post-acute services in the 30 days following HH discharge. Significant numbers of patients were admitted to inpatient facilities (29percent) and lesser numbers to skilled nursing facilities (7.6 percent), inpatient rehabilitation (1.5 percent) and home health (7.2 percent) or hospice (3.3 percent) within 30 days of HH discharge (Wolff et al., 2008). II. Feedback Overall, there is not enough detail on the measure specifications to provide adequate feedback. However, despite the lack of detail, a response is included below on the discharge to community measure for LTCH, IRF, SNF, and HH. Description Patients discharged to the community following PAC stay/episode and do not have an unplanned admission to an acute care hospital or LTCH in 31 days after community discharge and remain alive • Although not explicitly stated, it appears that this measure is only holding the last PAC setting responsible for the discharge? Given what is known about multiple sites of PAC use either from LTCH/IRF patients are often discharged to SNF before community discharge, as it is currently worded this measure will penalize the SNF for the more difficult patients that could not be discharged from the prior PAC setting. • Research has shown that PAC patients re-enter the healthcare system through multiple doors after the PAC community discharge why are these not captured in the measure as a failed transition I recommend that the numerator should not include individuals who were admitted to a SNF after PAC community discharge during the 30-day measure window. Numerator & Denominator There is not enough detail to provide feedback on the risk adjustment of the numerator. Without details on the variables included, the origin of those variables, and methods of risk adjustment it is difficult to provide feedback. Recommend providing further detail. Recommend including functional status, functional cognition, presence of social support (e.g., does the individual live alone), and comorbid conditions as risk adjustors Exclusion There is no discussion of excluding long-stay nursing home residents in this measure. Long-stay residents that have an acute medical event, are admitted to an acute care hospital, discharged to PAC for care, and then transitioned back to long-term care should be excluded from the measure. These individuals are not returning to the community, while the nursing home is their home, the effort and processes required for transitioning a long-stay nursing home resident from PAC back to custodial care is very different that transitioning a community- residing older adult back to their residence in the community after a PAC stay. Recommend excluding long stay nursing home residents from the measure. (Submitted by: American Occupational Therapy Association)
  
  
• This measure is currently in the early stage of development, has never been submitted to NQF and thus is not NQF-endorsed (Submitted by: AdvaMed)
  
  
• ARN is supportive of a measure that works to identify unplanned (re)admissions; however, we have concerns with identifying unplanned (re)admissions based on the planned readmissions algorithm used in NQF measure #2510: SNF 30-Day All-Cause Readmission Measure (SNFRM); NQF #2502: All-Cause Unplanned Readmission Measure for 30 Days Post Discharge from IRFs; NQF #2512: All-Cause Unplanned Readmission Measure for 30 Days Post Discharge from LTCHs; and NQF #2380: Re-hospitalization During the First 30 Days of Home Health. The exclusion criteria included within NQF #2510 for SNF stays where the patient had one or more intervening PAC admissions to an IRF that occurred either between the prior proximal hospital discharge and SNF admission or after the SNF discharge within the 30-day risk window fails to allow for a medically complex patient that is treated in an IRF and readmitted to the SNF within 30 days for a condition that may initially have been treated as a comorbidity. We disagree with the rationale provided for exclusion, for while the measure assesses readmission rates while accounting for patient demographics, principal diagnosis in the prior hospitalization, comorbidities, and other patient factors, often, this may not be the reason for admission to a SNF. ARN believes the measure should include the principal diagnosis during the prior proximal hospitalization, comorbidities based on the secondary medical diagnoses listed on the patient’s prior proximal hospital claim and diagnoses from prior hospitalizations that occurred in the previous 365 days, length of stay during the patient’s prior proximal hospitalization, length of stay in the intensive care unit (ICU), body system specific surgical indicators, End-Stage Renal Disease (ESRD) status, whether the patient was disabled, and the number of prior hospitalizations in the previous 365 days. It also would be beneficial to understand the comorbidities being evaluated in the risk-adjustment model. ARN urges the Centers for Medicare and Medicaid Services (CMS) to develop of a list of comorbidities, comparable to the IRF Prospective Payment System (PPS) list of comorbidities and we encourage CMS to categorize an intervening admission to an IRF as a proximal hospitalization. Additionally, ARN has serious concerns with CMS’s proposal to require PAC providers to utilize Medicare claims data to calculate their 30-day readmission rates. Using claims data to calculate readmission rates is difficult for health care providers, as claims data are cumbersome to use and access. Employing a 30-day readmission rate measure will not provide meaningful insight or have an impact on quality improvement efforts if the PAC settings do not have unrestricted access to the data. Further, ARN believes that patients who have been discharged to the community and expire within the post-discharge window should not be included within the quality measure, given the variation in patient characteristics across the four settings. For example, as compared to all Medicare beneficiaries, the SNF and LTCH patient populations represent the most disabled, elderly, and frail beneficiaries. The Medicare Payment Advisory Commission’s (MedPAC) March 2015 Report to Congress found that compared with other beneficiaries, “SNF users are older, frailer, and disproportionately female, disabled, living in an institution, and dually eligible for both Medicare and Medicaid.” Moreover, as compared with all Medicare beneficiaries, those admitted to LTCHs are “disproportionately disabled (under age 65), over age 85, or diagnosed with [ESRD]. They are also more likely to be African American.” ARN urges CMS to exclude patients that pass away within the post-discharge window after being discharged to the community from the discharge to community quality measure, as the types of patients treated in each setting greatly varies and can lead to an inaccurate reflection of the quality of care. ARN is pleased the CMS has proposed discharge measure exclusions; however, we have concerns with the proposed exclusion of post-acute stays that end in transfer to the same level of care, and specifically, CMS’s proposal to include only the final post-acute provider in the discharge to community measure. The Agency’s proposed exclusion criteria fails to consider when a patient’s “home” is a custodial nursing facility and the patient’s post-acute episode involves a discharge back to his or her “home.” In such circumstances, including the final post-acute provider in the discharge to community measure when a patient is discharged to the originating level of care, but in essence, is returning home, may distort the findings of the quality measure. We encourage CMS to design a quality measure that is capable of capturing the difference between a patient’s return to his or her home and a patient’s post-acute episode that involves transfer to the same level of care. (Submitted by: Association of Rehabilitation Nurses)
  
  
• LTCHs are highly specialized acute care facilities that treat complex and often critically ill patients who require care for an extended time. As such, an LTCH must meet Medicare’s conditions of participation for acute care hospitals and have an average length of stay of more than 25 days. Like short-term care hospitals (STCHs), LTCHs treat patients requiring critical, acute, or sub-acute levels of care1 and discharge patients that no longer require such high levels of care. As acute care hospitals, LTCHs’ goal is to discharge patients to the appropriate care setting when they no longer need treatment at the acute care level; their goal is not to keep the patient until they are ready to be discharged to community. As a result, the discharge to community measure as currently constructed is not an appropriate measure of quality for the LTCH setting. After successful LTCH and STCH care, some patients are discharged to lower levels of care such as skilled nursing facilities. The discharge to community measure would wrongly treat discharges to lower, non-acute care settings as unfavorable outcomes although these discharges are favorable outcomes from the perspective of patients and LTCHs, which are to discharge patients at the point where acute care is no longer needed. Just as the discharge to community measure is not used an indicator of care quality for STCHs, it should not be used as a quality measure for LTCHs. While we understand the IMPACT Act requires the development of a discharge to community measure, we urge the measure development team consider the role of LTCHs as acute care hospitals and take a broader perspective that includes discharges to lower acuity settings and/or discharges to home after a period of time. Clearly, more time is needed to develop and test an appropriate measure that does not produce incentives to send patients to the community before it is medically appropriate to do so. (Submitted by: National Association of Long Term Hospitals)
  
  

• APTA supports the goal of improving the quality of health care. Physical therapists are committed to providing high-quality, timely care and to the promotion of evidence-based and patient-centered practice. Furthermore, APTA feels that it is essential that we move towards a common set of quality measures across the continuum of care. APTA supports the implementation of a discharge to community measure across the care settings. We believe that successful transitions to the community following discharge from the respective post-acute care settings will decrease potentially preventable readmissions. However, the APTA does have some concerns regarding the proposed measure methodology. These concerns are discussed below. APTA believes that a patient’s level of function does impact a patients ability to transition successfully back to the community. Recent evidence indicates that patient function is associated with increased risk of 30-day all-cause hospital readmissions and may be an important factor in preventing readmissions for Medicare seniors that is not currently accounted for in measure methodologies1. APTA was pleased to see “activity of daily living” scores in the home health setting included in the risk adjustment methodology for the readmissions measures, and we recently commented, encouraging the use of patient function in the risk adjustment methodology for the post-acute care setting readmissions measures. We believe that readmissions and discharge to community are closely related measures and that patient function may also be an important risk adjustment variable for discharge to community. APTA appreciates that CMS has strict deadlines for the implementation of measures under the IMPACT act, however, as these measures will be new to the respective post-acute care settings, we encourage that settings have the ability to review this data as early as possible in order to understand and, more importantly, so that the respective setting have time to implement strategies to decrease readmissions where necessary. As many of these settings do not always receive feedback on the readmissions of their patients post-discharge, this data will be new to many facilities. APTA recognizes that the overall goal of IMPACT is for PAC providers (HH, IRF, SNF and LTCH) to collect and report standardized and interoperable patient assessment data, quality and resource use measures. We acknowledge that during the initial IMPACT implementation years that there will be a transition period which will include the addition of new measures into all of the post-acute care settings. We believe that achieving a standardized and interoperable patient assessment data set and stable quality measures as quickly as possible will allow for better cross-setting comparisons as well as the evolution of better quality measures with uniform risk standardization, thus achieving the true aim of IMPACT. 1 Greysen SR, Cenzer IS, Auerbach AD, Covinsky KE. Functional Impairment and Hospital Readmission in Medicare Seniors. JAMA Intern Med. 2015;175(4):559-565. (Submitted by: American Physical Therapy Association)
  
  
• This measure is currently in early development and measure specifications are in preliminary stages of clarification. The proposed application could create inaccuracy in measurement of patient condition and payment. Exclusions have not been defined (i.e. “Short term acute stay”), tested regarding clinical reasonableness (i.e. SNFs stays that end in transfer to another SNF) or verified for CMS capabilities to avoid Provider burden for unnecessary and in appropriate data capture (ie. Patients not continuously enrolled in Part A FFS Medicare for 12 months prior to the SNF admission date and at least 31 days after SNF discharge). This measure needs extensive work to assure accuracy and to meet the purpose for which it was intended. Because there is potential for accuracy this measure should continue to be developed. (Submitted by: Genesis Health Care)
  
  
• We have the following concerns related to this measure: 1. The inclusion of unplanned readmissions in the discharge to community measure development and calculation. Our concerns related to the inclusion of unplanned readmissions in the discharge to community measure development and calculation can be defined as follows: a. Duplication/commingling of quality metrics With the inclusion of unplanned readmissions in the discharge to community measure development and calculation, there is the potential for there to be a strong correlation between the performances of the two measures. In other words, those PAC providers that perform poorly with the unplanned readmission measure are more likely to perform poorly on this measure, while those PAC providers that perform well with the unplanned readmission measure have a better opportunity to perform well on this measure. Take for instance the scenario of two facilities with similar patient populations (age, sex, etc.), where provider A discharges 75% to a community setting with a 5% readmission rate while provider B discharges 72% to a community setting with a 2% readmission rate. Which facility performs better at discharging patients to a community setting? Is the discharge to community measure truly representative of the ability of the post-acute care provider to provide services that get the patient back to a community setting, or is too much emphasis being placed upon the ability of the patient to stay in a community setting? b. Inclusion of a factor that is outside of post-acute care provider control While a post-acute care provider can provide services that impact the ability of a patient to be discharged back to a community setting, whether or not a patient returns to an acute care facility in the 30 days following that PAC discharge can be completely unrelated to the services furnished by the PAC provider. For instance, a patient is discharged home from a post-acute care provider but is re-admitted to acute care 10 days later for pneumonia (which is listed as a diagnosis category that cannot be considered planned). Should the post-acute care provider’s discharge to community percentage be penalized if the readmission is completely unrelated to the services provided within post-acute care? With the concerns noted above, we would recommend that the measure developers remove the unplanned readmission component from the discharge to community measurement, and provide side-by-side comparison of these two separate and unique quality measures. 2. Potential for site-specific risk adjustment for a “cross-setting” measure. While we note that the risk adjustment variables are “under consideration”, we are very concerned that the variables noted in the draft specification designate certain variables as being applicable to unique or specific post-acute care providers. The IMPACT Act requires the specification of quality measures and resource use metrics that are standardized and interoperable across PAC settings. In creating a measure where consideration is being given to risk adjustment variables that differ by post-acute care setting, is CMS truly meeting the definition of “standardized and interoperable”? As an example, “Activities of Daily Living” (ADLs) are noted as one of the Clinical conditions variables, yet is noted for “the HHA setting only”. If the measure developer can show that ADLs can be used to differentiate the performance within the HHA population, couldn’t ADLs also be used to differentiate the performance of all post-acute care settings and produce a measurement value that is “standardized and interoperable”? We urge CMS and the measure developers to introduce risk-adjustment variables that are “standardized and interoperable” in order to meet the IMPACT Act requirements. (Submitted by: UDSMR)
  
  
• AHCA is pleased to have the opportunity to comment on the List of Measures under Consideration for December 1, 2015”. The American Health Care Association (AHCA) represents more than 12,000 non-profit and proprietary skilled nursing centers and assisted living communities. Rather than having many of our individual members who have contacted us with comments and to more efficiently provide NQF with feedback, we have received from our various members, we have summarized their comments here. Thank you again for the opportunity to provide these comments. Please contact me at DGifford@ahca.org with any questions or for additional information. A preliminary draft yet incomplete specifications for this measure were provided by CMS for public comment, with the deadline for comment being December 8th 2015, which is after the MUC list was published. No information on reliability or validity of the measure is available and the risk adjustment methodology is not fully specified. Therefore, we believe this measure does not meet NQF MAP criteria for endorsement but rather should receive a vote of “encourage continued development”. Additionally, based on information available, we have the following concerns with measure MUC15-462. 1. The name of the measure should reflect the limited population to which it applies – fee-for service (FFS) Medicare beneficiaries. Since in many states, 40% or more of Medicare beneficiaries are enrolled in Medicare Advantage (MA) plans and for SNFs over half of SNF admissions and discharges are not enrolled in FFS Medicare, this measure may not reflect a SNF’s true discharge to community rate. 2. Since the determination of discharge to the community is self-reported on the claims, we suggest the use of PAC assessment tools as the source of data on discharge to the community, which will allow the inclusion of additional Medicare Beneficiaries enrolled in MA plans. The use of the PAC assessment instruments is consistent with the intent of IMPACT act. 3. The measure does not exclude individuals admitted to a SNF for Part A services but who have an expressed goal to remain in the SNF for long term care and never be discharged back to the community. In fact, when we have compared discharge to community rates as a relationship to the turnover of patients in Part A beds, those SNFs with low turnover that are consistent with being primarily long term care facilities have very low discharge to community rates while SNFs with higher turnover of beds, have higher discharge to community rates. 4. The denominator definition places SNFs at a disadvantage when comparing discharges to community for other PAC providers and the exclusions need to be modified: a. By including any admissions to SNFs, IRFs or LTCHs within 30 days post hospital discharge, more complex patients who are unlikely to be discharged home are added to SNFs, IRFs and LTCHs, but this differentially impacts SNFs since a large proportion of IRF and LTCH patients are discharged to SNFs within 30 days. b. For home health, by not requiring a hospital stay prior to HH stay, increases the number of individuals in their sample who are less likely to be institutionalized after HH services, making their discharge to the community measure look better than the other PAC settings simply due to the denominator definition. c. Failure to exclude individuals admitted to a SNF who were residents of a SNF prior to the hospital stay also is inappropriate since they are residents of long term care facility and would not be expected to be discharged to the community following Part A services. There are approximately, 1.1M long term residents of which nearly 1/6th are admitted to the hospital and sent back the SNF each year. 5. The specifications are based on ICD-9 but all providers as of October 2015 are required to use ICD-10 and no cross walk with ICD-10 is provided, yet the measure will be used during time periods when only ICD-10 data is available. 6. The numerator definition and methodology of calculating a “predicted actual” is extremely confusing, which makes the data less likely to be used and is of questionable benefit. Data showing how this approach is superior to using an actual rate divided by the expected rate needs to be provided. 7. The numerator time window for how long a PAC provider has to accomplish a discharge to the community is not specified. 8. The numerator definition of alive 30 days after discharge to the community will discourage discharging individuals to their home who are dying. We appreciate that those who die in the next 30 days after discharge who are also enrolled in hospice will be exclude but believe these should not be excluded. Rather, we believe they should be counted as successful discharge to the community rather than excluded. 9. The numerator should not count individuals who during the 30 day discharge window are admitted to a SNF just as any other rehospitalizations are not counted. 10. Risk adjustment is not specified other than to state “under consideration is a hierarchical logistic regression model” without any specification of the risk adjustment variables other than examples and categories. Risk adjustment needs to include and specify: a. Social Demographic Characteristics (SDS). b. Functional status (ADL, mobility, self-care, and cognitive function) are some of the strongest predictors of successful discharge to the community but are not included. The risk adjustment variables are not specifically specified with respect to data source and coding but are given as examples with just an overall descriptor. Without knowing the risk adjustment variables and how they are specified, it is hard to evaluate the proposed measures. c. Risk adjustment variables under consideration are not consistent across all PAC settings which is understandable for some settings and patients however, since LTCHs and IRFs are not in all markets, SNFs in many parts of the country serve the same population as IRFs and LTCHs. For example i. Ventilator use is only listed for LTCH setting. SNFs also care for individuals using ventilators. ii. Case mix groups is only listed for LTCH setting but should apply to IRFs and SNFs as well. (Submitted by: American Health Care Association )
  
  
• This measure is limited to FFS beneficaries only, therefore is misleading as a "measure of return to community". It also does not recognize those individuals who resided in the long stay NH PRIOR to their hospital stay and SNF services, who thus plan to stay. Risk adjustment, for chronic conditions, functional status and SES are critical in order for the measure to have any meaning, given the diverse populations served in the SNF setting (Submitted by: Leadingage)
  
  
• The ultimate goals of post-acute care are avoiding institutionalization and returning patients to their previous level of independence and functioning, with discharge to community being the primary goal for the majority of post-acute patients. For many, home is a symbol of independence, privacy, and competence. Discharge to community is considered a valuable outcome to measure because it is a multifaceted measure that captures the patient’s functional status, cognitive capacity, physical ability, and availability of social support at home. There is considerable variation in discharge to community rates within and across post-acute settings. Studies show geographic variation, variation across patient socioeconomic characteristics (for example, race and ethnicity), and variation by facility characteristics (for profit vs. nonprofit, freestanding vs. hospital-based, urban vs. rural). In the IRF setting, discharge to community rates vary across providers, ranging from about 60% to 75%. The 2015 MedPAC report shows that, in FY 2013, the facility-level, mean risk-adjusted discharge to community rate for IRFs within 100 days of admission was 75.8%, and the mean observed rate was 74.7%. Discharge to community rates also vary widely in the SNF setting, ranging from as low as 31% to as high as 65%. The 2015 MedPAC report shows a mean risk-adjusted discharge to community rate of 37.5% for SNFs within 100 days of admission, and mean observed rate of 40.1%. A multicenter study of 23 LTCHs reported that only 28.8% of 1,061 patients who were ventilator-dependent on admission were discharged to home or assisted living facility. A study of 66,510 Medicare beneficiaries during pre- and post-HH episodes, revealed that 64 percent of beneficiaries discharged from HH did not use any other Medicare-reimbursed acute or post-acute services in the 30 days following HH discharge. Significant numbers of patients were admitted to inpatient facilities (29percent) and lesser numbers to skilled nursing facilities (7.6 percent), inpatient rehabilitation (1.5 percent) and home health (7.2 percent) or hospice (3.3 percent) within 30 days of HH discharge (Wolff et al., 2008). II. Feedback Overall, there is not enough detail on the measure specifications to provide adequate feedback. However, despite the lack of detail, a response is included below on the discharge to community measure for LTCH, IRF, SNF, and HH. Description Patients discharged to the community following PAC stay/episode and do not have an unplanned admission to an acute care hospital or LTCH in 31 days after community discharge and remain alive • Although not explicitly stated, it appears that this measure is only holding the last PAC setting responsible for the discharge? Given what is known about multiple sites of PAC use either from LTCH/IRF patients are often discharged to SNF before community discharge, as it is currently worded this measure will penalize the SNF for the more difficult patients that could not be discharged from the prior PAC setting. • Research has shown that PAC patients re-enter the healthcare system through multiple doors after the PAC community discharge why are these not captured in the measure as a failed transition I recommend that the numerator should not include individuals who were admitted to a SNF after PAC community discharge during the 30-day measure window. Numerator & Denominator There is not enough detail to provide feedback on the risk adjustment of the numerator. Without details on the variables included, the origin of those variables, and methods of risk adjustment it is difficult to provide feedback. Recommend providing further detail. Recommend including functional status, functional cognition, presence of social support (e.g., does the individual live alone), and comorbid conditions as risk adjustors Exclusion There is no discussion of excluding long-stay nursing home residents in this measure. Long-stay residents that have an acute medical event, are admitted to an acute care hospital, discharged to PAC for care, and then transitioned back to long-term care should be excluded from the measure. These individuals are not returning to the community, while the nursing home is their home, the effort and processes required for transitioning a long-stay nursing home resident from PAC back to custodial care is very different that transitioning a community- residing older adult back to their residence in the community after a PAC stay. Recommend excluding long stay nursing home residents from the measure. (Submitted by: American Occupational Therapy Association)
  
  
• AMDA is pleased to have the opportunity to comment on the List of Measures under Consideration for December 1, 2015. AMDA - The Society for Post-Acute and Long-Term Care Medicine (AMDA) represents over 5,500 physician medical directors, attending physicians, and other practitioners who work in post-acute and long-term care settings. Thank you again for the opportunity to provide these comments. Please contact me at abardakh@amda.com with any questions or for additional information. A preliminary draft yet incomplete specifications for this measure were provided by CMS for public comment, with the deadline for comment being December 8th 2015, which is after the MUC list was published. No information on reliability or validity of the measure is available and the risk adjustment methodology is not fully specified. Therefore, we believe this measure does not meet NQF MAP criteria for endorsement but rather should receive a vote of “encourage continued development” Based on information available, we have the following concerns with measure MUC15-462. 1. The name of the measure should reflect the limited population to which it applies – fee-for service (FFS) Medicare beneficiaries. Since in many states, 40% or more of Medicare beneficiaries are enrolled in Medicare Advantage (MA) plans and for SNFs over half of SNF admissions and discharges are not enrolled in FFS Medicare, this measure may not reflect a SNF’s true discharge to community rate. 2. Since the determination of discharge to the community is self-reported on the claims, we suggest the use of PAC assessment tools as the source of data on discharge to the community, which will allow the inclusion of additional Medicare Beneficiaries enrolled in MA plans. The use of the PAC assessment instruments is consistent with the intent of IMPACT act. 3. The measure does not exclude individuals admitted to a SNF for Part A services but who have an expressed goal to remain in the SNF for long term care and never be discharged back to the community. In fact, when we have compared discharge to community rates as a relationship to the turnover of patients in Part A beds, those SNFs with low turnover that are consistent with being a primary long term care facilities have very low discharge to community rates while SNFs with higher turnover of beds, have higher discharge to community rates. 4. The denominator definition places SNFs at a disadvantage when comparing discharges to community for other PAC providers and the exclusions need to be modified: a. By including any admissions to SNFs, IRFs or LTCHs within 30 days post hospital discharge, more complex patients who are unlikely to be discharged home are added to SNFs, IRFs and LTCHs, but this differentially impacts SNFs since a large proportion of IRF and LTCH patients are discharged to SNFs within 30 days. b. For home health, by not requiring a hospital stay prior to HH stay, also increases individuals in their sample who are less likely to be institutionalized after HH services, making their discharge to the community measure look better than the other PAC settings simply due to the denominator definition. c. Failure to exclude individuals admitted to a SNF who were residents of a SNF prior to the hospital stay also is inappropriate since they are residents of long term care facility and would not be expected to be discharged to the community following Part A services. There are approximately, 1.1M long term residents of which nearly 1/6th are admitted to the hospital and sent back the SNF each year. 5. The specifications are based on ICD-9 but all providers as of October 2015 are required to use ICD-10 and no cross walk with ICD-10 is provided, yet the measure will be used during time periods when only ICD-10 data is available. 6. The numerator definition and methodology of calculating a “predicted actual” is extremely confusing, which makes the data less likely to be used and is of questionable benefit. Data showing how this approach is superior to using an actual rate divided by the expected rate needs to be provided. 7. The numerator time window for how long a PAC provider has to accomplish a discharge to the community is not specified. 8. The numerator definition of alive 30 days after discharge to the community will discourage discharging individuals to their home who are dying. Many patients with terminal diagnoses are reluctant to enroll in hospice programs or receive palliative care. We appreciate that those who die in the next 30 days after discharge who are also enrolled in hospice will be excluded, but believe these should not be excluded. Rather, we believe they should be counted as successful discharge to the community rather than excluded. 9. The numerator should not count individuals who during the 30 day discharge window are admitted to a SNF just as any other rehospitalizations are not counted. 10. Risk adjustment is not specified other than to state “under consideration is a hierarchical logistic regression model” without any specification of the risk adjustment variables other than examples and categories. Risk adjustment needs to include and specify: a. Social Demographic Characteristics (SDS). b. Functional status (ADL, mobility, self-care, and cognitive function) are some of the strongest predictors of successful discharge to the community but are not included. The risk adjustment variables are not specifically specified with respect to data source and coding but are given as examples with just an overall descriptor. Without knowing the risk adjustment variables and how they are specified, it is hard to evaluate the proposed measures. c. Risk adjustment variables under consideration are not consistent across all PAC settings, which is understandable for some settings and patients, however, since LTCHs and IRFs are not in all markets, SNFs in many parts of the country serve the same population as IRFs and LTCHs. For example: i. Ventilator use is only listed for LTCH setting. SNFs also care for individuals using ventilators. ii. Case mix groups is only listed for LTCH setting but should apply to IRFs and SNFs as well. iii. Risk adjustment for other comorbidities should include stroke, ESRD, COPD, advanced heart failure (high risk for frequent decompensation). (Submitted by: AMDA-The Society for Post-Acute and Long-Term Care Medicine)
  
  
• This measure is currently in the early stage of development, has never been submitted to NQF and thus is not NQF-endorsed (Submitted by: AdvaMed)
  
  
• ARN is supportive of a measure that works to identify unplanned (re)admissions; however, we have concerns with identifying unplanned (re)admissions based on the planned readmissions algorithm used in NQF measure #2510: SNF 30-Day All-Cause Readmission Measure (SNFRM); NQF #2502: All-Cause Unplanned Readmission Measure for 30 Days Post Discharge from IRFs; NQF #2512: All-Cause Unplanned Readmission Measure for 30 Days Post Discharge from LTCHs; and NQF #2380: Re-hospitalization During the First 30 Days of Home Health. The exclusion criteria included within NQF #2510 for SNF stays where the patient had one or more intervening PAC admissions to an IRF that occurred either between the prior proximal hospital discharge and SNF admission or after the SNF discharge within the 30-day risk window fails to allow for a medically complex patient that is treated in an IRF and readmitted to the SNF within 30 days for a condition that may initially have been treated as a comorbidity. We disagree with the rationale provided for exclusion, for while the measure assesses readmission rates while accounting for patient demographics, principal diagnosis in the prior hospitalization, comorbidities, and other patient factors, often, this may not be the reason for admission to a SNF. ARN believes the measure should include the principal diagnosis during the prior proximal hospitalization, comorbidities based on the secondary medical diagnoses listed on the patient’s prior proximal hospital claim and diagnoses from prior hospitalizations that occurred in the previous 365 days, length of stay during the patient’s prior proximal hospitalization, length of stay in the intensive care unit (ICU), body system specific surgical indicators, End-Stage Renal Disease (ESRD) status, whether the patient was disabled, and the number of prior hospitalizations in the previous 365 days. It also would be beneficial to understand the comorbidities being evaluated in the risk-adjustment model. ARN urges the Centers for Medicare and Medicaid Services (CMS) to develop of a list of comorbidities, comparable to the IRF Prospective Payment System (PPS) list of comorbidities and we encourage CMS to categorize an intervening admission to an IRF as a proximal hospitalization. Additionally, ARN has serious concerns with CMS’s proposal to require PAC providers to utilize Medicare claims data to calculate their 30-day readmission rates. Using claims data to calculate readmission rates is difficult for health care providers, as claims data are cumbersome to use and access. Employing a 30-day readmission rate measure will not provide meaningful insight or have an impact on quality improvement efforts if the PAC settings do not have unrestricted access to the data. Further, ARN believes that patients who have been discharged to the community and expire within the post-discharge window should not be included within the quality measure, given the variation in patient characteristics across the four settings. For example, as compared to all Medicare beneficiaries, the SNF and LTCH patient populations represent the most disabled, elderly, and frail beneficiaries. The Medicare Payment Advisory Commission’s (MedPAC) March 2015 Report to Congress found that compared with other beneficiaries, “SNF users are older, frailer, and disproportionately female, disabled, living in an institution, and dually eligible for both Medicare and Medicaid.” Moreover, as compared with all Medicare beneficiaries, those admitted to LTCHs are “disproportionately disabled (under age 65), over age 85, or diagnosed with [ESRD]. They are also more likely to be African American.” ARN urges CMS to exclude patients that pass away within the post-discharge window after being discharged to the community from the discharge to community quality measure, as the types of patients treated in each setting greatly varies and can lead to an inaccurate reflection of the quality of care. ARN is pleased the CMS has proposed discharge measure exclusions; however, we have concerns with the proposed exclusion of post-acute stays that end in transfer to the same level of care, and specifically, CMS’s proposal to include only the final post-acute provider in the discharge to community measure. The Agency’s proposed exclusion criteria fails to consider when a patient’s “home” is a custodial nursing facility and the patient’s post-acute episode involves a discharge back to his or her “home.” In such circumstances, including the final post-acute provider in the discharge to community measure when a patient is discharged to the originating level of care, but in essence, is returning home, may distort the findings of the quality measure. We encourage CMS to design a quality measure that is capable of capturing the difference between a patient’s return to his or her home and a patient’s post-acute episode that involves transfer to the same level of care. (Submitted by: Association of Rehabilitation Nurses)
  
  

• APTA supports the goal of improving the quality of health care. Physical therapists are committed to providing high-quality, timely care and to the promotion of evidence-based and patient-centered practice. Furthermore, APTA feels that it is essential that we move towards a common set of quality measures across the across the continuum of care. APTA supports the implementation of readmissions measures across the care settings, as approximately 20% of all Medicare patients are readmitted within 30 days of an acute care discharge and readmissions account for an estimated $17 billion in health care spending. APTA is pleased to see that the draft specifications for these measures align with existing methodologies of other readmissions measures. We believe that potentially preventable readmissions measures will focus providers on those patients who are expected to have successful transitions to the community follow in discharge from the respective post-acute care settings. However, the APTA does have some concerns regarding the proposed measure methodology. These concerns are discussed below. APTA believes that a patient’s level of function does impact the potential for readmissions. Recent evidence indicates that patient function is associated with increased risk of 30-day all-cause hospital readmissions and may be an important factor in preventing readmissions for Medicare seniors that is not currently accounted for in measure methodologies1. APTA is pleased to see “activity of daily living” scores included in the risk adjustment methodology, however, we note that this is only included for those patients in the home health setting. We would encourage measure developers to include this as a risk adjustment variable in all readmissions measures. APTA appreciates that CMS has strict deadlines for the implementation of measures under the IMPACT act, however, as these measures will be new to the respective post-acute care settings we encourage that settings have the ability to review this data as early as possible in order to understand the data and, more importantly, so that the respective setting have time to implement strategies to decrease readmissions where necessary. As many of these settings do not always receive feedback on the readmissions of their patients post-discharge, this data will be new to many facilities. Additionally, skilled nursing facilities and inpatient rehabilitation facilities will be trying to manage two measures, one within stay, and one post-discharge. APTA recognizes that the overall goal of IMPACT is for PAC providers (HH, IRF, SNF and LTCH) to collect and report standardized and interoperable patient assessment data, quality and resource use measures. We acknowledge that during the initial IMPACT implementation years that there will be a transition period which will include the addition of new measures into all of the post-acute care settings. We believe that achieving a standardized and interoperable patient assessment data set and stable quality measures as quickly as possible will allow for better cross-setting comparisons as well as the evolution of better quality measures with uniform risk standardization, thus achieving the true aim of IMPACT. 1 Greysen SR, Cenzer IS, Auerbach AD, Covinsky KE. Functional Impairment and Hospital Readmission in Medicare Seniors. JAMA Intern Med. 2015;175(4):559-565. (Submitted by: American Physical Therapy Association)
  
  
• This measure is currently in early development and measure specifications are in preliminary stages of clarification. The proposed application could create inaccuracy in measurement of patient condition and payment. Exclusions have not been defined for clinical reasonableness as well as payment logic (i.e. “Patients who did not have a short term acute-care stay within 30 days prior to the SNF admission” - Does this mean all patients must have a hospitalization prior to SNF admission to qualify for Part A benefits? If so, why? How does this overlap with other CMS programs where beneficiaries are not required to have a qualifying hospital stay such as the year two CJR program? ) nor have exclusions been verified for CMS capabilities to avoid Provider burden for unnecessary and in appropriate data capture (ie. Patients not continuously enrolled in Part A FFS Medicare for 12 months prior to the SNF admission date and at least 31 days after SNF discharge – How will CMS assure Providers do not capture unneeded and/or inappropriate patient information?). This measure needs extensive work to assure accuracy and to meet the purpose for which it was intended. Because there is potential for accuracy this measure should continue to be developed. (Submitted by: Genesis Health Care)
  
  
• We would like to note the following concerns regarding the development of these measures: 1. Measure development for the IMPACT Act appears to create 4 separate measures that are site specific with their own set of criteria and risk adjustment factors. The IMPACT Act aims to create quality measurement within Post-Acute Care (PAC) that is "standardized and interoperable". By developing measures which differ in their calculation, such as site-specific inclusion/exclusion criteria or risk-adjustment factors, PAC sites will be subjected to comparisons of quality that do not differ based upon the quality of care provided, but rather the differences inherent in the measure calculations. We strongly recommend that CMS and the measure developers produce a measure that is calculated in a "standardized and interoperable" manner with inclusion/exclusion criteria and risk-adjustment factors the are applied consistently across all PAC sites. 2. The measures developed for the IMPACT Act hold PAC providers responsible for circumstances that occur outside of their control. While PAC providers are responsible for providing care that allows patients safe transitions to their next setting or home, holding these providers accountable fora time period in which they are not furnishing care should not differentiate the quality of care they actually provided. In Appendix A, Table A1 lists various conditions that define potentially preventable hospital readmission for 30-days post-PAC discharge. In this list, conditions such as Asthma, Congestive Heart Failure, and Dehydration are listed as a reason for a potentially preventable readmission. So if the patient fails to care for themselves post-discharge (against the advice and discharge instructions provided by a PAC provider) and requires readmission to Acute Care for one of these conditions, the PAC provider is penalized even though the patient is no longer being cared for by the PAC provider. We strongly recommend that CMS and the measure developers produce a measure that represents the quality of care provided by PAC providers while the patient is in their care. (Submitted by: UDSMR)
  
  
• AHCA is pleased to have the opportunity to comment on the List of Measures under Consideration for December 1, 2015”. The American Health Care Association (AHCA) represents more than 12,000 non-profit and proprietary skilled nursing centers and assisted living communities. Rather than having many of our individual members who have contacted us with comments and to more efficiently provide NQF with feedback, we have received from our various members, we have summarized their comments here. Thank you again for the opportunity to provide these comments. Please contact me at DGifford@ahca.org with any questions or for additional information. A preliminary draft yet incomplete specifications for this measure were provided by CMS for public comment, with the deadline for comment being December 1st, 2015, which is after the MUC list was published. No information on reliability or validity of the measure is available and the risk adjustment methodology is not fully specified. In addition, CMS has provided inadequate time for public review and comment of this measure. In fact, the TEP never saw the preliminary full draft specification of this measure until it was made available for public comment in late November 2015 after the TEP meeting. Therefore, for these reasons and our concerns detailed below, we believe this measure does not meet NQF MAP criteria for endorsement but rather should receive a vote of “encourage continued development” This measure (MUC15-495) double counts readmissions with other SNF potentially preventable readmission measure (MUC15-1048). Measure MUC15-1048 includes readmission that occur within a SNF stay and after a SNF stay. Thus, this measure will capture the same readmissions as in measure MUC15-1048 resulting in a double counting of measures. We have additional concerns with this measure that mirror our comments on measure MUC15-1048 which include: 1. The name of the measure should reflect the limited population to which it applies – fee-for service (FFS) Medicare beneficiaries; particularly since in many states 40% or more of Medicare beneficiaries are enrolled in MA plans and for SNFs over half of SNF admissions and discharges are not enrolled in FFS Medicare. 2. The specifications are based on ICD-9 but all providers as of October 2015 are required to use ICD-10 and no cross walk with ICD-10 is provided. 3. The numerator definition and methodology of calculating a “predicted actual” is extremely confusing, which makes the data less likely to be used and is of questionable benefit. Data showing how this approach is superior to using an actual rate divided by the expected rate needs to be provided. 4. Many of the potentially preventable admissions contains diagnoses that do not meet the proposed definition of potentially preventable readmissions; therefore, these should be dropped. 5. The process for developing the list of potentially preventable admissions used existing literature but would have benefited from a more formal process such as the RAND appropriateness rating system (e.g. modified Delphi approach) rather than ad hoc TEP and consultant experts. 6. Risk adjustment does not included: a. SES characteristics, an issue identified in the last round of NQF readmission measure reviews as a requirement. b. Functional status, one of the strongest predictors of hospitalizations. All the PAC settings are now required to utilize standard functional status assessment – Section GG from the CARE tool. c. The risk adjustment variables are not specifically specified with respect to data source and categories but are given as examples with just a descriptor. For example is age continuous or categorical variable and if categorical how and from what source? d. Prior utilization variables indicate they “vary by measure” but how they vary by measure is not provided. 7. The exclusions are confusing and need some modifications. a. The denominator exclusions appear confusing and overlapping so that it is unclear if discharges from IRF or LTCH to SNF or HH are included in the IRF and LTCH denominator. Similarly, if SNF discharges to HH are included. b. We agree that AMA discharges from PAC provider should be excluded but so should hospital discharges that are AMA but end up in PAC provider. 8. The list of related to other NQF endorsed measures is incomplete and only compares to CMS or RTI endorsed measures. All other NQF endorsed measures should be included. 9. We agree with the expansion of unplanned readmission diagnoses from YALE’s list, but will this also be applied to the SNFRM and hospital readmission measures? (Submitted by: American Health Care Association )
  
  
• This measure seems to be in overlap with MUC15-1048 as it captures the SAME readmissions. As with many of the other measures, it should be acknowledged that this measure only addresses FFS. Many SNF providers have a significant % of their admissions and discharges under managed care, thus making this a far less useful measure of performance. Also, there needs to be risk adjustment for clinical complexity, functional status and SES. (Submitted by: LeadingAge)
  
  
• This measure will encourage post-acute care providers to optimize care for patients before discharge. (Submitted by: Smith & Nephew )
  
  
• ARN is pleased CMS has proposed discharge measure exclusions; however, we disagree with the proposed exclusion criteria of patients less than 18 years old. Many IRFs treat patients younger than 21 when necessary, is reflected by the Functional Independence Measure (FIM) and IRF-Patient Assessment Instrument (PAI), both of which are used to assess patients age seven or older. ARN encourages CMS not to exclude patients under 18 years old from the discharge measures. We also recommend that CMS ensure the data collection time frame is the same for all PAC settings. Currently, the time frame for the initial data collection for the project varies from one year (SNFs), two years (IRFs and LTCHs), and three years (HHAs), we urge the Agency to implement one standard reporting time frame across PAC providers. While reporting may be based on either one year of data, two years of data, etc., we strongly recommend CMS and its subcontractors address the inconsistent reporting periods. Further, ARN questions the rationale behind the exclusion for SNF stays where the patient had one or more intervening PAC admissions which occurred either between the prior proximal hospital discharge and SNF admission or after the SNF discharge, within the 30-day risk window as well as SNF admissions where the patient had multiple SNF admissions after the prior proximal hospitalization, within the 30-day risk window. The rationale states that “when patients have multiple PAC admissions, evaluating quality of care coordination is confounded and even controversial in terms of attributing responsibility for a readmission among multiple PAC providers. Similarly, assigning responsibility for a readmission for patients who have multiple SNF admissions subsequent to their prior proximal hospitalization is also controversial.” ARN believes that this rationale could apply to any PAC setting and therefore, disagrees with having this as an exclusion from the SNF denominator. We also have concerns with the exclusion criteria of SNF stays with a gap of greater than one day between discharge from the prior proximal hospitalization and admission to a SNF. As currently outlined, the exclusion criteria fails to consider a medically complex patient that is treated in an IRF and subsequently readmitted within 30 days for an issue that may have been treated as a comorbidity. Given that a prior proximal hospitalization is defined as an inpatient admission to an acute care hospital, critical access hospital (CAH), or a psychiatric hospital, and IRFs are licensed as hospitals, we believe that admission to an IRF should be considered a proximal hospitalization and disagree that patients who are clinically different should be excluded. Moreover, in regards to the within-stay criteria, IRFs have experienced a rise in the number of patients who must return to the acute inpatient care setting within 48-72 hours of admission due to the disparity between the level of care which their condition (either a co-morbidity or complication secondary to the presenting diagnosis) requires and the level of care that an IRF is able to provide. An example of this is leukemia patients as many times they require inpatient rehabilitation to discharge them from the hospital and that the patient’s survival may be related to their functional status and rehabilitation can improve function and symptoms such as fatigue, well-being, and pain. To this end, ARN disagrees with the within-stay criteria delineated for IRFs and believes that in specific instances, readmissions are a necessity for patient safety, and not necessarily preventable. Finally, while ARN is generally supportive of the potentially preventable readmissions measure specifications, we have several concerns. As ARN has stated in previous comment letters, the IRF measure is based on 24 months of data while the SNF measure is based on 12 months of data. PAC facilities should not be penalized for conditions that prompt readmissions that are unrelated to the patient’s initial reason for admission. Moreover, as previously stated, we oppose CMS’s proposal to require PAC providers to utilize 30-day readmission claims data to determine their readmission rates. Using claims data to calculate readmission rates will be difficult for IRFs and other PAC settings, as claims data are cumbersome to use and access. Employing a 30-day readmission rate measure will not provide meaningful insight or have an impact on quality improvement efforts if PAC settings do not have unrestricted access to the data. (Submitted by: Association of Rehabilitation Nurses)
  
  
• AMDA is pleased to have the opportunity to comment on the List of Measures under Consideration for December 1, 2015”. AMDA - The Society for Post-Acute and Long-Term Care Medicine (AMDA) represents over 5,500 physician medical directors, attending physicians, and other practitioners who work in post-acute and long-term care settings. Thank you again for the opportunity to provide these comments. Please contact me at abardakh@amda.com with any questions or for additional information. A preliminary draft yet incomplete specifications for this measure were provided by CMS for public comment, with the deadline for comment being December 1st, 2015, which is after the MUC list was published. No information on reliability or validity of the measure is available and the risk adjustment methodology is not fully specified. In addition, CMS has provided inadequate time for public review and comment of this measure. In fact, the TEP never saw the preliminary draft full specification of this measure until it was made available for public comment in late November 2105 after the TEP meeting. Therefore, for these reasons and our concerns detailed below, we believe this measure does not meet NQF MAP criteria for endorsement but rather should receive a vote of “encourage continued development” This measure (MUC15-495) double counts readmissions with other SNF potentially preventable readmission measure (MUC15-1048). Measure MUC15-1048 includes readmission that occur within a SNF stay and after a SNF stay. Thus, this measure will capture the same readmissions as in measure MUC15-1048 resulting in a double counting of measures. We have additional concerns with this measure that mirror our comments on measure MUC15-1048 which include: 1. The name of the measure should reflect the limited population to which it applies – fee-for service (FFS) Medicare beneficiaries; particularly since in many states 40% or more of Medicare beneficiaries are enrolled in MA plans and for SNFs over half of SNF admissions and discharges are not enrolled in FFS Medicare. 2. The specifications are based on ICD-9 but all providers as of October 2015 are required to use ICD-10 and no cross walk with ICD-10 is provided. 3. The numerator definition and methodology of calculating a “predicted actual” is extremely confusing, which makes the data less likely to be used and is of questionable benefit. Data showing how this approach is superior to using an actual rate divided by the expected rate needs to be provided. 4. Many of the potentially preventable admissions contains diagnoses that do not meet the proposed definition of potentially preventable readmissions; therefore, these should be dropped. 5. The process for developing the list of potentially preventable admissions used existing literature but would have benefited from a more formal process such as the RAND appropriateness rating system (e.g. modified Delphi approach) rather than ad hoc TEP and consultant experts. 6. Risk adjustment does not included: a. SES characteristics, an issue identified in the last round of NQF readmission measure reviews as a requirement. b. Functional status, one of the strongest predictors of hospitalizations. All the PAC settings are now required to utilize standard functional status assessment – Section GG from the CARE tool. c. The risk adjustment variables are not specifically specified with respect to data source and categories but are given as examples with just a descriptor. For example is age continuous or categorical variable and if categorical how and from what source? d. Prior utilization variables indicate they “vary by measure” but how they vary by measure is not provided. 7. The exclusions are confusing and need some modifications. a. The denominator exclusions appear confusing and overlapping so that it is unclear if discharges from IRF or LTCH to SNF or HH are included in the IRF and LTCH denominator. Similarly, if SNF discharges to HH are included. b. We agree that AMA discharges from PAC provider should be excluded but so should hospital discharges that are AMA but end up in PAC provider. 8. The list of related to other NQF endorsed measures is incomplete and only compares to CMS or RTI endorsed measures. All other NQF endorsed measures should be included. 9. We agree with the expansion of unplanned readmission diagnoses from YALE’s list, but will this also be applied to the SNFRM and hospital readmission measures? (Submitted by: AMDA-The Society for Post-Acute and Long-Term Care Medicine)
  
  
• PointRight Inc., the leading provider of web-based analytic services to SNFs nationwide, and the developer of the NQF-endorsed PointRight Pro 30 measure of rehospitalizations from SNFs, is pleased to offer comments on the proposed measure of potentially preventable readmissions from SNFs (SNFPPR). We see the potential for enhancements in the reporting of the SNFPPR that would make it much more valuable to providers, payers and consumers, and would like to recommend these enhancements NQF as the developers finalize the specification for their measure. While other commenters have submitted comments on the methodology for calculation of the SNFPPR, this is not our aim here. Instead, we are advocating that the SNFPPR be reported in a distinctive way that will enhance its acceptance, its clinical utility, and ultimately its impact on provider-level and system-level performance improvement. We believe that there will always be legitimate disagreements on optimal methods of defining provider performance measures and adjusting them for clinical risk factors and for measurement error. Our belief is that many such disagreements can be mitigated by making the calculation methodology transparent to the measure’s users, and by reporting the measure in a way that optimizes its integration with complementary measures of the same domain of performance. In the case of the SNFPPS, the complementary NQF-endorsed measure of rehospitalization is the PointRight Pro 30 measure; this measure is based on MDS data rather than on Medicare FFS claims. Our comments here will reference research that PointRight has conducted using MDS data from CY2014, from a national sample of 1,766 Medicare-certified SNFs that were at least 97% complete in reporting the outcomes of their post-acute stays. A more detailed report on that study and its methodology is available to NQF upon request. The SNFPPR has certain characteristics that make it an accurate measure of performance in many circumstances but a potentially misleading one in others. These characteristics are as follows: 1) The SNFPPR is based on Medicare FFS claims. To the extent that a SNF has a relatively large number of post-acute admissions that have other payers, and to the extent that its performance differs by payer, the SNFPPR will be an inaccurate – or at the very least incomplete – measure of a SNF’s performance. In general, post-acute stays of Medicare Advantage and other non-Medicare FFS patients are shorter than Medicare FFS stays, because financial incentives for SNFs are different, and because proactive case management is more common when the payer is at risk. A potential consequence of shorter SNF stays is a higher rate of rehospitalizations after SNF discharge but within the 30 day window. Some SNFs with a high managed care volume are able to reduce these indirect rehospitalizations through intensive management of care transitions and optimal timing of SNF discharge. Other SNFs do not succeed in doing this. The SNFPPR measure does not distinguish these situations because it does not cover non-FFS patients. A health plan that built its preferred provider network or selected post-acute providers based on the SNFPPR alone might make choices that actually led to an increase in its rehospitalization rates. This issue has national scope and significance. For 25% of all SNFs in our national sample the proportion of post-acute admissions that were not Medicare FFS exceeded 29%. The 10% of SNFs with the highest proportion of non-Medicare FFS post-acute admissions had a proportion in excess of 41%. The median 30-day direct rehospitalization rate was 26% for Medicare FFS patients, but only 19% for non-Medicare FFS patients. 2) Research on rehospitalizations across the entire post-acute continuum has shown that issues like functional status, nutritional status and cognitive impairment that are not necessarily captured by coded diagnoses can have a powerful effect on rehospitalization rates. Facilities that treat a frail and debilitated elderly population will look worse on the SNFPPR measure even after risk adjustment, and facilities treating a more vigorous elderly population will look better on the SNFPPR regardless of their true clinical performance. If discharge planners direct patients based on the SNFPPR alone they may send frail patients to SNFs that actually don’t do a good job with such patients, while reducing the census and operating margins of some SNFs that do excel at managing highly vulnerable patients. 3) The SNFPPR’s definition of a rehospitalization as potentially preventable is based on an algorithm that may be more or less applicable to particular patient populations, due to the interaction of non-measured covariates with the claims-based information used in the algorithm. 4) The use of predicted rather than observed numerators can mitigate problems of non-representative ratios when denominators are small, but at the expense of effacing genuine differences between facilities. A facility that persistently has no potentially preventable readmissions, quarter after quarter, has meaningfully different performance than one with an observed rate of 10%, even if the predicted rate of the former facility is 10% when calculated one quarter at a time. Predicted rates are not a substitute for observed rates but rather a complement to them; aggregation of observed rates over time is an alternative approach to dealing with errors related to small denominators. The SNFPPR measure additional three features that limit its value for supporting SNFs’ ongoing quality improvement programs: 5) A facility’s performance on the SNFPPR can only be determined in retrospect. Thus, a facility cannot use the measure’s methodology to prospectively identify patients at high risk, nor can it continuously monitor the measure to quickly see the impact of a change in its care processes. 6) The numerator does not distinguish between rehospitalizations directly from the SNF and rehospitalizations from the community following discharge from the SNF (but within 30 days of the hospital discharge). Separating the two components of the numerator would enable a SNF to know whether it has problems with discharge planning and care transitions that contribute disproportionately to rehospitalizations. This is important because care transitions have traditionally not been a major focus of attention and resource deployment for SNFs. 7) The exclusion of rehospitalizations from the denominator as “planned” is based on a claims-based algorithm that does not necessarily agree with facilities’ own categorization of rehospitalizations as planned. This can make it difficult for a facility to target a change in its number of unplanned rehospitalizations as categorized by the SNFPPR measure. In the PointRight study cited above the proportion of direct rehospitalizations within 30 days that were self-characterized as “planned” had a significantly skewed distribution. (Proportion of rehospitalizations that were self-described as planned: 10th percentile – 0%, 25th percentile – 3.4%, median – 6.8%, 75th percentile – 11.1%, 90th percentile 16.8%) Since the distribution of the same proportion determined by algorithm was not published, the extent of disagreement cannot be determined. However, if the target of quality improvement is reduction of unplanned rehospitalizations, this aim would be better attained if facilities’ determinations of planned status and NQF’s determination of planned status could be aligned. Recommendation: With these considerations in mind PointRight suggests a change in reporting methodology for the SNFPPR measure that will make the measure more transparent and more useful for quality improvement, for payers’, hospitals’ and managed care organizations’ development of post-acute care networks, and planning of care transitions. The recommended changes in reporting will enable users to take full advantage of the contemporaneous reporting of the SNFPPR with other, complementary NQF-endorsed quality measures based on other data sources. The recommended change in reporting would in fact be applicable, and valuable, for other NQF-endorsed rehospitalization measures that are based on Medicare FFS claims and have algorithm-based exclusion criteria. Concurrent interpretation of complementary measures helps identify and make sense of situations where claims-based expected rates may over-adjust or under-adjust for risk and situations were where algorithm-based determinations of planned status or potential preventability do not comport with clinical experience. Finally, the separation of rehospitalizations into direct and indirect readmissions makes the measure much more useful not only to SNFs but also to other users such as hospitals, primary care physicians, payers and consumers. It enables them to know whether a SNF’s primary weakness is with transitions of care from SNF to community – a problem that parties upstream or downstream from the SNF may be able to address proactively. Specifically, PointRight recommends that reporting of the SNFPPR include the following items in addition to each SNF’s predicted rate, expected rate, and adjusted rate of 30-day rehospitalizations of the denominator population over the measurement period: 1) The observed rate of rehospitalizations in the denominator population. 2) The observed rate of rehospitalizations in the denominator population that took place directly from the SNF to the acute care hospital, as opposed to from the community following discharge from the SNF to the community. 3) The percentage of total admissions to the SNF from acute care hospitals (i.e. potential members of the denominator population) that were Medicare FFS. 4) The percentage of Medicare FFS admissions to the SNF from acute care hospitals that was included in the denominator population (i.e. not excluded from the denominator population because of meeting one or more exclusion criteria). 5) The percentage of total rehospitalizations of the denominator population excluded because the hospitalizations were determined by algorithm to be planned. 6) The percentage of total unplanned rehospitalizations of the denominator population excluded from the numerator because the hospitalizations were determined by algorithm to be not potentially preventable. 7) The actual count of the denominator population, if this is permissible under HIPAA. In other words, the additional data to be reported would provide the observed rate (and not just the predicted rate), and would explain the relationship of the measure denominator to the entirety of post-acute admissions, the relationship of the measure numerator to the entirety of all rehospitalizations of the denominator population, and the breakdown of the numerator into direct and indirect rehospitalizations. (Submitted by: PointRight Inc.)
  
  
• At least one NRHI member has expressed concerns regarding this measure. Specifically, clinical staff at HealthInsight have concerns with inclusion of this measure due to the validity of and complicated methodologies used for "risk adjustment for patient characteristics and a statistical estimate of the facility effect beyond patient mix." This methodology is frequently cited as the basis for determining potentially avoidable hospital facility admissions. This methodology is not well-established. (Submitted by: Network for Regional Healthcare Improvement)
  
  

• APTA supports the goal of improving the quality of health care. Physical therapists are committed to providing high-quality, timely care and to the promotion of evidence-based and patient-centered practice. Furthermore, APTA feels that it is essential that we move towards a common set of quality measures across the across the continuum of care. APTA supports the implementation of readmissions measures across the care settings, as approximately 20% of all Medicare patients are readmitted within 30 days of an acute care discharge and readmissions account for an estimated $17 billion in health care spending. APTA is pleased to see that the draft specifications for these measures align with existing methodologies of other readmissions measures. We believe that potentially preventable readmissions measures will focus providers on those patients who are expected to have successful transitions to the community follow in discharge from the respective post-acute care settings. However, the APTA does have some concerns regarding the proposed measure methodology. These concerns are discussed below. APTA believes that a patient’s level of function does impact the potential for readmissions. Recent evidence indicates that patient function is associated with increased risk of 30-day all-cause hospital readmissions and may be an important factor in preventing readmissions for Medicare seniors that is not currently accounted for in measure methodologies1. APTA is pleased to see “activity of daily living” scores included in the risk adjustment methodology, however, we note that this is only included for those patients in the home health setting. We would encourage measure developers to include this as a risk adjustment variable in all readmissions measures. APTA appreciates that CMS has strict deadlines for the implementation of measures under the IMPACT act, however, as these measures will be new to the respective post-acute care settings we encourage that settings have the ability to review this data as early as possible in order to understand the data and, more importantly, so that the respective setting have time to implement strategies to decrease readmissions where necessary. As many of these settings do not always receive feedback on the readmissions of their patients post-discharge, this data will be new to many facilities. Additionally, skilled nursing facilities and inpatient rehabilitation facilities will be trying to manage two measures, one within stay, and one post-discharge. APTA recognizes that the overall goal of IMPACT is for PAC providers (HH, IRF, SNF and LTCH) to collect and report standardized and interoperable patient assessment data, quality and resource use measures. We acknowledge that during the initial IMPACT implementation years that there will be a transition period which will include the addition of new measures into all of the post-acute care settings. We believe that achieving a standardized and interoperable patient assessment data set and stable quality measures as quickly as possible will allow for better cross-setting comparisons as well as the evolution of better quality measures with uniform risk standardization, thus achieving the true aim of IMPACT. 1 Greysen SR, Cenzer IS, Auerbach AD, Covinsky KE. Functional Impairment and Hospital Readmission in Medicare Seniors. JAMA Intern Med. 2015;175(4):559-565. (Submitted by: American Physical Therapy Association)
  
  
• We would like to note the following concerns regarding the development of these measures: 1. Measure development for the IMPACT Act appears to create 4 separate measures that are site specific with their own set of criteria and risk adjustment factors. The IMPACT Act aims to create quality measurement within Post-Acute Care (PAC) that is "standardized and interoperable". By developing measures which differ in their calculation, such as site-specific inclusion/exclusion criteria or risk-adjustment factors, PAC sites will be subjected to comparisons of quality that do not differ based upon the quality of care provided, but rather the differences inherent in the measure calculations. We strongly recommend that CMS and the measure developers produce a measure that is calculated in a "standardized and interoperable" manner with inclusion/exclusion criteria and risk-adjustment factors the are applied consistently across all PAC sites. 2. The measures developed for the IMPACT Act hold PAC providers responsible for circumstances that occur outside of their control. While PAC providers are responsible for providing care that allows patients safe transitions to their next setting or home, holding these providers accountable fora time period in which they are not furnishing care should not differentiate the quality of care they actually provided. In Appendix A, Table A1 lists various conditions that define potentially preventable hospital readmission for 30-days post-PAC discharge. In this list, conditions such as Asthma, Congestive Heart Failure, and Dehydration are listed as a reason for a potentially preventable readmission. So if the patient fails to care for themselves post-discharge (against the advice and discharge instructions provided by a PAC provider) and requires readmission to Acute Care for one of these conditions, the PAC provider is penalized even though the patient is no longer being cared for by the PAC provider. We strongly recommend that CMS and the measure developers produce a measure that represents the quality of care provided by PAC providers while the patient is in their care. (Submitted by: UDSMR)
  
  
• AMRPA has historically supported readmissions measures within a post-discharge window such as 30 days. A similar measure is included in the IRF QRP, an all cause unplanned 30 day post IRF discharge readmission measure. CMS should clarify its intent to replace the existing IRF QRP measure with this new measure once it is adopted. A properly constructed readmission measure has the potential to improve the outcomes for the patients IRFs treat. However, before MUC 496 is adopted, we encourage CMS to modify how it identifies a readmission as potentially preventable. The draft measure specifications state that the potentially preventable readmission (PPR) is based on the acute care hospital diagnosis. But this diagnosis, as represented by the ICD code, is not always the rationale for admission to the IRF. Instead, using data from the Inpatient Rehabilitation Facility Patient Assessment Instrument (IRF PAI) might more accurately demonstrate when a planned readmission to the acute care hospital is appropriate. For example, Patient A is a C6 tetraplegic who is unexpectedly admitted to acute care from the IRF for autonomic dysreflexia. In this scenario, the rationale for the unplanned readmission to the acute care is rarely seen in the acute care hospital but is not uncommon in the IRF setting. Additionally, the readmission might not have been preventable meaning the IRF may be inappropriately penalized for an unplanned readmission. The measure also fails to account for planned readmissions that might be inappropriate. With regard to a PPR, the clinical literature does not support the inclusion of inadequate prophylaxis as a rationale for deeming a readmission as potentially preventable. For example, increasingly chest guidelines are demonstrating the role of deep vein thrombosis and pulmonary embolism prevention in PAC settings. However, the literature is less clear on the ability to prevent stress ulcers in IRFs. Therefore, it cannot be said that inadequate prophylaxis is inextricably linked to PPRs in all cases. We also believe that atrial fibrillation and flutter should be separated for the purposes of PPR. While we appreciate the Agency’s attempt to minimize data collection and reporting burden by using two years of claims data, we believe patient characteristic data including diagnosis, comorbidities and complexities, length of stay, motor and cognitive function, and discharge destination are also critically important to collect. Fortunately the industry already collects such information via the IRF PAI. We also recommend that the risk adjustment methodology be modified to include socioeconomic and sociodemographic factors. Finally, CMS recently closed a public comment process seeking feedback on the draft measure specifications. Given that this is a newly developed measure that has not yet completed the CMS development process, it may be premature to endorse this measure for the purpose of the IRF QRP. (Submitted by: American Medical Rehabilitation Providers Association)
  
  
• This measure will encourage post-acute care providers to optimize care for patients before discharge. (Submitted by: Smith & Nephew )
  
  
• ARN is pleased CMS has proposed discharge measure exclusions; however, we disagree with the proposed exclusion criteria of patients less than 18 years old. Many IRFs treat patients younger than 21 when necessary, is reflected by the Functional Independence Measure (FIM) and IRF-Patient Assessment Instrument (PAI), both of which are used to assess patients age seven or older. ARN encourages CMS not to exclude patients under 18 years old from the discharge measures. We also recommend that CMS ensure the data collection time frame is the same for all PAC settings. Currently, the time frame for the initial data collection for the project varies from one year (SNFs), two years (IRFs and LTCHs), and three years (HHAs), we urge the Agency to implement one standard reporting time frame across PAC providers. While reporting may be based on either one year of data, two years of data, etc., we strongly recommend CMS and its subcontractors address the inconsistent reporting periods. Moreover, in regards to the within-stay criteria, IRFs have experienced a rise in the number of patients who must return to the acute inpatient care setting within 48-72 hours of admission due to the disparity between the level of care which their condition (either a co-morbidity or complication secondary to the presenting diagnosis) requires and the level of care that an IRF is able to provide. An example of this is leukemia patients as many times they require inpatient rehabilitation to discharge them from the hospital and that the patient’s survival may be related to their functional status and rehabilitation can improve function and symptoms such as fatigue, well-being, and pain. To this end, ARN disagrees with the within-stay criteria delineated for IRFs and believes that in specific instances, readmissions are a necessity for patient safety, and not necessarily preventable. Finally, while ARN is generally supportive of the potentially preventable readmissions measure specifications, we have several concerns. As ARN has stated in previous comment letters, the IRF measure is based on 24 months of data while the SNF measure is based on 12 months of data. PAC facilities should not be penalized for conditions that prompt readmissions that are unrelated to the patient’s initial reason for admission. Moreover, as previously stated, we oppose CMS’s proposal to require PAC providers to utilize 30-day readmission claims data to determine their readmission rates. Using claims data to calculate readmission rates will be difficult for IRFs and other PAC settings, as claims data are cumbersome to use and access. Employing a 30-day readmission rate measure will not provide meaningful insight or have an impact on quality improvement efforts if PAC settings do not have unrestricted access to the data. (Submitted by: Association of Rehabilitation Nurses)
  
  
• This set of potentially preventable readmission (PPR) measures for post-acute care (PAC) estimates the risk-standardized rate of unplanned, potentially preventable readmissions for patients (Medicare fee-for-service [FFS] beneficiaries) who receive services in one of the following post-acute care provider types: skilled nursing facilities (SNFs), inpatient rehabilitation facilities (IRFs), long-term care hospitals (LTCH), and home health agencies (HHA). These outcome measures reflect readmission rates for patients who are readmitted to a short-stay acute-care hospital or an LTCH with a principal diagnosis considered to be unplanned and potentially preventable. Six PPR PAC measures are being developed: • Four of these measures assess PPR within a 30-day window following discharge from PAC—one measure for each PAC setting (i.e., SNF, IRF, LTCH, and HH)—and are being developed to meet the requirements of the IMPACT Act. • An additional SNF measure (SNF PPR), which is being developed to meet the PAMA requirements, assesses PPR during the 30-day period following a hospital discharge to a SNF setting. • An additional IRF measure assesses PPR during the IRF stay (referred to as the within-stay window) which is being developed for use in the IRF Quality Reporting Program. Of note with respect to potentially preventable readmissions and a patient’s occupational therapy needs, several recent studies consider whether returning to the community from a recent hospitalization with unmet activities of daily living (ADL) need was associated with probability of readmission. The findings from these studies indicate that this indeed may be a considerable risk factor. The studies reveal that any older patients are discharged from the hospital with ADL disability. Those who report unmet need for new ADL disabilities after they return home from the hospital are particularly vulnerable to readmission. This area is not typically addressed in a thorough manner through current discharge practices. This needs to change. Patients' functional needs after discharge should be carefully evaluated and addressed. Factors such as enabling self-management and ensuring appropriate medication management and ADLs, such as cooking and eating are addressed, can have a direct effect on readmissions The profession of occupational therapy is built on delivering patient-centered care, seeking to keep the patient at the highest functional level in the least restrictive setting and to reduce caregiver burden and health care system resource utilization. Occupational therapy directly addresses the enablement of successful performance of ADLs. This focus, experience and research base in occupational therapy must be fully tapped to address this component of readmission prevention. A further issue is that Self-management is a key element in successful care, and occupational therapists are experts in motivation, task analysis, and psychosocial contexts, which all contribute to enabling positive outcomes. In order to successfully re-establish or establish new routines and habits to meet health care needs, such as medication management, proper sleep hygiene, and following other medical directives, is within the scope and proven effectiveness of occupational therapy. AOTA would encourage CMS to examine more fully, perhaps through pilot testing, the value of occupational therapy evaluation and intervention participation as part of discharge planning. This could identify more clearly specific ADL limitations prior to and after discharge and assure proper interventions are provided that address fully and completely the range of ADL and other activity restrictions and capacities to enable optimum recovery from the condition as well as optimum participation of the client in their own care. With respect to the IRF Readmission measure, AOTA makes the following comments: • IRFs are subject to a reduced payment if the patient is transferred to the acute care hospital before meeting the average length of stay for the condition. Therefore, if there is a measure that tracks readmissions occurring during the IRF stay, the IRF could be subject to two payment reductions simultaneously: one payment reduction for the transfer and a second for the readmission. For this reason, AOTA recommends that the PPR penalty only apply to IRF transfers not subject to the transfer policy. • AOTA recommends that the draft Specifications for IRF PPR measure risk adjustment consider the following factors: o Age; o Sex; o Original reason for Medicare entitlement (age, disability, or ESRD); o Surgery category, if present; o Receiving dialysis in prior short-term stay; o Principal diagnosis on prior short-term claim; o Comorbidities from secondary diagnoses on the prior short-term claim and diagnoses from earlier short-term stays up to one year before the post-acute care admission; o Aggregates of the IRF case-mix groups (IRF-specific exclusion) AOTA has reviewed the readmissions measures released for comment and makes the following general comments: • The PPR measures Numerator and Denominator definition language are consistent from previous materials. • The Readmissions Measures are based on two years of claims data. Use of multiple years is acceptable in the measure review process. • The exclusion criteria for the various measures are associated with the Yale/New Haven readmission measures. Co-developed by the Centers for Medicare & Medicaid Services (CMS) and researchers at Yale University, estimates the risk-standardized rate of unplanned, all-cause readmissions to a hospital for any eligible condition within 30 days of hospital discharge for patients aged 18 and older. The CMS/Yale measure is specified for evaluating hospital performance. AOTA would prefer for the exclusion criteria to be specified for evaluating performance in post-acute care settings. (Submitted by: American Occupational Therapy Association)
  
  
• At least one NRHI member has expressed concerns regarding this measure. Specifically, clinical staff at HealthInsight have concerns with inclusion of this measure due to the validity of and complicated methodologies used for "risk adjustment for patient characteristics and a statistical estimate of the facility effect beyond patient mix." This methodology is frequently cited as the basis for determining potentially avoidable hospital facility admissions. This methodology is not well-established. (Submitted by: Network for Regional Healthcare Improvement)
  
  
• We are generally supportive of measuring PAC providers on potentially preventable readmissions and believe that much of the work performed for these measures is positive, such as the risk adjustment methodology. We believe that the PAC industry should embrace measures, such as these PPR measures, that help patients stay at home – a central tenant of post-acute care. Accordingly, we value these PPR measures as useful tools in advancing that goal. We have several comments that will serve as constructive additions to the development of these measures. We hope that NQF and the measure developers will analyze and consider these comments and how they could improve the PPR measure development for IRFs. B. USE OF UPSTREAM DRGS; DIFFERENCES IN SETTING-SPECIFIC MEASURES COULD CAUSE DIFFICULTY IN DATA COMPARISONS The various PPR measure proposals each propose to rely on DRGs from the initial hospital stay as the anchor condition for a PPR in a PAC setting. We are concerned that the high level of variability in DRG coding practices amongst upstream hospitals will present problems in the resulting PAC PPR data because lack of consistency in initial DRG coding will have an effect on the number of patients at a given PAC provider are “eligible” for inclusion in a condition-based PPR measure. This dynamic would create a barrier to making sound apples-to-apples PPR data comparisons. Another issue with using DRG codes for the initial hospital stay is that such codes are often unreliable indicators as to why a particular patient is referred to a particular PAC setting, not to mention why they are eventually readmitted after the PAC stay. For example, a patient admitted to the hospital for congestive heart failure and who is treated in a subsequent PAC setting may end up returning to the hospital during the 30-day post-PAC discharge window not for reasons associated with congestive heart failure, but because he was not properly hydrating. Although this readmission would count under the PPR measure, it would not capture the actual reason the patient was readmitted. By implementing a post-PAC discharge PPR measure that is coded all the way back to the initial hospital DRG risks missing the clinical reason that was present within the particular PAC setting at the time of readmission. Similarly, we note that IRFs are the only provider for which a “within stay” measure is being developed (for the IRF QRP, not under the IMPACT Act per se). Although already in place for HHAs, we believe that such “within stay” PPR measures could also be helpful and are possible for LTCHs as well, and ask CMS and RTI to consider creating such a measure to enable more congruity into future PPR comparisons between all PAC providers. The SNF 30-day post hospital discharge measure, while covering at least some part of a SNF stay, is also incongruent because it would fail to capture any “within stay” PPRs that occur when a patient is still in a SNF 30 days after discharge from the acute hospital. Although these differences are minor, they present concerns when considering the ability of future policy makers to make direct comparisons within the PAC industry. C. RISK ADJUSTMENT FOR SOCIODEMOGRAPHIC AND FUNCTIONAL STATUS The draft specifications indicate that the PPR measures will all be risk-adjusted for multiple variables, including age, sex, dialysis status, and for the IRF measures, aggregates of the IRF case-mix groups. We support this broad range of risk adjustment, but also believe that CMS and RTI should consider including risk adjustment for sociodemographic factors and functional status as well. Recent academic literature has added evidence to the notion that readmissions back to hospitals are driven by more directly by patient status factors and not by the quality of care delivered. Furthermore, NQF in April began a two-year trial program of a temporary policy change that would allow risk adjustment of performance measures for socioeconomic and other demographic factors, a departure from earlier quality measurement positions that viewed sociodemographic risk adjustment as inappropriate. However, with evidence that sociodemographic risk is real and impact readmission rates, we recommend that the measure developers consider including it in the list of risk adjustment factors for these measures. We also believe adjusting for functional status would benefit the accuracy of the PPR measures. Functional status has a direct correlation with a patient’s ability to remain healthy at home after PAC services have ended, meaning that PAC providers who treat more functionally impaired patients will likely have a higher PPR rate. Accordingly, we recommend that the measure developers consider how to apply an additional risk-adjustment factor for functional status in the overall risk adjustment methodology. D. RISK OF CONFUSION IN MULTIPLE READMISSION MEASURES As part of the existing IRF PPS Quality Reporting Program (“IRF QRP”), IRFs already report All-cause Unplanned Readmission (NQF #2502) and are scheduled to have this measure publicly reported beginning next year. Other PAC types will also require public reporting of a general all-cause readmissions measure. We are concerned that, as required by the IMPACT Act, the eventual public reporting of these PPR measures (which are essentially a subset of the more general all-cause measures) will result in substantial confusion amongst members of the public. For example, if both measures are publicly reported, PAC providers will have one all-cause unplanned readmission rate, but likely a different PPR rate. We question whether the crucial distinctions between these two different, but related, readmissions measures will be readily apparent to members of the public who take the time to assess different PAC providers based on readmission rates. Similarly, providers may have trouble accurately understanding the purpose and ultimate use of additional readmissions measures. IRF providers already receive annual PEPPER reports and also report on the All-cause measure (NQF #2502). With the addition of a “within stay” measure and a post-discharge PPR measure to these existing readmission measures, CMS should take deliberate steps to clearly communicate the intended use for each of these four readmission tools and how they will relate to one another. Without such clarifying communication, providers may find themselves adrift in various readmissions data without a clear idea of how it is all being used by the Agency. This is another example, similar to the IRF short-stay/transfer case considerations discussed above, of the reconciliation issues that CMS must face in overlaying IMPACT Act requirements with existing PAC regulations, and insofar as NQF and the measure developers can eliminate the duplicative requirements of IMPACT Act implementation, including overlapping public reporting of two “flavors” of readmissions, we highly encourage them to do so. Indeed, the IMPACT Act encourages CMS to refrain from duplication in methods for data collection, and we recommend that this theme be extended to other parts of IMPACT Act implementation as well. NQF also has policies against endorsing measures that have unnecessarily duplicative functions with existing measures. (Submitted by: HealthSouth)
  
  

• AMRPA would refer NQF and the MAP to the comments made regarding MUC 496, as they relate to our suggested modifications to the concept of a potentially preventable readmissions (PPR) and the risk adjustment methodology. As drafted, we believe the definition of a PPR has the potential to unfairly penalize IRFs and will need to be adjusted to truly improve outcomes for the patients we treat. We also believe that the collection of data beyond claims data will be critical. We are concerned that the confluence of payment and quality policies has the potential to inappropriately target IRFs in two ways if there is a PPR during the IRF stay. Specifically, we do not think that a “within stay” measure should be applicable until day four or after from the date of admission to the IRF. We recommend this for two reasons. First, if a within stay measure is applied too early after the date of admission to the IRF, it might be a barrier to admission for patients especially for sicker patients who are more likely to be readmitted to the acute care hospital would negatively impact the IRF’s quality score. Second, at this time the Inpatient Rehabilitation Facility Prospective Payment System (IRF PPS) includes a payment policy that provides a lower payment to the IRF if the patient is transferred back to the acute care hospital before day four after admission to the IRF. Therefore, if the within stay measure as drafted does not account for this payment policy, an IRF who transfers a patient back to the acute care hospital prior to day four after admission would face double potential payment reductions: one under the IRF PPS transfer policy and a second for the readmission quality metric. It is unreasonable to apply two financial penalties simultaneously for the same event. We remain concerned that this measure is not fully developed. Therefore, it may be premature to endorse this measure for the purpose of the IRF QRP. (Submitted by: American Medical Rehabilitation Providers Association)
  
  
• While we appreciate CMS' consideration for a readmission measure that an IRF may be able to control, we are highly concerned that measure development has not yet been completed, and that measure level testing has not been completed in such a way as to determine whether measure results will meet criteria for being reliable and valid. We are also concerned that the measure materials provided to date do not provide sufficient information related to how the measure will account for "program interruptions" that occur within the IRF stay and do not require the IRF to discharge the patient. We also wonder whether testing results will show differences between the different facility types, and whether risk-adjustment methodologies are being considered to account for differences between facilities. (Submitted by: UDSMR)
  
  
• In regards to the within-stay criteria, IRFs have experienced a rise in the number of patients who must return to the acute inpatient care setting within 48-72 hours of admission due to the disparity between the level of care which their condition (either a co-morbidity or complication secondary to the presenting diagnosis) requires and the level of care that an IRF is able to provide. An example of this is leukemia patients as many times they require inpatient rehabilitation to discharge them from the hospital and that the patient’s survival may be related to their functional status and rehabilitation can improve function and symptoms such as fatigue, well-being, and pain. To this end, ARN disagrees with the within-stay criteria delineated for IRFs and believes that in specific instances, readmissions are a necessity for patient safety, and not necessarily preventable. (Submitted by: Association of Rehabilitation Nurses)
  
  
• This measure is currently in the early stage of development, has never been submitted to NQF and thus is not NQF-endorsed (Submitted by: AdvaMed)
  
  
• At least one NRHI member has expressed concerns regarding this measure. Specifically, clinical staff at HealthInsight have concerns with inclusion of this measure due to the validity of and complicated methodologies used for "risk adjustment for patient characteristics and a statistical estimate of the facility effect beyond patient mix." This methodology is frequently cited as the basis for determining potentially avoidable hospital facility admissions. This methodology is not well-established. (Submitted by: Network for Regional Healthcare Improvement)
  
  
• We are generally supportive of measuring PAC providers on potentially preventable readmissions and believe that much of the work performed for these measures is positive, such as the risk adjustment methodology. We believe that the PAC industry should embrace measures, such as these PPR measures, that help patients stay at home – a central tenant of post-acute care. Accordingly, we value these PPR measures as useful tools in advancing that goal. We have several comments that will serve as constructive additions to the development of these measures. We hope that NQF and the measure developers will analyze and consider these comments and how they could improve the PPR measure development for IRFs. A. THE IRF “WITHIN STAY” MEASURE SHOULD ACCOUNT FOR THE EXISTING THREE-DAY SHORT STAY AND TRANSFER CARE POLICIES FOR IRFS The measure specification document states that the IRF “within stay” measure “is intended to capture readmissions during the IRF stay (i.e. program interruptions where the patient is readmitted to the acute care setting) and readmissions (i.e. [sic] acute care transfers) at the end of the IRF stay.” This specification does not detail precisely when the measure would begin capturing such readmissions, but the “within stay” measure should not apply until day four or after from a patient’s date of admission to the IRF. IRFs are currently subject to a policy under the IRF Prospective Payment System (“PPS”) that provides a uniform payment for certain cases with a length-of-stay not exceeding 3 days. This three-day window was designed by CMS to help ensure that IRFs have enough time to fully assess whether a patient admitted to the IRF remains in need of IRF services. According to CMS, “an IRF is eligible to receive the IRF short stay payment for 3 days or less if a patient’s thorough preadmission screening shows that the patient is an appropriate candidate for IRF care but then something unexpected happens between the preadmission screening and the IRF admission such that the patient is no longer an appropriate candidate for IRF care on admission and the day count is greater than 3.” Therefore, since the IRF PPS already accounts for admission of some patients who are subsequently determined by day 3 of the IRF stay as no longer needing IRF services by reimbursing the IRF by paying a lower rate, the IRF “within stay” PPR measure should not further penalize IRFs for readmissions that occur within the same time period. There is precedent for this accommodation for short-stay policies within other PAC readmissions measures already: the existing home health “within stay” rehospitalization measure specifically excludes stays that begin with a Low-Utilization Payment Adjustment (“LUPA”) (i.e., stays with four or fewer home health visits, the equivalent of short-stays in IRFs). Short stays in any PAC provider should be excluded from the PPR measure because the PAC provider will not have the opportunity to meaningfully deliver care to a patient. As such, it would be unfair to attribute the cause of the readmission to the PAC provider when they are so limited in what they can do within the short-stay window. To the extent that CMS and RTI develop true “within stay” PPR measures for SNFs and LTCHs in the future, we believe those measures should also account for short stays policies. Similarly, the existing IRF transfer policy provides that, for patients who are transferred from an IRF to a hospital, LTCH, SNF, or other IRF, and whose length of stay is less than the average length of stay for the pertinent CMG, the transferring IRF receives a per diem payment, the rate of which is calculated by dividing the typical per discharge CMG payment by the average length of stay. Therefore, because this policy reduces IRF payment automatically from the normal CMG payment for cases that are discharged early, it functions to discourage IRFs from making such early discharges. If a “within stay” PPR measure were to start too early into IRF stays, these two policies would overlap and any transfers would effectively cause a double penalty – a lower payment under the transfer policy as well as a PPR “ding.” Accordingly, in order to avoid applying two separate payment disincentives to a single event, an IRF “within stay” PPR measure should either begin capturing readmissions after the average length of stay passes, or currently-defined transfers that would also qualify as a PPR should no longer be counted as transfers. We recommend that NQF and the measure developers consider the intersections of these potentially overlapping policies and structure the IRF “within stay” PPR measure to begin on or after day four of the IRF stay, and not to conflict with the existing transfer policy. B. USE OF UPSTREAM DRGS; DIFFERENCES IN SETTING-SPECIFIC MEASURES COULD CAUSE DIFFICULTY IN DATA COMPARISONS The various PPR measure proposals each propose to rely on DRGs from the initial hospital stay as the anchor condition for a PPR in a PAC setting. We are concerned that the high level of variability in DRG coding practices amongst upstream hospitals will present problems in the resulting PAC PPR data because lack of consistency in initial DRG coding will have an effect on the number of patients at a given PAC provider are “eligible” for inclusion in a condition-based PPR measure. This dynamic would create a barrier to making sound apples-to-apples PPR data comparisons. Another issue with using DRG codes for the initial hospital stay is that such codes are often unreliable indicators as to why a particular patient is referred to a particular PAC setting, not to mention why they are eventually readmitted after the PAC stay. For example, a patient admitted to the hospital for congestive heart failure and who is treated in a subsequent PAC setting may end up returning to the hospital during the 30-day post-PAC discharge window not for reasons associated with congestive heart failure, but because he was not properly hydrating. Although this readmission would count under the PPR measure, it would not capture the actual reason the patient was readmitted. By implementing a post-PAC discharge PPR measure that is coded all the way back to the initial hospital DRG risks missing the clinical reason that was present within the particular PAC setting at the time of readmission. Similarly, we note that IRFs are the only provider for which a “within stay” measure is being developed (for the IRF QRP, not under the IMPACT Act per se). Although already in place for HHAs, we believe that such “within stay” PPR measures could also be helpful and are possible for LTCHs as well, and ask CMS and RTI to consider creating such a measure to enable more congruity into future PPR comparisons between all PAC providers. The SNF 30-day post hospital discharge measure, while covering at least some part of a SNF stay, is also incongruent because it would fail to capture any “within stay” PPRs that occur when a patient is still in a SNF 30 days after discharge from the acute hospital. Although these differences are minor, they present concerns when considering the ability of future policy makers to make direct comparisons within the PAC industry. C. RISK ADJUSTMENT FOR SOCIODEMOGRAPHIC AND FUNCTIONAL STATUS The draft specifications indicate that the PPR measures will all be risk-adjusted for multiple variables, including age, sex, dialysis status, and for the IRF measures, aggregates of the IRF case-mix groups. We support this broad range of risk adjustment, but also believe that CMS and RTI should consider including risk adjustment for sociodemographic factors and functional status as well. Recent academic literature has added evidence to the notion that readmissions back to hospitals are driven by more directly by patient status factors and not by the quality of care delivered. Furthermore, NQF in April began a two-year trial program of a temporary policy change that would allow risk adjustment of performance measures for socioeconomic and other demographic factors, a departure from earlier quality measurement positions that viewed sociodemographic risk adjustment as inappropriate. However, with evidence that sociodemographic risk is real and impact readmission rates, we recommend that the measure developers consider including it in the list of risk adjustment factors for these measures. We also believe adjusting for functional status would benefit the accuracy of the PPR measures. Functional status has a direct correlation with a patient’s ability to remain healthy at home after PAC services have ended, meaning that PAC providers who treat more functionally impaired patients will likely have a higher PPR rate. Accordingly, we recommend that the measure developers consider how to apply an additional risk-adjustment factor for functional status in the overall risk adjustment methodology. D. RISK OF CONFUSION IN MULTIPLE READMISSION MEASURES As part of the existing IRF PPS Quality Reporting Program (“IRF QRP”), IRFs already report All-cause Unplanned Readmission (NQF #2502) and are scheduled to have this measure publicly reported beginning next year. Other PAC types will also require public reporting of a general all-cause readmissions measure. We are concerned that, as required by the IMPACT Act, the eventual public reporting of these PPR measures (which are essentially a subset of the more general all-cause measures) will result in substantial confusion amongst members of the public. For example, if both measures are publicly reported, PAC providers will have one all-cause unplanned readmission rate, but likely a different PPR rate. We question whether the crucial distinctions between these two different, but related, readmissions measures will be readily apparent to members of the public who take the time to assess different PAC providers based on readmission rates. Similarly, providers may have trouble accurately understanding the purpose and ultimate use of additional readmissions measures. IRF providers already receive annual PEPPER reports and also report on the All-cause measure (NQF #2502). With the addition of a “within stay” measure and a post-discharge PPR measure to these existing readmission measures, CMS should take deliberate steps to clearly communicate the intended use for each of these four readmission tools and how they will relate to one another. Without such clarifying communication, providers may find themselves adrift in various readmissions data without a clear idea of how it is all being used by the Agency. This is another example, similar to the IRF short-stay/transfer case considerations discussed above, of the reconciliation issues that CMS must face in overlaying IMPACT Act requirements with existing PAC regulations, and insofar as NQF and the measure developers can eliminate the duplicative requirements of IMPACT Act implementation, including overlapping public reporting of two “flavors” of readmissions, we highly encourage them to do so. Indeed, the IMPACT Act encourages CMS to refrain from duplication in methods for data collection, and we recommend that this theme be extended to other parts of IMPACT Act implementation as well. NQF also has policies against endorsing measures that have unnecessarily duplicative functions with existing measures. (Submitted by: HealthSouth)
  
  

• We would like to note the following concerns regarding the development of these measures: 1. Measure development for the IMPACT Act appears to create 4 separate measures that are site specific with their own set of criteria and risk adjustment factors. The IMPACT Act aims to create quality measurement within Post-Acute Care (PAC) that is "standardized and interoperable". By developing measures which differ in their calculation, such as site-specific inclusion/exclusion criteria or risk-adjustment factors, PAC sites will be subjected to comparisons of quality that do not differ based upon the quality of care provided, but rather the differences inherent in the measure calculations. We strongly recommend that CMS and the measure developers produce a measure that is calculated in a "standardized and interoperable" manner with inclusion/exclusion criteria and risk-adjustment factors the are applied consistently across all PAC sites. 2. The measures developed for the IMPACT Act hold PAC providers responsible for circumstances that occur outside of their control. While PAC providers are responsible for providing care that allows patients safe transitions to their next setting or home, holding these providers accountable fora time period in which they are not furnishing care should not differentiate the quality of care they actually provided. In Appendix A, Table A1 lists various conditions that define potentially preventable hospital readmission for 30-days post-PAC discharge. In this list, conditions such as Asthma, Congestive Heart Failure, and Dehydration are listed as a reason for a potentially preventable readmission. So if the patient fails to care for themselves post-discharge (against the advice and discharge instructions provided by a PAC provider) and requires readmission to Acute Care for one of these conditions, the PAC provider is penalized even though the patient is no longer being cared for by the PAC provider. We strongly recommend that CMS and the measure developers produce a measure that represents the quality of care provided by PAC providers while the patient is in their care. (Submitted by: UDSMR)
  
  
• This measure will encourage post-acute care providers to optimize care for patients before discharge. (Submitted by: Smith & Nephew )
  
  
• This set of potentially preventable readmission (PPR) measures for post-acute care (PAC) estimates the risk-standardized rate of unplanned, potentially preventable readmissions for patients (Medicare fee-for-service [FFS] beneficiaries) who receive services in one of the following post-acute care provider types: skilled nursing facilities (SNFs), inpatient rehabilitation facilities (IRFs), long-term care hospitals (LTCH), and home health agencies (HHA). These outcome measures reflect readmission rates for patients who are readmitted to a short-stay acute-care hospital or an LTCH with a principal diagnosis considered to be unplanned and potentially preventable. Six PPR PAC measures are being developed: • Four of these measures assess PPR within a 30-day window following discharge from PAC—one measure for each PAC setting (i.e., SNF, IRF, LTCH, and HH)—and are being developed to meet the requirements of the IMPACT Act. • An additional SNF measure (SNF PPR), which is being developed to meet the PAMA requirements, assesses PPR during the 30-day period following a hospital discharge to a SNF setting. • An additional IRF measure assesses PPR during the IRF stay (referred to as the within-stay window) which is being developed for use in the IRF Quality Reporting Program. Of note with respect to potentially preventable readmissions and a patient’s occupational therapy needs, several recent studies consider whether returning to the community from a recent hospitalization with unmet activities of daily living (ADL) need was associated with probability of readmission. The findings from these studies indicate that this indeed may be a considerable risk factor. The studies reveal that any older patients are discharged from the hospital with ADL disability. Those who report unmet need for new ADL disabilities after they return home from the hospital are particularly vulnerable to readmission. This area is not typically addressed in a thorough manner through current discharge practices. This needs to change. Patients' functional needs after discharge should be carefully evaluated and addressed. Factors such as enabling self-management and ensuring appropriate medication management and ADLs, such as cooking and eating are addressed, can have a direct effect on readmissions The profession of occupational therapy is built on delivering patient-centered care, seeking to keep the patient at the highest functional level in the least restrictive setting and to reduce caregiver burden and health care system resource utilization. Occupational therapy directly addresses the enablement of successful performance of ADLs. This focus, experience and research base in occupational therapy must be fully tapped to address this component of readmission prevention. A further issue is that Self-management is a key element in successful care, and occupational therapists are experts in motivation, task analysis, and psychosocial contexts, which all contribute to enabling positive outcomes. In order to successfully re-establish or establish new routines and habits to meet health care needs, such as medication management, proper sleep hygiene, and following other medical directives, is within the scope and proven effectiveness of occupational therapy. AOTA would encourage CMS to examine more fully, perhaps through pilot testing, the value of occupational therapy evaluation and intervention participation as part of discharge planning. This could identify more clearly specific ADL limitations prior to and after discharge and assure proper interventions are provided that address fully and completely the range of ADL and other activity restrictions and capacities to enable optimum recovery from the condition as well as optimum participation of the client in their own care. AOTA has reviewed the readmissions measures released for comment and makes the following general comments: • The PPR measures Numerator and Denominator definition language are consistent from previous materials. • The Readmissions Measures are based on two years of claims data. Use of multiple years is acceptable in the measure review process. • The exclusion criteria for the various measures are associated with the Yale/New Haven readmission measures. Co-developed by the Centers for Medicare & Medicaid Services (CMS) and researchers at Yale University, estimates the risk-standardized rate of unplanned, all-cause readmissions to a hospital for any eligible condition within 30 days of hospital discharge for patients aged 18 and older. The CMS/Yale measure is specified for evaluating hospital performance. AOTA would prefer for the exclusion criteria to be specified for evaluating performance in post-acute care settings. (Submitted by: American Occupational Therapy Association)
  
  
• At least one NRHI member has expressed concerns regarding this measure. Specifically, clinical staff at HealthInsight have concerns with inclusion of this measure due to the validity of and complicated methodologies used for "risk adjustment for patient characteristics and a statistical estimate of the facility effect beyond patient mix." This methodology is frequently cited as the basis for determining potentially avoidable hospital facility admissions. This methodology is not well-established (Submitted by: Network for Regional Healthcare Improvement)
  
  
• According to the RTI measure specifications, The PPHR measures are constructed by multiplying a standardized risk ratio by the unadjusted average rate of readmission in the specific PAC setting’s population. The average readmission rates used in the calculation are not adjusted for patient clinical differences between PAC settings. As a result, the differences in the PPHRs between PAC settings (e.g., LTCH, SNF, IRF, and HHA) may reflect differences in patient clinical differences rather than differences in care quality. There exist significant differences in patient severity and acuity across PAC provider settings. For example, according to RTI analysis of 2012-2013 Medicare claims data, the unadjusted unplanned 30-day mean readmission rate among LTCHs with at least 25 index stays was 24.3% whereas the comparable rate for inpatient rehabilitation facilities was 12.4%. While focusing on PPHRs may close some of this gap, patients admitted to LTCHs may be more susceptible to some of the PPHRs than other patients. The PPHR measures require a short-term acute-care stay within 30 days prior to a PAC admission. This requirement would mostly exclude patients discharged from LTCHs to less intensive care settings in calculating the readmission rates of those less intensive care settings. For example, if a patient is discharged from a short term acute care (STCH) to a LTCH and spends more than 30 days in the LTCH before being discharged to a skilled nursing facility (SNF), that patient would not be included in calculating readmission rates for that SNF. Patients who transition from more intensive care settings (such as LTCHs) to less intensive care settings (such as SNFs and HHA) are likely to have higher observed and unobserved severity relative to those who transition from acute care stay to the less intensive PAC setting directly or within a 30-day period. Therefore, this requirement would cause the PPHR measure for the less intensive care settings to be based on a limited and less severe portion of their broader population, potentially exacerbating the differences in patient acuity across PAC settings described above. We recommend that this requirement is changed so that episodes in which a patient moves through the continuum of care following discharge from an acute care hospital are not systematically excluded from the measure sample. This could be done looking back at contiguous inpatient stays prior to admission to the PAC admission (pre-PAC episode). Any admission to a short-term acute care hospital within that pre-PAC episode would serve as the anchor stay. These cases would be included in the measure even if the STCH stay occurred more than 30 days prior to admission to the PAC. This revised requirement would ensure that the PPHR measures are based on a patient population that has experienced a STCH stay without selecting a less severe portion of the population for the measure computation. In defining potentially preventable hospital readmissions, RTI's draft measure specifications do not distinguish across PAC settings. The draft measure specifications cite studies on readmission from SNFs and inpatient rehabilitation facilities, but do not cite any studies on for LTCHs. It would be important to include potentially preventable readmissions that have been identified for the LTCH setting and using readmissions that are specific to each PAC setting in constructing the measure. (Submitted by: National Association of Long Term Hospitals)
  
  

• APTA supports the goal of improving the quality of health care. Physical therapists are committed to providing high-quality, timely care and to the promotion of evidence-based and patient-centered practice. Furthermore, APTA feels that it is essential that we move towards a common set of quality measures across the continuum of care. APTA supports the implementation of a discharge to community measure across the care settings. We believe that successful transitions to the community following discharge from the respective post-acute care settings will decrease potentially preventable readmissions. However, the APTA does have some concerns regarding the proposed measure methodology. These concerns are discussed below. APTA believes that a patient’s level of function does impact a patients ability to transition successfully back to the community. Recent evidence indicates that patient function is associated with increased risk of 30-day all-cause hospital readmissions and may be an important factor in preventing readmissions for Medicare seniors that is not currently accounted for in measure methodologies1. APTA was pleased to see “activity of daily living” scores in the home health setting included in the risk adjustment methodology for the readmissions measures, and we recently commented, encouraging the use of patient function in the risk adjustment methodology for the post-acute care setting readmissions measures. We believe that readmissions and discharge to community are closely related measures and that patient function may also be an important risk adjustment variable for discharge to community. APTA appreciates that CMS has strict deadlines for the implementation of measures under the IMPACT act, however, as these measures will be new to the respective post-acute care settings, we encourage that settings have the ability to review this data as early as possible in order to understand and, more importantly, so that the respective setting have time to implement strategies to decrease readmissions where necessary. As many of these settings do not always receive feedback on the readmissions of their patients post-discharge, this data will be new to many facilities. APTA recognizes that the overall goal of IMPACT is for PAC providers (HH, IRF, SNF and LTCH) to collect and report standardized and interoperable patient assessment data, quality and resource use measures. We acknowledge that during the initial IMPACT implementation years that there will be a transition period which will include the addition of new measures into all of the post-acute care settings. We believe that achieving a standardized and interoperable patient assessment data set and stable quality measures as quickly as possible will allow for better cross-setting comparisons as well as the evolution of better quality measures with uniform risk standardization, thus achieving the true aim of IMPACT. 1 Greysen SR, Cenzer IS, Auerbach AD, Covinsky KE. Functional Impairment and Hospital Readmission in Medicare Seniors. JAMA Intern Med. 2015;175(4):559-565. (Submitted by: American Physical Therapy Association)
  
  
• • MUC15-523: Discharge to Community-Post Acute Care (PAC) Home Health Quality Reporting Program (Required under the IMPACT Act). There is a difference between the measure noted in the 2015 National Impact Assessment of the Centers for Medicare & Medicaid Services (CMS) Quality Measures Report of Patients Able to Live in the Community at Discharge and the measure under consideration, Discharge to Community-Post Acute Care. The numerator in the measure under consideration includes a risk adjustment for patient characteristics. There is no information on what patient characteristics were taken into account. The List of Measures under Consideration for December 1, 2015, stated there is considerable variation in discharge to community rates within and across post-acute settings. Studies show geographic variation, variation across patient socioeconomic characteristics (for example, race and ethnicity), and variation by facility characteristics (for profit vs. nonprofit, freestanding vs. hospital-based, urban vs. rural). The lack of information on what patient characteristics would be included in the risk adjustment, and the considerable variation of outcomes noted in studies leads us to recommend that this measure requires continued development. (Submitted by: LeadingAge)
  
  
• We have the following concerns related to this measure: 1. The inclusion of unplanned readmissions in the discharge to community measure development and calculation. Our concerns related to the inclusion of unplanned readmissions in the discharge to community measure development and calculation can be defined as follows: a. Duplication/commingling of quality metrics With the inclusion of unplanned readmissions in the discharge to community measure development and calculation, there is the potential for there to be a strong correlation between the performances of the two measures. In other words, those PAC providers that perform poorly with the unplanned readmission measure are more likely to perform poorly on this measure, while those PAC providers that perform well with the unplanned readmission measure have a better opportunity to perform well on this measure. Take for instance the scenario of two facilities with similar patient populations (age, sex, etc.), where provider A discharges 75% to a community setting with a 5% readmission rate while provider B discharges 72% to a community setting with a 2% readmission rate. Which facility performs better at discharging patients to a community setting? Is the discharge to community measure truly representative of the ability of the post-acute care provider to provide services that get the patient back to a community setting, or is too much emphasis being placed upon the ability of the patient to stay in a community setting? b. Inclusion of a factor that is outside of post-acute care provider control While a post-acute care provider can provide services that impact the ability of a patient to be discharged back to a community setting, whether or not a patient returns to an acute care facility in the 30 days following that PAC discharge can be completely unrelated to the services furnished by the PAC provider. For instance, a patient is discharged home from a post-acute care provider but is re-admitted to acute care 10 days later for pneumonia (which is listed as a diagnosis category that cannot be considered planned). Should the post-acute care provider’s discharge to community percentage be penalized if the readmission is completely unrelated to the services provided within post-acute care? With the concerns noted above, we would recommend that the measure developers remove the unplanned readmission component from the discharge to community measurement, and provide side-by-side comparison of these two separate and unique quality measures. 2. Potential for site-specific risk adjustment for a “cross-setting” measure. While we note that the risk adjustment variables are “under consideration”, we are very concerned that the variables noted in the draft specification designate certain variables as being applicable to unique or specific post-acute care providers. The IMPACT Act requires the specification of quality measures and resource use metrics that are standardized and interoperable across PAC settings. In creating a measure where consideration is being given to risk adjustment variables that differ by post-acute care setting, is CMS truly meeting the definition of “standardized and interoperable”? As an example, “Activities of Daily Living” (ADLs) are noted as one of the Clinical conditions variables, yet is noted for “the HHA setting only”. If the measure developer can show that ADLs can be used to differentiate the performance within the HHA population, couldn’t ADLs also be used to differentiate the performance of all post-acute care settings and produce a measurement value that is “standardized and interoperable”? We urge CMS and the measure developers to introduce risk-adjustment variables that are “standardized and interoperable” in order to meet the IMPACT Act requirements. (Submitted by: UDSMR)
  
  
• Similar to the PPR measure, NAHC has concern that HH providers will be held accountable for unavoidable readmissions within 30 days post discharge to the community. Also, the PPR and discharge to community measures overlap both in structure and purpose. Both measure readmissions 30 days post-discharge. (Submitted by: National Assocition for Home Care & Hospice (NAHC))
  
  
• The Alliance supports the development of this measure, but has concerns that it articulated to the measure development contractors in recent comments. Most significantly, the discharge to community measure is structured as a single measure, but the target populations are not standardized among the various settings. Specifically, the target population for the home health setting is all Medicare fee-for-service persons admitted to home health care. An acute care discharge in the 30 days preceding the start of the home health episode is not required; by contrast, for the SNF, IRF and LTCH settings, the target population is only those who were admitted within 30 days of discharge from an acute care hospital. As a result, for home health settings, the discharge to community measure is not solely a post-acute care measure. Further, as drafted in the specifications, the measure as applied to home health care would be a unique home health measure that is inconsistent with the intent of the IMPACT Act to standardize patient assessment data in post-acute care. If the intent of the IMPACT Act is to be able to compare patient outcomes and characteristics across post-acute care settings, the unique target population for home health care will confound the ability to achieve the goals of the IMPACT Act. The Alliance recommends that the target population for home health match that of the other settings so that only those admitted to home health within 30 days of discharge from an acute care hospital are included in the target population. The Alliance’s full comments on this measure can be found at: http://ahhqi.org/images/uploads/Alliance_Comments_on_Discharge_to_Comm_112315.pdf. In addition, the Alliance notes that for the other settings’ (SNF, IRF and LTCH) discharge to community measures, both to home health and discharge home (without home health) are considered a discharge to community. The Alliance urges stratification that will enable identification of those discharged to home health and those sent to home without home health care. This will enable improved analysis of provider performance and practice in bundled payment arrangements such as the comprehensive care for joint replacement model. The Alliance urges testing of this measure and reconsideration before it is finalized. Both this measure and the potentially preventable readmission measure are based on risk-adjusted estimates. Testing and validation should be no less than six months with an opportunity to modify the measure prior to finalizing it. A similar approach was used for many of the OASIS-based measures that CMS uses for home health agencies. (Submitted by: Alliance for Home Health Quality and Innovation)
  
  
• We support the goal of the IMPACT act to align quality measures across post-acute providers and to promote patient-centeredness in quality efforts. With respect to the draft measure developed by RTI we have particular concerns related to the definition of discharge to the community for other post-acute care providers that includes patients discharged from their setting with home health services while discharges from home health includes only those patients without home health services. Patients discharged from SNFs, IRFs, and LTCHs with home health services will have a greater chance of achieving the discharge to the community measure, not incurring an unplanned readmission, and potentially remaining alive for the 31 day period. Cross-setting comparisons of this quality measure would not be appropriate. Similarly, we believe that the inclusion of patients who did not have a hospitalization within the past 30 days only in the home health calculation and not for other post-acute care providers would create different measures that would not be comparable. We support the exclusion of patients discharged to hospice from the measure calculation. We have always disagreed with the Discharge to the Community OASIS-based quality measure’s definition that patients transferred to a non-institutional hospice are considered an unfavorable outcome. We would recommend the addition of any patient that elects hospice during the 31-day window and not just those that are discharged directly to hospice. Patients may be hospice-eligible but choose to delay their election for a variety of reasons. This measure does not acknowledge the role that personal care/non-medical home care services can provide in keeping a patient in the community. We support the collection of this information across the PAC providers and its inclusion in the risk adjustment model. We support the inclusion of this measure in the CASPER reports as soon as possible to provide agencies with data on their performance. (Submitted by: Association of Home & Hospice Care of NC/SC Home Care & Hospice Association)
  
  
• The ultimate goals of post-acute care are avoiding institutionalization and returning patients to their previous level of independence and functioning, with discharge to community being the primary goal for the majority of post-acute patients. For many, home is a symbol of independence, privacy, and competence. Discharge to community is considered a valuable outcome to measure because it is a multifaceted measure that captures the patient’s functional status, cognitive capacity, physical ability, and availability of social support at home. There is considerable variation in discharge to community rates within and across post-acute settings. Studies show geographic variation, variation across patient socioeconomic characteristics (for example, race and ethnicity), and variation by facility characteristics (for profit vs. nonprofit, freestanding vs. hospital-based, urban vs. rural). In the IRF setting, discharge to community rates vary across providers, ranging from about 60% to 75%. The 2015 MedPAC report shows that, in FY 2013, the facility-level, mean risk-adjusted discharge to community rate for IRFs within 100 days of admission was 75.8%, and the mean observed rate was 74.7%. Discharge to community rates also vary widely in the SNF setting, ranging from as low as 31% to as high as 65%. The 2015 MedPAC report shows a mean risk-adjusted discharge to community rate of 37.5% for SNFs within 100 days of admission, and mean observed rate of 40.1%. A multicenter study of 23 LTCHs reported that only 28.8% of 1,061 patients who were ventilator-dependent on admission were discharged to home or assisted living facility. A study of 66,510 Medicare beneficiaries during pre- and post-HH episodes, revealed that 64 percent of beneficiaries discharged from HH did not use any other Medicare-reimbursed acute or post-acute services in the 30 days following HH discharge. Significant numbers of patients were admitted to inpatient facilities (29percent) and lesser numbers to skilled nursing facilities (7.6 percent), inpatient rehabilitation (1.5 percent) and home health (7.2 percent) or hospice (3.3 percent) within 30 days of HH discharge (Wolff et al., 2008). II. Feedback Overall, there is not enough detail on the measure specifications to provide adequate feedback. However, despite the lack of detail, a response is included below on the discharge to community measure for LTCH, IRF, SNF, and HH. Description Patients discharged to the community following PAC stay/episode and do not have an unplanned admission to an acute care hospital or LTCH in 31 days after community discharge and remain alive • Although not explicitly stated, it appears that this measure is only holding the last PAC setting responsible for the discharge? Given what is known about multiple sites of PAC use either from LTCH/IRF patients are often discharged to SNF before community discharge, as it is currently worded this measure will penalize the SNF for the more difficult patients that could not be discharged from the prior PAC setting. • Research has shown that PAC patients re-enter the healthcare system through multiple doors after the PAC community discharge why are these not captured in the measure as a failed transition I recommend that the numerator should not include individuals who were admitted to a SNF after PAC community discharge during the 30-day measure window. Numerator & Denominator There is not enough detail to provide feedback on the risk adjustment of the numerator. Without details on the variables included, the origin of those variables, and methods of risk adjustment it is difficult to provide feedback. Recommend providing further detail. Recommend including functional status, functional cognition, presence of social support (e.g., does the individual live alone), and comorbid conditions as risk adjustors Exclusion There is no discussion of excluding long-stay nursing home residents in this measure. Long-stay residents that have an acute medical event, are admitted to an acute care hospital, discharged to PAC for care, and then transitioned back to long-term care should be excluded from the measure. These individuals are not returning to the community, while the nursing home is their home, the effort and processes required for transitioning a long-stay nursing home resident from PAC back to custodial care is very different that transitioning a community- residing older adult back to their residence in the community after a PAC stay. Recommend excluding long stay nursing home residents from the measure. (Submitted by: American Occupational Therapy Association)
  
  
• ARN is supportive of a measure that works to identify unplanned (re)admissions; however, we have concerns with identifying unplanned (re)admissions based on the planned readmissions algorithm used in NQF measure #2510: SNF 30-Day All-Cause Readmission Measure (SNFRM); NQF #2502: All-Cause Unplanned Readmission Measure for 30 Days Post Discharge from IRFs; NQF #2512: All-Cause Unplanned Readmission Measure for 30 Days Post Discharge from LTCHs; and NQF #2380: Re-hospitalization During the First 30 Days of Home Health. The exclusion criteria included within NQF #2510 for SNF stays where the patient had one or more intervening PAC admissions to an IRF that occurred either between the prior proximal hospital discharge and SNF admission or after the SNF discharge within the 30-day risk window fails to allow for a medically complex patient that is treated in an IRF and readmitted to the SNF within 30 days for a condition that may initially have been treated as a comorbidity. We disagree with the rationale provided for exclusion, for while the measure assesses readmission rates while accounting for patient demographics, principal diagnosis in the prior hospitalization, comorbidities, and other patient factors, often, this may not be the reason for admission to a SNF. ARN believes the measure should include the principal diagnosis during the prior proximal hospitalization, comorbidities based on the secondary medical diagnoses listed on the patient’s prior proximal hospital claim and diagnoses from prior hospitalizations that occurred in the previous 365 days, length of stay during the patient’s prior proximal hospitalization, length of stay in the intensive care unit (ICU), body system specific surgical indicators, End-Stage Renal Disease (ESRD) status, whether the patient was disabled, and the number of prior hospitalizations in the previous 365 days. It also would be beneficial to understand the comorbidities being evaluated in the risk-adjustment model. ARN urges the Centers for Medicare and Medicaid Services (CMS) to develop of a list of comorbidities, comparable to the IRF Prospective Payment System (PPS) list of comorbidities and we encourage CMS to categorize an intervening admission to an IRF as a proximal hospitalization. Additionally, ARN has serious concerns with CMS’s proposal to require PAC providers to utilize Medicare claims data to calculate their 30-day readmission rates. Using claims data to calculate readmission rates is difficult for health care providers, as claims data are cumbersome to use and access. Employing a 30-day readmission rate measure will not provide meaningful insight or have an impact on quality improvement efforts if the PAC settings do not have unrestricted access to the data. Further, ARN believes that patients who have been discharged to the community and expire within the post-discharge window should not be included within the quality measure, given the variation in patient characteristics across the four settings. For example, as compared to all Medicare beneficiaries, the SNF and LTCH patient populations represent the most disabled, elderly, and frail beneficiaries. The Medicare Payment Advisory Commission’s (MedPAC) March 2015 Report to Congress found that compared with other beneficiaries, “SNF users are older, frailer, and disproportionately female, disabled, living in an institution, and dually eligible for both Medicare and Medicaid.” Moreover, as compared with all Medicare beneficiaries, those admitted to LTCHs are “disproportionately disabled (under age 65), over age 85, or diagnosed with [ESRD]. They are also more likely to be African American.” ARN urges CMS to exclude patients that pass away within the post-discharge window after being discharged to the community from the discharge to community quality measure, as the types of patients treated in each setting greatly varies and can lead to an inaccurate reflection of the quality of care. ARN is pleased the CMS has proposed discharge measure exclusions; however, we have concerns with the proposed exclusion of post-acute stays that end in transfer to the same level of care, and specifically, CMS’s proposal to include only the final post-acute provider in the discharge to community measure. The Agency’s proposed exclusion criteria fails to consider when a patient’s “home” is a custodial nursing facility and the patient’s post-acute episode involves a discharge back to his or her “home.” In such circumstances, including the final post-acute provider in the discharge to community measure when a patient is discharged to the originating level of care, but in essence, is returning home, may distort the findings of the quality measure. We encourage CMS to design a quality measure that is capable of capturing the difference between a patient’s return to his or her home and a patient’s post-acute episode that involves transfer to the same level of care. (Submitted by: Association of Rehabilitation Nurses)
  
  
• This measure is currently in the early stage of development, has never been submitted to NQF and thus is not NQF-endorsed (Submitted by: AdvaMed)
  
  

• This measure is currently in early development and measure specifications are in preliminary stages of clarification. The proposed application could create inaccuracy in measurement of patient condition when used in other PAC settings. Terms have not been defined within the exclusions (i.e. "incomplete stays") or clarified regarding time parameters (i.e. “residents who are not Medicare Fee for Service beneficiaries” – does this mean they have never been, recently changed coverage prior to care or changed after care was provided?). No testing of the measure in non-IRF settings has been done. These could create inaccurate data capture throughout PAC venues. The Technical Expert Panel indicated that this measure needs extensive work to assure accuracy and to meet the purpose for which it was intended. Because there is potential for accuracy this measure should continue to be developed. (Submitted by: Genesis Health Care)
  
  
• AHCA is pleased to have the opportunity to comment on the List of Measures under Consideration for December 1, 2015”. The American Health Care Association (AHCA) represents more than 12,000 non-profit and proprietary skilled nursing centers and assisted living communities. Rather than having many of our individual members who have contacted us with comments and to more efficiently provide NQF with feedback, we have received from our various members, we have summarized their comments here. Thank you again for the opportunity to provide these comments. Please contact me at DGifford@ahca.org with any questions or for additional information. 1. While this measure has been endorsed by NQF, it was developed, tested and validated in the IRF setting for IRF patients. It was not developed for SNF setting. We do not think it is appropriate to take a measure from one setting and assume it is reliable and valid in another setting. If this logic is to be followed through, there are several other NQF endorsed functional measure that rely on the CARE tool item set, which is what is used for this measure. We believe, the MAP process must review and consider all NQF endorsed measures that are similar, in particular NQF measure 2612 and 2287. 2. In addition, the proposed measure relies on CARE tool items that were not added to section GG in the IRF, SNF or LTCH setting. As such, the specifications of this measure do not match the available data in the PAC assessment tools for SNF, IRF, and LTCH setting. 3. The specifications are based on ICD-9 but all providers as of October 2015 are required to use ICD-10 and no cross walk with ICD-10 is provided, yet the measure will be used during time periods when only ICD-10 data is available. 4. AHCA recommends that, before a measure developed in one setting is applied to another, that cross-setting measure testing be performed. Such testing is critically important because of the clinical sorting that goes on between the settings, with benefit rules, Medicare enrollment requirements, and the general suitability of the settings resulting in certain kinds of patients tending to go to SNFs, and other patients tending to go to IRFs. For example, the IRF benefit rule that patients can receive three or more hours of therapy each day means that certain groups of patients who would have difficulty walking will tend to sort into the SNF setting rather than the IRF setting. This phenomena will also occur on self-care domains. For example, differences in the proportion and types of stroke patients between SNF and IRF make application of an IRF measure inaccurate in the SNF setting. Because CMS has not tested whether the distribution of risk adjusted measure scores in the SNF setting are the same or widely different from those in the IRF setting, we are left with a serious concern that the measure’s distribution will be shifted up in the IRF setting and down in the SNF setting, without anything to do with the quality of care being provided in each setting. Last, we recognize that it is entirely possible that the measures perform adequately after cross-setting measure testing; however, if this testing is not performed before the measures are rolled out into live public use, CMS risks promoting a risk adjustment model that does not work as well for the SNFs compared to the IRF Quality Reporting Programs. Therefore, for these reasons, we believe this measure does not meet NQF MAP criteria for endorsement but rather should receive a vote of “encourage continued development” only after NQF MAP process has reviewed the other NQF endorsed functional measures to determine which the best to be recommended is. (Submitted by: American Health Care Association )
  
  
• This is a measure tested and validated in the IRF setting. it cannot be assumed to have validity in the SNF setting without testing. The populations are different and mobility may not be the goal for the SNF patient receiving skilled services. This measure also needs risk adjustment for acuity, as many patients are sent to SNF for skilled services related to complex medical care, in addition to rehab services - but these may not all address mobility. Thus risk adjustment for acuity, functional status prior to admission and pt goals much be taken into account (Submitted by: Leadingage)
  
  
• First, the IRF endorsement of this measure was conditional, requiring updates when data is available (after October 2016) to display the reliability and validity of the measure, among other things. We strongly urge the MAP PAC/LTC committee to consider whether or not to approve and/or implement a measure that has not been shown to meet NQF requirements for reliability and validity. Second, this measure should be identified as a competing measure with NQF #2612 - CARE: Improvement in Mobility, developed by AHCA and endorsed earlier this year by NQF. These measures are similar in that they utilize functional items creates as part of the CARE tool development project or PAC-PRD, but have differences in the functional items included, inclusion/exclusion criteria, and risk-adjustment methodologies. Third, a "best-in-class" determination related to the functional improvement or change in self-care is required and necessary. This measure has already been found to compete with a UDSMR developed measure (NQF #2321), and NQF staff chose not to make a "best-in-class" decision until such a time as additional measure testing was available. With the additional consideration for the AHCA measure that is already NQF endorsed noted previously, there exist 3 different ways for measuring the same functional construct. The IMPACT Act requirements indicate that measures for PAC settings should be "standardized and interoperable", yet implementation of the IMPACT Act is burdening PAC providers with the data collection of functional items for payment and similar but different functional items for quality. PAC providers should not bear the burden of data collection for measures that may not be identified as "best-in-class". (Submitted by: UDSMR)
  
  
• This measure was developed for the IRF and as such may not be usable in the SNF without further development and testing. The admission criteria for a patient entering an IRF is different than a SNF, therefore IRF patients may be different than SNF patients. Therefore, measuring mobility could use different items for a SNF patient than measuring mobility for an IRF patient. This measure requires testing and validation before it could be used for SNF patients. (Submitted by: National Association for the Support of Long Term Care)
  
  
• ARN is supportive of the IRF functional measures; however, there are several key differences between the existing and new proposed function items that may result in variation in the resident assessment results including: (1) data collection and associated data collection instructions; (2) rating scales used to score a resident’s level of independence; and (3) item definitions. Utilizing multiple data collection systems and instruments as well as a different rating scale and item definitions is both cumbersome and burdensome. Additionally, functional measures are not a “one size fits all,” and measures should account for the benefits of the quality of life domain for patients with these extreme conditions. ARN encourages the Agency to revise the measures to reflect the following attributes: a low collection burden for providers and beneficiaries; comprehensibility for beneficiaries; a high level of significance to patients and providers; and data that is routinely captured. (Submitted by: Association of Rehabilitation Nurses)
  
  
• AMDA is pleased to have the opportunity to comment on the List of Measures under Consideration for December 1, 2015”. AMDA - The Society for Post-Acute and Long-Term Care Medicine (AMDA) represents over 5,500 physician medical directors, attending physicians, and other practitioners who work in post-acute and long-term care settings. Thank you again for the opportunity to provide these comments. Please contact me at abardakh@amda.com with any questions or for additional information. 1. While this measure has been endorsed by NQF, it was developed, tested and validated in the IRF setting for IRF patients. It was not developed for SNF setting. We do not think it is appropriate to take a measure from one setting and assume it is reliable and valid in another setting. If this logic is to be followed through, there are several other NQF endorsed functional measure that rely on the CARE tool item set, which is what is used for this measure. We believe, the MAP process must review and consider all NQF endorsed measures that are similar. In particular, NQF measure 2612 and 2687 which rely on the CARE tool items. 2. In addition, the proposed measure relies on CARE tool items that were not added to section GG in the IRF, SNF or LTCH setting. As such, the specifications of this measure does not match the available data in the PAC assessment tools for SNF, IRF, and LTCH setting. 3. The specifications are based on ICD-9 but all providers as of October 2015 are required to use ICD-10 and no cross walk with ICD-10 is provided, yet the measure will be used during time periods when only ICD-10 data is available. 4. AMDA recommends that, before a measure developed in one setting is applied to another, that cross-setting measure testing be performed. Such testing is critically important because of the clinical sorting that goes on between the settings, with benefit rules, Medicare enrollment requirements, and the general suitability of the settings resulting in certain kinds of patients tending to go to SNFs, and other patients tending to go to IRFs. For example, the IRF benefit rule that patients can receive three or more hours of therapy each day means that certain groups of patients who would have difficulty walking will tend to sort into the SNF setting rather than the IRF setting. This phenomena will also occur on self-care domains. For example, differences in the proportion and types of stroke patients between SNF and IRF make application of an IRF measure inaccurate in the SNF setting. Because CMS has not tested whether the distribution of risk adjusted measure scores in the SNF setting are the same or widely different from those in the IRF setting, we are left with a serious concern that the measure’s distribution will be shifted up in the IRF setting and down in the SNF setting, without anything to do with the quality of care being provided in each setting. Last, we recognize that it is entirely possible that the measures perform adequately after cross-setting measure testing; however, if this testing is not performed before the measures are rolled out into live public use, CMS risks promoting a risk adjustment model that does not work as well for the SNFs compared to the IRF Quality Reporting Programs. Therefore, for these reasons, we believe this measure does not meet NQF MAP criteria for endorsement but rather should receive a vote of “encourage continued development” only after NQF MAP process has reviewed the other NQF endorsed functional measure to determine which the best to be recommended is. (Submitted by: AMDA-The Society for Post-Acute and Long-Term Care Medicine)
  
  
• This measure is currently in the early stage of development, has never been submitted to NQF and thus is not NQF-endorsed (Submitted by: AdvaMed)
  
  
• At least one NRHI member has expressed concerns regarding this measure. Specifically, clinical staff at HealthInsight have concerns with designation of functional status scores such as "self-care score" and "mobility score" that will rate the performance of facilities. (Submitted by: Network for Regional Healthcare Improvement)
  
  

• This measure is currently in early development and measure specifications are in preliminary stages of clarification. The proposed application could create inaccuracy in measurement of patient condition when used in other PAC settings. Terms have not been defined within the exclusions (i.e. "incomplete stays") and patient coverage parameters have not been clarified (i.e. “residents not covered by the Medicare Fee for Service program”---does this mean never under Medicare or having recently changed from Medicare or having changed coverage after services were provided?). Also, there is no clinical explanation or justification for excluding patients discharged to hospice. No testing of the measure in non-IRF settings has been done. These could create inaccurate data capture throughout PAC venues. The Technical Expert Panel indicated that this measure needs work to assure accuracy and to meet the purpose for which it was intended. Because there is potential for accuracy this measure should continue to be developed. (Submitted by: Genesis Health Care)
  
  
• AHCA is pleased to have the opportunity to comment on the List of Measures under Consideration for December 1, 2015”. The American Health Care Association (AHCA) represents more than 12,000 non-profit and proprietary skilled nursing centers and assisted living communities. Rather than having many of our individual members who have contacted us with comments and to more efficiently provide NQF with feedback, we have received from our various members, we have summarized their comments here. Thank you again for the opportunity to provide these comments. Please contact me at DGifford@ahca.org with any questions or for additional information. 1. While this measure has been endorsed by NQF, it was developed, tested and validated in the IRF setting for IRF patients. It was not developed for SNF setting. We do not think it is appropriate to take a measure from one setting and assume it is reliable and valid in another setting. If this logic is to be followed through, there are several other NQF endorsed functional measure that rely on the CARE tool item set, which is what is used for this measure. We believe, the MAP process must review and consider all NQF endorsed measures that are similar, in particular NQF measures 2613 and 2286. 2. In addition, the proposed measure relies on CARE tool items that were not added to section GG in the IRF, SNF or LTCH setting. As such, the specifications of this measure do not match the available data in the PAC assessment tools for SNF, IRF, and LTCH setting. 3. AHCA recommends that, before a measure developed in one setting is applied to another, that cross-setting measure testing be performed. Such testing is critically important because of the clinical sorting that goes on between the settings, with benefit rules, Medicare enrollment requirements, and the general suitability of the settings resulting in certain kinds of patients tending to go to SNFs, and other patients tending to go to IRFs. For example, the IRF benefit rule that patients can receive three or more hours of therapy each day means that certain groups of patients who would have difficulty walking will tend to sort into the SNF setting rather than the IRF setting. This phenomena will also occur on self-care domains. For example, differences in the proportion and types of stroke patients between SNF and IRF make application of an IRF measure inaccurate in the SNF setting Because CMS has not tested whether the distribution of risk adjusted measure scores in the SNF setting are the same or widely different from those in the IRF setting, we are left with a serious concern that the measure’s distribution will be shifted up in the IRF setting and down in the SNF setting, without anything to do with the quality of care being provided in each setting. Last, we recognize that it is entirely possible that the measures perform adequately after cross-setting measure testing; however, if this testing is not performed before the measures are rolled out into live public use, CMS risks degrading the integrity of the SNF and IRF Quality Reporting Programs by introducing measures that may become publicly known to be unusable for their intended purpose, or at worst need to be retracted and redeveloped. 4. The specifications are based on ICD-9 but all providers as of October 2015 are required to use ICD-10 and no cross walk with ICD-10 is provided, yet the measure will be used during time periods when only ICD-10 data is available. Therefore, for these reasons, we believe this measure does not meet NQF MAP criteria for endorsement but rather should receive a vote of “encourage continued development” only after NQF MAP process has reviewed the other NQF endorsed functional measures to determine which the best to be recommended is. (Submitted by: American Health Care Association )
  
  
• While this measure was previously endorsed for IRFs, we would note that endorsement of a similar measure (NQF #2633) utilizing the same measure items and risk adjustment methodology was conditional with requirements for updates on the reliability and validity of the measure. We strongly urge the MAP committee to review and determine whether or not this measure meets NQF requirements for reliability and validity, especially in the SNF setting. Second, we would note that the functional items identified in this measure differ from those endorsed to measure improvement in self-care as developed by AHCA. While a majority of these items are the same, we question which set of items can be shown to produce reliable and valid results for measuring quality. Finally, we are very concerned that the implementation of these measures is causing the need to collect one set of functional items for payment and a similar but different set of functional items for quality. Until such a time as a best-in-class decision is made on which functional items are most representative of quality and payment, PAC providers are burdened with dual data collection. (Submitted by: UDSMR)
  
  
• This measure compares an actual score with an expected score. This is an IRF measure that is proposed to be used for SNF. SNF patients are going to have different expected scores than IRF patients. So, this IRF measure cannot be used for SNF patients without testing its reliability and validity. (Submitted by: National Association for the Support of Long Term Care)
  
  
• ARN is supportive of the IRF functional measures; however, there are several key differences between the existing and new proposed function items that may result in variation in the resident assessment results including: (1) data collection and associated data collection instructions; (2) rating scales used to score a resident’s level of independence; and (3) item definitions. Utilizing multiple data collection systems and instruments as well as a different rating scale and item definitions is both cumbersome and burdensome. Additionally, functional measures are not a “one size fits all,” and measures should account for the benefits of the quality of life domain for patients with these extreme conditions. ARN encourages the Agency to revise the measures to reflect the following attributes: a low collection burden for providers and beneficiaries; comprehensibility for beneficiaries; a high level of significance to patients and providers; and data that is routinely captured. (Submitted by: Association of Rehabilitation Nurses)
  
  
• AMDA is pleased to have the opportunity to comment on the List of Measures under Consideration for December 1, 2015”. AMDA - The Society for Post-Acute and Long-Term Care Medicine (AMDA) represents over 5,500 physician medical directors, attending physicians, and other practitioners who work in post-acute and long-term care settings. Thank you again for the opportunity to provide these comments. Please contact me at abardakh@amda.com with any questions or for additional information. 1. While this measure has been endorsed by NQF, it was developed, tested and validated in the IRF setting for IRF patients. It was not developed for SNF setting. We do not think it is appropriate to take a measure from one setting and assume it is reliable and valid in another setting. If this logic is to be followed through, there are several other NQF endorsed functional measure that rely on the CARE tool item set, which is what is used for this measure. We believe, the MAP process must review and consider all NQF endorsed measures that are similar. In particular, NQF measure 2613 and 2686 which rely on the CARE tool items. 2. AMDA recommends that, before a measure developed in one setting is applied to another, that cross-setting measure testing be performed. Such testing is critically important because of the clinical sorting that goes on between the settings, with benefit rules, Medicare enrollment requirements, and the general suitability of the settings resulting in certain kinds of patients tending to go to SNFs, and other patients tending to go to IRFs. For example, the IRF benefit rule that patients can receive three or more hours of therapy each day means that certain groups of patients who would have difficulty walking, and/or significant deconditioning, functional quadriplegia, dysphagia, or gastrostomy, will tend to sort into the SNF setting rather than the IRF setting. This phenomena will also occur on self-care domains. For example, differences in the proportion and types of stroke patients between SNF and IRF make application of an IRF measure inaccurate in the SNF setting Because CMS has not tested whether the distribution of risk adjusted measure scores in the SNF setting are the same or widely different from those in the IRF setting, we are left with a serious concern that the measure’s distribution will be shifted up in the IRF setting and down in the SNF setting, without anything to do with the quality of care being provided in each setting. Last, we recognize that it is entirely possible that the measures perform adequately after cross-setting measure testing; however, if this testing is not performed before the measures are rolled out into live public use, CMS risks degrading the integrity of the SNF and IRF Quality Reporting Programs by introducing measures that may become publicly known to be unusable for their intended purpose, or at worst need to be retracted and redeveloped. 3. In addition, the proposed measure relies on CARE tool items that were not added to section GG in the IRF, SNF or LTCH setting. As such, the specifications of this measure does not match the available data in the PAC assessment tools for SNF, IRF, and LTCH setting. 4. The specifications are based on ICD-9 but all providers as of October 2015 are required to use ICD-10 and no cross walk with ICD-10 is provided, yet the measure will be used during time periods when only ICD-10 data is available. 5. AMDA recommends that, before a measure developed in one setting is applied to another, that cross-setting measure testing be performed. Such testing is critically important because of the clinical sorting that goes on between the settings, with benefit rules, Medicare enrollment requirements, and the general suitability of the settings resulting in certain kinds of patients tending to go to SNFs, and other patients tending to go to IRFs. For example, the IRF benefit rule that patients can receive three or more hours of therapy each day means that certain groups of patients who would have difficulty walking will tend to sort into the SNF setting rather than the IRF setting. This phenomena will also occur on self-care domains. For example, differences in the proportion and types of stroke patients between SNF and IRF make application of an IRF measure inaccurate in the SNF setting. Because CMS has not tested whether the distribution of risk adjusted measure scores in the SNF setting are the same or widely different from those in the IRF setting, we are left with a serious concern that the measure’s distribution will be shifted up in the IRF setting and down in the SNF setting, without anything to do with the quality of care being provided in each setting. Last, we recognize that it is entirely possible that the measures perform adequately after cross-setting measure testing; however, if this testing is not performed before the measures are rolled out into live public use, CMS risks promoting a risk adjustment model that does not work as well for the SNFs compared to the IRF Quality Reporting Programs. Therefore, for these reasons, we believe this measure does not meet NQF MAP criteria for endorsement but rather should receive a vote of “encourage continued development” only after NQF MAP process has reviewed the other NQF endorsed functional measure to determine which the best to be recommended is. (Submitted by: AMDA-The Society for Post-Acute and Long-Term Care Medicine)
  
  
• The measure does not appear to consider performance skills and cognitive capacities of self-care such as sequencing, problem-solving, temporal appropriateness (e.g., whether to dress for day or bed), memory, and activity planning. Nor does the measure consider performance of activities of daily living, including the broader instrumental activities of daily living (IADLs), which significantly impact a patient’s ability to function and live independently in the community. This IRF measure has been risk adjusted for the SNF setting. It remains unclear how this measure is distinctly different from the "Change in" Self Care measure. (Submitted by: American Occupational Therapy Association)
  
  
• This measure is currently in the early stage of development, has never been submitted to NQF and thus is not NQF-endorsed (Submitted by: AdvaMed)
  
  
• At least one NRHI member has expressed concerns regarding this measure. Specifically, clinical staff at HealthInsight have concerns with designation of functional status scores such as "self-care score" and "mobility score" that will rate the performance of facilities. (Submitted by: Network for Regional Healthcare Improvement)
  
  

• This measure is currently in early development and measure specifications are in preliminary stages of clarification. The proposed application could create inaccuracy in measurement of patient condition when used in other PAC settings. Terms have not been defined within the exclusions (i.e. "incomplete stays") and patient coverage parameters have not been clarified (i.e. “residents not covered by the Medicare Fee for Service program”---does this mean never under Medicare or having recently changed from Medicare or having changed coverage after services were provided?). Also, there is no clinical explanation or justification for excluding patients discharged to hospice. No testing of the measure in non-IRF settings has been done. These could create inaccurate data capture throughout PAC venues. The Technical Expert Panel indicated that this measure needs work to assure accuracy and to meet the purpose for which it was intended. Because there is potential for accuracy this measure should continue to be developed. (Submitted by: Genesis Health Care)
  
  
• While this is a validated measure, it is only tested in the IRF setting. Furthermore, it is not relying on assessment data available in the PAC setting. the measure does not adjust for clinical risk, nor does it take into account an individual's goals for outcome. While IRF patients are selected because of their ability to participate in 3 or more hours of therapy, SNF pts may have much more modest functional goals, and also may be discharge to home with home health before 100% of goals are met. (Submitted by: LeadingAge)
  
  
• AHCA is pleased to have the opportunity to comment on the List of Measures under Consideration for December 1, 2015”. The American Health Care Association (AHCA) represents more than 12,000 non-profit and proprietary skilled nursing centers and assisted living communities. Rather than having many of our individual members who have contacted us with comments and to more efficiently provide NQF with feedback, we have received from our various members, we have summarized their comments here. Thank you again for the opportunity to provide these comments. Please contact me at DGifford@ahca.org with any questions or for additional information. 1. While this measure has been endorsed by NQF, it was developed, tested and validated in the IRF setting for IRF patients. It was not developed for SNF setting. We do not think it is appropriate to take a measure from one setting and assume it is reliable and valid in another setting. If this logic is to be followed through, there are several other NQF endorsed functional measure that rely on the CARE tool item set, which is what is used for this measure. We believe, the MAP process must review and consider all NQF endorsed measures that are similar, in particular NQF measures 2612 and 2287. 2. In addition, the proposed measure relies on CARE tool items that were not added to section GG in the IRF, SNF or LTCH setting. As such, the specifications of this measure do not match the available data in the PAC assessment tools for SNF, IRF, and LTCH setting. 3. AHCA recommends that, before a measure developed in one setting is applied to another, that cross-setting measure testing be performed. Such testing is critically important because of the clinical sorting that goes on between the settings, with benefit rules, Medicare enrollment requirements, and the general suitability of the settings resulting in certain kinds of patients tending to go to SNFs, and other patients tending to go to IRFs. For example, the IRF benefit rule that patients can receive three or more hours of therapy each day means that certain groups of patients who would have difficulty walking will tend to sort into the SNF setting rather than the IRF setting. Because CMS has not tested whether the distribution of risk adjusted measure scores in the SNF setting are the same or widely different from those in the IRF setting, we are left with a serious concern that the measure’s distribution will be shifted up in the IRF setting and down in the SNF setting, without anything to do with the quality of care being provided in each setting. Last, we recognize that it is entirely possible that the measures perform adequately after cross-setting measure testing; however, if this testing is not performed before the measures are rolled out into live public use, CMS risks degrading the integrity of the SNF and IRF Quality Reporting Programs by introducing measures that may become publicly known to be unusable for their intended purpose, or at worst need to be retracted and redeveloped. 4. The specifications are based on ICD-9 but all providers as of October 2015 are required to use ICD-10 and no cross walk with ICD-10 is provided, yet the measure will be used during time periods when only ICD-10 data is available. Therefore, for these reasons, we believe this measure does not meet NQF MAP criteria for endorsement but rather should receive a vote of “encourage continued development” only after NQF MAP process has reviewed the other NQF endorsed functional measures to determine which the best to be recommended is. (Submitted by: American Health Care Association )
  
  
• While this measure was previously endorsed for IRFs, we would note that endorsement of a similar measure (NQF #2634) utilizing the same measure items and risk adjustment methodology was conditional with requirements for updates on the reliability and validity of the measure. We strongly urge the MAP committee to review and determine whether or not this measure meets NQF requirements for reliability and validity, especially in the SNF setting. Second, we would note that the functional items identified in this measure differ from those endorsed to measure improvement in mobility as developed by AHCA. While a majority of these items are the same, we question which set of items can be shown to produce reliable and valid results for measuring quality. Finally, we are very concerned that the implementation of these measures is causing the need to collect one set of functional items for payment and a similar but different set of functional items for quality. Until such a time as a best-in-class decision is made on which functional items are most representative of quality and payment, PAC providers are burdened with dual data collection. (Submitted by: UDSMR)
  
  
• This measure compares an expected score against an actual score. This measure was designed for the IRF patient and is proposed to be used for the SNF patient. The SNF patient will have a different expected score than the IRF patient. This measure cannot be transferred to be used in the SNF without testing reliability and validity. (Submitted by: National Association for the Support of Long Term Care)
  
  
• ARN is supportive of the IRF functional measures; however, there are several key differences between the existing and new proposed function items that may result in variation in the resident assessment results including: (1) data collection and associated data collection instructions; (2) rating scales used to score a resident’s level of independence; and (3) item definitions. Utilizing multiple data collection systems and instruments as well as a different rating scale and item definitions is both cumbersome and burdensome. Additionally, functional measures are not a “one size fits all,” and measures should account for the benefits of the quality of life domain for patients with these extreme conditions. ARN encourages the Agency to revise the measures to reflect the following attributes: a low collection burden for providers and beneficiaries; comprehensibility for beneficiaries; a high level of significance to patients and providers; and data that is routinely captured. (Submitted by: Association of Rehabilitation Nurses)
  
  
• AMDA is pleased to have the opportunity to comment on the List of Measures under Consideration for December 1, 2015”. AMDA - The Society for Post-Acute and Long-Term Care Medicine (AMDA) represents over 5,500 physician medical directors, attending physicians, and other practitioners who work in post-acute and long-term care settings. Thank you again for the opportunity to provide these comments. Please contact me at Abardakh@amda.com with any questions or for additional information. 1. While this measure has been endorsed by NQF, it was developed, tested and validated in the IRF setting for IRF patients. It was not developed for SNF setting. We do not think it is appropriate to take a measure from one setting and assume it is reliable and valid in another setting. If this logic is to be followed through, there are several other NQF endorsed functional measure that rely on the CARE tool item set, which is what is used for this measure. We believe, the MAP process must review and consider all NQF endorsed measures that are similar. In particular, NQF measure 2612 and 2687 which rely on the CARE tool items. 2. In addition, the proposed measure relies on CARE tool items that were not added to section GG in the IRF, SNF or LTCH setting. As such, the specifications of this measure does not match the available data in the PAC assessment tools for SNF, IRF, and LTCH setting. 3. AMDA recommends that, before a measure developed in one setting is applied to another, that cross-setting measure testing be performed. Such testing is critically important because of the clinical sorting that goes on between the settings, with benefit rules, Medicare enrollment requirements, and the general suitability of the settings resulting in certain kinds of patients tending to go to SNFs, and other patients tending to go to IRFs. For example, the IRF benefit rule that patients can receive three or more hours of therapy each day means that certain groups of patients who would have difficulty walking, and/or significant deconditioning, functional quadriplegia, dysphagia, or gastrostomy, will tend to sort into the SNF setting rather than the IRF setting. Because CMS has not tested whether the distribution of risk adjusted measure scores in the SNF setting are the same or widely different from those in the IRF setting, we are left with a serious concern that the measure’s distribution will be shifted up in the IRF setting and down in the SNF setting, without anything to do with the quality of care being provided in each setting. Last, we recognize that it is entirely possible that the measures perform adequately after cross-setting measure testing; however, if this testing is not performed before the measures are rolled out into live public use, CMS risks degrading the integrity of the SNF and IRF Quality Reporting Programs by introducing measures that may become publicly known to be unusable for their intended purpose, or at worst need to be retracted and redeveloped. 4. In addition, the proposed measure relies on CARE tool items that were not added to section GG in the IRF, SNF or LTCH setting. As such, the specifications of this measure does not match the available data in the PAC assessment tools for SNF, IRF, and LTCH setting. 5. The specifications are based on ICD-9 but all providers as of October 2015 are required to use ICD-10 and no cross walk with ICD-10 is provided, yet the measure will be used during time periods when only ICD-10 data is available. Therefore, for these reasons, we believe this measure does not meet NQF MAP criteria for endorsement but rather should receive a vote of “encourage continued development” only after NQF MAP process has reviewed the other NQF endorsed functional measure to determine which the best to be recommended is. (Submitted by: AMDA-The Society for Post-Acute and Long-Term Care Medicine)
  
  
• This measure is currently in the early stage of development, has never been submitted to NQF and thus is not NQF-endorsed (Submitted by: AdvaMed)
  
  
• At least one NRHI member has expressed concerns regarding this measure. Specifically, clinical staff at HealthInsight have concerns with designation of functional status scores such as "self-care score" and "mobility score" that will rate the performance of facilities. (Submitted by: Network for Regional Healthcare Improvement)
  
  

• This measure is currently in the early stage of development, has never been submitted to NQF and thus is not NQF-endorsed (Submitted by: AdvaMed)
  
  

Navigating the Discussion Guide

• How do I get back to the section I was just looking at?
  The easiest way is to use the back button on your browser. Other options are using your backspace button (which works for many browsers on laptops), or using the permanent links at the upper right hand corner of the discussion guide. But the back button is the best choice in most situations.
  
  
• Can I print the discussion guide out?
  You can, but we don't recommend it. Besides using a lot of paper (probably a couple hundred pages at least), you'll lose all the links that allow you to move around the document. For instance, if you're scrolling through the agenda and want to see more information about a particular measure, the electronic format will allow you to click a link, read more, and then bo back. If you're on paper, there will be a lot of flipping through paper.
  
  
• If I can't print this out, how can I read it on the plane?
  We will send you a pdf/Adobe Acrobat file a few days before the meeting, which will hopefully be useful when you're reviewing the discussion guide as you travel to Washington, DC.
  
  
• How do I know that I'm looking at the most recent version?
  At the top left corner of the discussion guide is a version number. At the beginning of the in person meetings, the NQF staff will ask everyone to load the most recent discussion guide version and will check that everyone has the same version loaded.
  
  
• What electronic devices can I use to view the discussion guide?
  We tried to make this as universal as possible, so it should work on your laptop (PC, Mac, Linux), your tablet (iPad, Android), or your phone (iPhone, Android). It should also work on many types of browsers (IE, Firefox, Chrome, Safari, Opera, Dolphin,....). Please let us know if you have any problems, and we'll troubleshoot with you (and improve the discussion guide for the next go around).
  
  
• Why do I see weird characters in some places?
  Because we're joining data from many different sources, we do find some technical challenges. This generally shows up as strange characters--extra question marks, accented characters, or otherwise unusual items. We've been able to fix many of these problems, but not all. We ask that you bear with us as we improve this over time!
  
  

Content

• What is included in the discussion guide?
  There are four sections within this document:
  • Agenda, with summaries of each measure under consideration
  • Full information about each measure, including its specifications, preliminary analysis of how this measure can advance the program's goals, and the rationale by HHS for being included in the list
  • Summaries for each federal health program being considered
  • Public comments that have been received to date (Note that the discussion guide may be released before the public comment period is finished, in which case there will just be a placeholder for where comments will go)
  
  
• How are the meeting discussions organized?
  The meeting sessions are organized around consent calendars, which are groups of measures being considered for a particular program or groups of measures for a particular condition or topic area. For each measure being discussed, this document will show you the description, the public comments (if any), the summary of the preliminary analysis, and the result of the preliminary analysis algorithm.

Remote Participation Instructions:

• Direct your web browser to: http://nqf.commpartners.com/.
• Under “Enter a Meeting” type in the meeting number for Day 1: 162261 or for Day 2: 522440
• In the “Display Name” field, type in your first and last names and click “Enter Meeting.”

• Dial 888-802-7237 for workgroup members or 877-303-9138 for public participants ; use conference ID code for Day 1: 76199779 or Day 2: 76199828 to access the audio platform.