E
Pay for Performance in Various Care Settings
Each care setting—dialysis facilities, hospitals, ambulatory physicians, health plans, home health agencies, and skilled nursing facilities—has specific characteristics that need to be considered when planning a pay-for-performance program. This appendix briefly describes each care setting and discusses how rewards could be distributed.
REWARDING DIALYSIS FACILITIES
Background
The treatment of end-stage renal disease (ESRD) is unique in that almost all ESRD patients are covered under Medicare, with only minimal coverage being provided by the private sector or out-of-pocket payment by beneficiaries. Because of the historical placement of dialysis facilities in the Medicare program, however, payment issues are complicated by the fact that payments come from both Part A and Part B. Facility payments (Part A) are capitated, but other, Part B services may be reimbursed in addition to the facility payment. The committee believes that the following three domains should be the focus of initial efforts to provide rewards to dialysis facilities (see Chapter 4):
-
Clinical quality: Since 1988, a partnership comprising the National Institutes of Health, the Centers for Medicare and Medicaid Services (CMS), and the United States Renal Data System has acted to collect data on this population. (The United States Renal Data System tracks the incidence and
-
prevalence of ESRD and acts to drive the ESRD research agenda.) Five of these measures are currently collected for the Agency for Healthcare Research and Quality’s (AHRQ) National Healthcare Quality Report (NHQR). Three of these are outcome measures derived from the University of Michigan. The other two are process measures from the United States Renal Data System. Additionally, the three outcome measures are currently reported on CMS’s Dialysis Facility Compare website. Also, as a requirement for payment, all facilities must already be reporting on hematocrit levels as a part of normal reimbursement procedures. The committee believes that dialysis facilities should begin reporting on the measures collected for AHRQ’s NHQR (five measures), which could be combined into an equally weighted composite score.
-
Patient-centeredness: Patients’ experiences of care will be measured by a Consumer Assessment of Healthcare Providers and Systems (CAHPS) survey. As of August 2006, an In-Center Hemodialysis (ICH) CAHPS was being finalized to capture the patient’s perspective of care in dialysis facilities.
-
Efficiency: There is a dearth of efficiency measures, and until valid measures are developed, the committee believes that a system should be developed in which dialysis facilities meeting certain thresholds for both clinical quality and patient-centeredness measures are given an additional reward if they are among the most efficient one-third of dialysis facilities. The most efficient third could be calculated using methods for calculating standardized costs for Medicare. For example, Medicare standardized costs over time would be calculated using charges for Medicare Parts A and B (starting at the time of hospitalization and following charges for 90 days) using standard national prices such as an “average” payout per diagnosis-related group or resource-based relative value scale. This would be uniform for all providers and would not include disproportiante share or graduate medical education payments. Using this method, efficiency could be rewarded only when both clinical quality and patient-centered measures were available.
Timing of Pay for Performance
For dialysis facilities, measurement of the three domains is at different levels of development. As the dialysis facilities have been reporting on clinical quality measures as discussed above, the committee believes that rewards could be provided for meeting performance criteria in this domain at the beginning of year 2 (2009). As ICH-CAHPS data were being collected, dialysis facilities would be rewarded for publicly reporting performance data through Dialysis Facility Compare. Beginning in year 3 (2010), as ICH-CAHPS data became available, patient-centeredness would be rewarded based on performance. Efficiency measures could thus begin to be rewarded
TABLE E-1 Dialysis Facility Phasing
|
Year 1 (2008) |
Year 2 (2009) |
Year 3 (2010) |
Clinical quality |
NHQR measures— pay for public reporting |
NHQR measures— pay for performance |
NHQR measures— pay for performance |
Patient-centeredness |
|
ICH-CAHPS— pay for public reporting |
ICH-CAHPS—pay for performance |
Efficiency |
|
|
Additional payout to the most efficient 1/3 of facilities meeting thresholds for both clinical quality and patient-centeredness measures |
only beginning in year 3. As more measures for each domain were developed, they would be considered for payment based first on public reporting and then on performance. Rewards for public reporting would be smaller than those for performance (see Table E-1).
Example of Pay for Performance for Dialysis Facilities
Measurement of the performance of dialysis facilities and physicians treating ESRD would be based on clinical quality, patient-centeredness, and efficiency. The clinical quality measures would include the following:
Outcomes/Process Measures Table
Outcome Measures |
Process Measures |
% of hemodialysis patients with urea reduction ratio of 65 or greater |
% of dialysis patients registered on a waiting list for transplantation |
% of patients with hematocrit of 33 or greater |
% of patients with treated chronic kidney failure who receive a transplant within 3 years of renal failure |
Patient survival rate |
|
-
Quality: A composite score for treatment of ESRD would be assessed to determine whether patients received all the care they should have received. Each measure could be equally weighted, and a straight average could be taken for all five measures.
-
Resource use: To capture resource use, the committee chose to use longitudinal measures (risk-adjusted mortality) and resource use (standard-
-
ized costs for Medicare Parts A and B) for all patients at a given dialysis facility.
-
Eligibility for rewards: The dialysis facility and physicians who billed above a threshold number of evaluation and management (E&M) claims during the subsequent year would be eligible.
-
How to distribute rewards: There are two choices: (1) rewards could go to dialysis facilities for distribution; or (2) rewards could be split between dialysis facilities (X percent) and physicians (100–X percent), with physician rewards being distributed in proportion to the share of E&M claims.
REWARDING HOSPITALS
Background
Hospitals participating in Medicare are reimbursed by both Medicare Parts A and B. Part A covers facility use, while Part B covers physician payments. Since the adoption of the Medicare Prescription Drug, Improvement, and Modernization Act of 2003 (Public Law 108-173), approximately 4,200 U.S. hospitals have been reporting data on a set of measures agreed upon by the Hospital Quality Alliance (HQA) (hospitalcompare.hhs.gov).
-
Clinical quality: Previously, hospitals had to report on a set of 10 measures to receive 0.4 percent of their Medicare reimbursement. Now, hospitals are reporting on a set of 20 measures. As these are widely endorsed measures (face validity and relatively strong evidence base), the committee believes that the most recent version of the HQA measure set should be the basis for rewards. These measures should then be combined into a composite for each condition; the composite could be calculated as a sum of the scores of each measure for the given condition. These measures would cut across settings when applicable (see the example for acute myocardial infarction [AMI] in the section below on ambulatory physician care).
-
Patient-centeredness: Currently the best measures of patients’ perspectives on the care they receive derive from the Hospital CAHPS survey. The Hospital CAHPS survey instrument was recently validated and approved for use by CMS. Hospitals have not consistently been collecting and reporting the results of this survey; training will be completed and results will begin to be collected by 2007.
-
Efficiency: See the above section on dialysis facility care for one possible method.
Timing of Pay for Performance
For hospitals, measurement of the three domains is at different levels of development. As the majority of hospitals have been reporting on clinical
TABLE E-2 Hospital Phasing
|
Year 1 |
Year 2 |
Year 3 |
Clinical quality |
Hospital Quality Alliance measures— pay for public reporting |
Hospital Quality Alliance measures— pay for performance |
Hospital Quality Alliance measures— pay for performance |
Patient-centeredness |
Hospital CAHPS— pay for public reporting |
Hospital CAHPS— pay for performance |
Hospital CAHPS— pay for performance |
Efficiency |
|
Additional payout to the most efficient 1/3 of hospitals meeting thresholds for both clinical quality and patient-centeredness measures |
Additional payout to the most efficient 1/3 of hospitals meeting thresholds for both clinical quality and patient-centeredness measures |
quality measures as discussed above, the committee believes rewards could be provided for meeting performance criteria for these measures beginning in year 2 (2009). As Hospital CAHPS would just have gotten off the ground, hospitals could be rewarded for publicly reporting these data through Hospital Compare. Beginning in year 2, patient-centeredness could be rewarded based on performance. Efficiency measures could thus begin to be rewarded only beginning in year 2. As more measures for each dimension were developed, they could be considered for payment based first on public reporting and then on performance. Rewards for public reporting would be smaller than those for performance (see Table E-2).
Example of Hospital Pay for Performance
See the example of AMI at the end of the section on ambulatory physicians below.
REWARDING AMBULATORY CARE
Background
Efforts to hold individual physicians accountable for the care they provide are in their early stages because of the basic difficulties involved and the fact that such efforts have never been undertaken on a large scale. There have, however, been some successful smaller-scale examples. An important step toward being able to attribute care in ambulatory settings is a collabo-
rative initiative launched in January 2006 by CMS—the Physician Voluntary Reporting Program. No portion of physician reimbursements has been linked to this initiative (physicians are paid out of Medicare Part B).
-
Clinical quality: The committee believes that physicians should begin reporting on the measures proposed for the starter set of Ambulatory care Quality Alliance (AQA) measures (currently consisting of 26 measures). Over time, as reporting became more widespread, these measures could be aligned with those used in the Physician Voluntary Reporting Program (currently consisting of 36 measures) as appropriate. Composite scores should be created for each condition.
-
Patient-centeredness: Patients’ experiences of care are currently measured by the CAHPS survey. A survey specifically targeting patients seen at the individual clinician level as part of the Ambulatory CAHPS survey was expected to be released by end of 2006.
-
Efficiency: See the earlier discussion of dialysis facilities for one possible method.
Timing of Pay for Performance
Reporting on quality measures is not widespread at the level of the individual physician as the basic infrastructure needed to collect and report these data has not yet been broadly adopted at this level. Therefore, the committee proposes the following timeline for implementation.
The first part of year 1 (2008) would likely be spent finalizing performance measures for clinical quality and collecting data, with data cleanup and validation in the latter part of the year. In year 2 (2009), performance reports would be distributed to physicians, and feedback could be provided to reporting physicians, which would be followed by paying for public reporting. In year 3, physicians would be rewarded based on their level of performance. Patient-centered measures for physicians would not be ready for widespread use until 2007; data would have to be collected in the beginning of year 1, and therefore payment for performance based on the results of these patient experience surveys would begin in year 3. Efficiency would be rewarded on for the top one-third of physicians in the nation meeting thresholds for both clinical quality and patient-centeredness. Additional measures would follow the same timeline for implementation as soon as they had been deemed valid (see Table E-3).
Example for Acute Myocardial Infarction
One example of an episode of care that could be paid for according to performance in year 1 is AMI. Measurement of the performance of the
TABLE E-3 Ambulatory Phasing
|
Year 1 (2008) |
Year 2 (2009) |
Year 3 (2010) |
Clinical quality |
|
Data back to providers/ feedback period and Ambulatory care Quality Alliance measures— pay for public reporting |
Ambulatory care Quality Alliance measures— pay for performance |
Patient-centeredness |
|
Ambulatory CAHPS— Pay for public reporting |
Ambulatory CAHPS— pay for performance |
Efficiency |
|
|
Additional payout to the most efficient 1/3 of physicians meeting thresholds for both clinical quality and patient-centeredness measures |
physicians treating such patients would be based on clinical quality, patient-centeredness, and efficiency. The clinical quality measures would include the following:
Acute Myocardial Infarction Measures
Hospital Quality Alliance |
Ambulatory care Quality Alliance |
Aspirin at arrival for acute myocardial infarction (AMI) Aspirin prescribed at discharge for AMI Beta-blocker at arrival for AMI Beta-blocker prescribed at discharge for AMI AMI inpatient mortality Angiotensin-converting enzyme (ACE) inhibitor for left ventricular systolic dysfunction (LVSD) Percutaneous coronary intervention (PCI) within 120 minutes of arrival for AMI Thrombolytic agent within 30 minutes of arrival for AMI |
Drug therapy for lowering LDL cholesterol Beta-blocker treatment after heart attack Beta-blocker treatment post– myocardial infarction |
Acute Myocardial Infarction Measures
-
Quality: A composite score for AMI in each care setting could be formulated to determine whether patients had received all the care they should have received. Each measure could be equally weighted, and a straight average could be taken—one for HQA measures and one for AQA measures.
-
Resource use: To capture resource use, the committee proposes longitudinal measures (risk-adjusted mortality) and resource use (standardized costs for Medicare Parts A and B) for all patients at a given hospital with AMI.
-
Eligibility for rewards: The hospital and all physicians who billed above a minimum threshold number of E&M claims for the hospital’s AMI patients during the subsequent year would be eligible.
-
How to distribute: There are two choices: (1) rewards could go to hospitals for distribution; or (2) rewards could be split between hospitals (X percent) and physicians (100–X percent), with physician rewards being distributed in proportion to the share of E&M claims.
REWARDING HEALTH PLAN CARE
Background
Health plans have been reporting data for more than 10 years through the National Committee for Quality Assurance’s (NCQA) Health Plan Employer Data and Information Set (HEDIS). HEDIS, first released by the HMO Group in 1991 and revised by NCQA in 1993, measures the performance of health plans on member satisfaction and delivery of chronic and preventive care for the purpose of accreditation and certification. HEDIS is used by over 90 percent of managed care organizations in the United States. A subset of health plans that work with Medicare, called Medicare Advantage plans, are paid out of Medicare Part C.
-
Clinical quality: HEDIS measures are updated annually and are widely endorsed and used. They reflect the following aspects of care: effectiveness of care (preventive screenings; immunizations; treatment of heart attacks, depression, asthma), access/availability of care (access to primary health care and dentistry, timeliness of claims), satisfaction with the experience of care (surveys for adult and child care), health plan stability, use of service, cost of care, informed health care choices, and health plan descriptive information. These data are reported publicly as the quality data within Medicare’s Personal Plan Finder, a website dedicated to comparing Medicare health plans.
TABLE E-4 Health Plan Phasing
|
Year 1 |
Year 2 |
Year 3 |
Clinical quality |
HEDIS measures—pay for public reporting |
HEDIS measures— pay for performance |
HEDIS measures— pay for performance |
Patient-centeredness |
CAHPS Health Plan Survey—pay for public reporting |
CAHPS Health Plan Survey—pay for performance |
CAHPS Health Plan Survey—pay for performance |
Efficiency |
Additional payout to the most efficient 1/3 of health plans meeting thresholds for both clinical quality and patient-centeredness measures |
Additional payout to the most efficient 1/3 of health plans meeting thresholds for both clinical quality and patient-centeredness measures |
Additional payout to the most efficient 1/3 of health plans meeting thresholds for both clinical quality and patient-centeredness measures |
-
Patient-centeredness: Health plans have been collecting patient-satisfaction data for years through use of the original CAHPS survey. A more specific survey, the CAHPS health plan survey, is part of the group of Ambulatory CAHPS surveys. It can be used to determine patients’ experiences of care provided by their health plan and will be ready for use in 2007.
-
Efficiency: A system should be developed to supplement the current lack of efficiency measures. See the example for rewarding efficiency in the section on dialysis facilities.
Timing of Pay for Performance
Health plans are very experienced at reporting quality and patient experience data; many have also already begun to provide incentives to their physicians based on performance. Pay for performance could, therefore, be implemented in health plans now. Pay-for-performance programs should include the most up-to-date measures of performance; new measures should be rewarded as they are introduced (see Table E-4).
REWARDING HOME HEALTH CARE
Background
Home health agencies have been able to attribute care to individual facilities, as illustrated by the fact that they have been publicly reporting
performance data through CMS since 2003. Federal support for home health care is provided through Medicare, paid out of Part A.
-
Clinical quality: Home health care has been measured largely through use of the Outcome and Assessment Instrument Set (OASIS). OASIS, implemented in 2000, measures both short- and long-term care in home health agencies. These measures are publicly reported on the Medicare Home Health Compare website.
-
Patient-centeredness: There currently are no patient-centeredness measures designed specifically for home health care. Efforts to assess patient experiences of care should use the original CAHPS measures, which have long been in use and have widespread support.
-
Efficiency: See the section on dialysis facilities for one possible method.
Timing of Pay for Performance
As measures of clinical quality are available now, pay for performance on those specific measures could begin in year 1 (2008). CAHPS measures from the original surveys for health plans would be used to characterize patient experiences until a more specific set was available. Because both clinical quality and patient-centeredness measures are available now, efficiency could also be assessed (see Table E-5).
TABLE E-5 Home Health Care Phasing
|
Year 1 (2008) |
Year 2 (2009) |
Year 3 (2010) |
Clinical quality |
OASIS measures— pay for performance |
OASIS measures— pay for performance |
OASIS measures— pay for performance |
Patient-centeredness |
CAHPS measures— pay for performance |
CAHPS measures— pay for performance |
CAHPS measures— pay for performance |
Efficiency |
Additional payout to the most efficient 1/3 of home health agencies meeting thresholds for both clinical quality and patient-centeredness measures |
Additional payout to the most efficient 1/3 of home health agencies meeting thresholds for both clinical quality and patient-centeredness measures |
Additional payout to the most efficient 1/3 of home health agencies meeting thresholds for both clinical quality and patient-centeredness measures |
Example of Pay for Performance for Home Health Care
Measurement of the performance of the physicians treating home health care patients would be based on clinical quality, patient-centeredness, and efficiency. The clinical quality measures would include the following:
-
Quality: A composite score for home health care in each care setting would be formulated to determine whether patients had received all the care they should have received. Each measure would be equally weighted, and a straight average would be taken.
-
Resource use: To capture resource use, the committee proposes standardized costs for Medicare Part A for all patients at a given home health agency.
-
Eligibility for rewards: Home health agencies that billed above a minimum threshold number of claims for patients during the subsequent year would be eligible.
-
How to distribute: Rewards would be distributed to the home health agencies based on performance.
OASIS Measures |
|
|
Patients who get better at walking or moving around in a wheelchair safely |
|
Patients who get better at getting in and out of bed |
|
Patients who get better at getting to and from the toilet |
|
Patients who have less pain when moving interfering with activity around |
|
Patients who get better at bathing |
|
Patients who get better at taking their medications correctly (by mouth) |
|
Patients who get better at getting dressed |
|
Patients who stay the same (don’t get worse) at bathing |
|
Percentage of patients who had to be admit |
|
Percentage of patients who need urgent, unplanned medical care |
|
Patients who are confused less often |
REWARDING SKILLED NURSING FACILITY CARE
Background
The majority of nursing home care is paid for by Medicaid and private payers. Medicare pays for a specific type of nursing home care, called skilled nursing care, through Medicare Part A. This care constitutes 25 percent of all nursing home care.
-
Clinical quality: The Minimum Data Set evaluates care in nursing homes. There are only three measures in this set that pertain to skilled nursing facilities. Evaluation of nursing homes is available through the Medicare Nursing Home Compare website.
-
Patient-centeredness: A Nursing Home CAHPS survey is still being developed to assess both patient and family experiences of care. Field testing was completed in January 2006; the final approval date for the instrument has yet to be determined.
-
Efficiency: See the section on hospital care for one possible method.
Timing of Pay for Performance
As described in Chapter 5, measures are not yet available that can adequately characterize care provided by skilled nursing facilities. Until such measures are developed, pay for performance should not be implemented in this setting. Once the necessary measures were available, implementation would proceed in a fashion similar to that for the other settings where pay for public reporting would precede pay for performance.