Selected Medicare Prospective Payment Systems
ACUTE CARE HOSPITALS, INPATIENT
The inpatient hospital prospective payment system (PPS), which was established in 1983, uses a preset payment schedule based on a patient’s principal diagnosis at discharge, comorbidities, and complications. The service unit is a patient stay. The fixed payment amounts are intended to cover the average costs of all services, supplies, and elements of care an efficient hospital would need to treat the average patient in a specified diagnosis-related group (DRG). There are 524 distinct DRGs, each of which encompasses “patients with similar clinical problems that are expected to require similar amounts of hospital resources” (MedPAC, 2005a). Some cases will cost the hospital more and others less than actual Medicare payments for a particular diagnosis. While these “bundled” payments are indexed to account for geographic differences in labor and other input costs, all hospitals within an area receive the same base payments for DRGs regardless of quality (Worzala et al., 2003). Hospital-specific adjustments that take the form of a percentage increase in all payments are made for institutions serving a disproportionate share of low-income and uninsured patients and teaching hospitals.
The acute inpatient hospital PPS provides an incentive to manage the costs of inputs needed for care. Hospitals can manage their costs by eliminating unnecessary services, reducing the intensity of services per case, bargaining hard over input prices, shortening lengths of stay, increasing the volume of less complex cases within any particular DRG, and reducing the volume of cases in DRGs for which Medicare’s preset payment does not
cover the costs of the average case. The impact of the hospital inpatient PPS on the quality of hospital care is unclear. Early concerns that the DRG payment system would lead to stinting on care and an inappropriate shortening of hospital stays appear largely to have been unfounded. MedPAC’s most recent assessment of trends found lower in-hospital and 30-day mortality rates, improvement in measures of appropriateness of care and clinical effectiveness, and some increases and some decreases in measures of adverse events (MedPAC, 2006).
ACUTE CARE HOSPITALS, OUTPATIENT
The implementation of inpatient PPS promoted a shift of care to outpatient departments because hospitals continued to be reimbursed for such care on a retrospective, cost basis. To reduce this incentive, a hospital outpatient PPS was introduced in 2000. Like the inpatient PPS, the outpatient system groups services that are similar clinically and costwise into one of about 850 ambulatory payment classification (APC) categories, each with its own payment rate. Additional APCs are designated for new technologies—those for which the Centers for Medicare and Medicaid Services has insufficient data—which are grouped together by cost, not clinical similarity. There are also pass-through payments that cover the costs of particular new drugs; costs of biologicals and devices used in the delivery of services; outlier payments for cases that are unusually expensive relative to the preset payment rate; and adjustments for rural, low-volume facilities.
Compared with the inpatient PPS, the outpatient system provides somewhat weaker incentives for efficiencies because the APC service bundles are not as broad as those of the DRGs, and the use of certain new technologies is encouraged. The volume of outpatient services continues to grow rapidly as procedures once requiring a hospital stay can now be performed safely on an outpatient basis. There is little systematic knowledge about trends in the quality of outpatient care or the impact that the outpatient PPS may have had on the quality of care.
SKILLED NURSING FACILITIES
Skilled nursing facilities (SNFs) shifted from cost-based reimbursement to a PPS in 1998. Under the PPS, SNFs are paid set per diem rates for each patient. Based on periodic assessments, patients with similar needs and characteristics are placed in one of 53 resource utilization groups, each with its own payment rate. These rates, which are the sum of a nursing component, a therapy component, and a routine services component, are intended to cover all routine care and ancillary services. Additional payments are made for certain rare but high-cost ancillary services, such as an outpatient hospi-
tal scan. The per diem base payments are adjusted for geographic differences in labor costs. Nontherapy ancillary services, such as tracheostomy and ventilator care, and certain prescription drugs, while included in the nursing component of the base rate, were not included in the case mix indexes, so the base payments do not appropriately reflect the resource needs of certain extensive care patients (MedPAC, 2006). Recently, payments for patients needing extensive (nontherapy) services were reduced absolutely and in comparison with those for patients needing therapies, and that change may explain the relatively longer delays such patients encounter in accessing SNF care (MedPAC, 2005b). Information is insufficient for making judgments about the quality of care provided in individual facilities or industrywide; only three measures collected through the patient assessment and reporting system known as the nursing home Minimum Data Set relate to the quality of SNF care.
Medicare’s market share nationally for free-standing home health care was 32 percent in 2003; the percent for hospital-based home health care was not available (MedPAC, 2006). Until the mid-1990s, home health agencies received cost-based reimbursement from Medicare. The number of Medicare patients, the number of services provided per case, and the number of agencies grew quickly in the early 1990s. Although the number of visits for each case had to be approved by the Medicare fiscal intermediary, the cost-based system encouraged agencies to provide as many visits as would be allowed. In 1998 Medicare implemented a 2-year interim payment system to provide a transition to prospective payment. The interim system included a financial incentive to cut the number of visits per case when possible. The number of participating agencies dropped by approximately 30 percent from 1997 to 2000, indicating the impact of Medicare’s payment policy, along with strong new regulatory efforts to control fraud and abuse in the home health industry.
Under the home health PPS, implemented in 2000, the unit of payment is an episode, which includes all services needed during a 60-day period. If the patient needs care for a longer period, the home health agency receives another episode payment. The episode includes skilled nursing care; physical, occupational, and speech therapy; and medical social work and aide services. Patients are assigned to one of 80 Home Health Resource Groups based on their functional status, clinical condition, and likely use of various services. Outlier payments are made for particularly costly episodes, and reduced payments are made for episodes that require fewer than five visits, for transfer cases, and for patients that have a significant change in status during an episode.
Lacking a sophisticated case mix adjustment that accurately reflects the likely resource needs of the patient, providers have an incentive to decrease the number of visits per episode and to increase the number of episodes per patient. The impact of the payment system can be seen during the transition period, between 1997 and 2002: the average visits per episode dropped from 36 to 19 and the average length of stay from 106 to 56 days (MedPAC, 2005d).
Performance measures of home health services show quality has improved slightly or remained stable over the last couple of years, but that trend cannot be related to specific changes in payment. What confounds judgments about the impact of payment incentives on the use of home health services and their quality is the lack of specific methods and guidelines for identifying patients in need of such services, what specific services they need, and whether home health is the most appropriate source for that care. This situation is not unique to home health care; it is typical of all post–acute care settings in which patients can receive similar services from different types of providers that receive different amounts and types of payments.
OUTPATIENT DIALYSIS SERVICES
The payment for outpatient dialysis services has two parts: the slightly larger part is a prospective composite payment that covers the bundle of services associated with a dialysis treatment, and the smaller part covers certain separately billable drugs and supplements, such as erythropoietin, vitamin D, and iron. Dialysis facilities also receive payments for laboratory tests not included in the composite rate. The composite payment is adjusted for the patient’s age, body mass, and body surface area and for geographic differences in wages and other costs. While the base payment is the same for hemodialysis and peritoneal dialysis and for in-center or home dialysis, hospital-based facilities receive $4 more per treatment.
The payment system had created incentives for providers to increase the efficiency of the composite-covered services and to increase use of the separately covered drugs and supplements because the latter payment had been based on the average wholesale price of the products, which exceeded the centers’ acquisition costs. The Medicare Prescription Drug, Improvement, and Modernization Act of 2003 (Public Law 108-173) required that beginning in 2006, the drug payment be the average sales price plus 6 percent, and that the savings be added to the composite rate. This less generous drug payment was intended to slow the growth in drug spending without jeopardizing the quality of care for dialysis patients. While the volume of services has been growing, the proportion of people receiving services at home has been shrinking (MedPAC, 2006). Because the payments are location neutral, the latter trend may indicate a perception on the part of pro-
viders that home dialysis is more costly to provide or that patients find it less desirable. The dialysis program has its own quality improvement organizations that have documented improvements in care between 1999 and 2003, but it is too soon to tell whether recent changes in the payment system have had any effect on this trend.
Medicare pays for services provided by physicians and other medical professionals who can charge directly—such as dentists, optometrists, podiatrists, chiropractors, psychologists, clinical social workers, nurse midwives, certified registered nurse anesthetists, physician assistants, nurse practitioners, clinical nurse specialists, physical therapists, occupational therapists, and registered dietitians—under the physician fee schedule. This schedule assigns a payment amount to each of more than 7,000 procedures, visit types, and other diagnostic and therapeutic services. The scope of some of the services—such as a flu shot—is quite narrow, while that of others— such as the office and hospital visits and surgery associated with a hip replacement—is quite broad. Over 74 percent of public and private payers, including state Medicaid programs, have adopted components of the Medicare system for reimbursing physicians (AAP, 2005).
The specific amount of the payment is based on the relative costliness of the professional work, practice expenses, and liability insurance needed to provide the service. The latter two components are adjusted for geographic differences in costs. Because the initial methodology for determining the relative value of the professional work involved in providing specific services was based on calculations of the resources used in care, the time involved, the complexity of procedures, and physician training,1 the fee schedule tends to reward specialty care at higher rates relative to primary care.
The annual updates to the physician fee schedule are governed by the sustainable growth rate, which, broadly speaking, limits the growth in per beneficiary physician fee schedule expenditures to the growth in per capita gross domestic product. Recent rapid growth in the volume and intensity of services and congressional interventions to stave off payment reductions have created a situation in which the updates for the physician fee schedule are projected to be negative for at least the next 7 years.
Since the system is based on paying per service provided, it tends to penalize rather than reward physicians who use fewer resources or services to achieve a given level of quality or outcome (Wilensky, 2005). But the physician or other provider will receive additional payment if the service resulted in a complication and had to be repeated or followed by corrective procedures. The payment system also tends to offer higher rates for new technologies while providing few incentives to use older, lower-cost technologies that may be equally effective.
Medicare Advantage (MA) plans are paid a capitated amount for each participant each month, the size of which depends on the plan’s bid, a benchmark, and the risk characteristics of the beneficiary. For local plans, those that offer their services on a county-by-county basis, the benchmark is the county MA payment rate that existed before 2006, updated each year by the increase in national per beneficiary fee-for-service (FFS) spending. The old payment rates generally exceeded the average per capita expenditures for FFS, in some counties by a considerable amount. Some MA plans benefited from minimum payment floors or blended local–national rates that were intended to attract MA plans into areas with low FFS spending; others were boosted by a guaranteed minimum update; and still others were advantaged by inconsistencies in the methodology used to set the rate. For regional plans, the preferred provider organizations that offer service throughout one of the 26 state or multistate regions, the benchmark is a weighted average of the local benchmarks.
Each year, Medicare compares plan bids with the relevant benchmark for providing the comprehensive bundle of services mandated by Medicare Parts A and B (except hospice services) for an average beneficiary. If the bid is below the benchmark, the plan is paid its bid plus three-fourths of the difference between the benchmark and the bid. The plan must use the “excess” payment for additional (nonrequired) benefits, reduced cost sharing, or Part B or Part D premium reductions. Plans bidding above the benchmark are paid the benchmark and must charge their participants the difference between their bid and the benchmark. The basic payment is adjusted for an enrollee’s risk profile using a methodology that incorporates information on the individual’s demographic characteristics and previous use of hospital inpatient and ambulatory services. The quality of care in MA plans is measured more fully and there is a longer history of performance data collection than is the case for FFS Medicare. The Health Plan Employer Data and Information Set is used for data collection, and those measures show general improvement over time; some measures remain low, however, and there is substantial variation among overall plan scores (MedPAC,
2005c). It is unclear how the payment methods have affected quality and, given the major changes in those methods over the last decade, it would be difficult to attribute quality change to any specific incentive in the payment system.
AAP (American Academy of Pediatrics). 2005. 2006 RBRVS: What Is It and How Does It Affect Pediatrics? [Online]. Available: http://www.aap.org/visit/rbrvsbrochure.pdf [accessed December 7, 2005].
MedPAC (Medicare Payment Advisory Commission). 2005a. Hospital Acute Inpatient Services Payment System. Washington, DC: MedPAC.
MedPAC. 2005b. Report to the Congress: Medicare Payment Policy. Washington, DC: MedPAC.
MedPAC. 2005c. A Data Book: Healthcare Spending and the Medicare Program. Washington, DC: MedPAC.
MedPAC. 2005d. MedPAC Data Runs. Washington, DC: MedPAC.
MedPAC. 2006. Report to the Congress: Medicare Payment Policy. Washington, DC: MedPAC.
Wilensky GR. 2005. The twin policy challenges of Medicare physician payment and Medicaid. Health Affairs w5-333–w5-334.
Worzala C, Pettengill J, Ashby J. 2003. Challenges and opportunities for Medicare’s original prospective payment system. Health Affairs 22(6):175–182.