Medicare Fee-for-Service Issues and Innovations

This issue focuses on selected developments in the Medicare fee-for-service program. Two of the articles provide new estimates of the effect of Medicare supplemental insurance on total Medicare utilization and costs: One addresses utilization differences under alternative forms of supplemental insurance, and the other reports on utilization experience under the Medicare SELECT Demonstration. Two other articles discuss specific payment innovations: The first presents findings from the Medicare Participating Heart Bypass Center Demonstration, and the second describes the appeal and the challenges of moving from administered fee setting to competitive bidding for clinical laboratory services. Finally, there are two articles about tools with potential for improving the management of care under fee-for-service (FFS). The first of these describes how distinct but closely related patient classification systems for medical rehabilitation might be used for quality and outcomes monitoring as well as payment. The second analyzes the relationship between types of case manager activities and service utilization in the Medicare Alzheimer's Disease Demonstration.

The Balanced Budget Act of 1997 attempts to increase beneficiary cost-sharing in the FFS program through the creation of two new high-deductible, standard medigap policies. Research has consistently found that medigap supplemental insurance is associated with higher Medicare utilization and expenditures—a result usually associated with the fact that in its most common form, supplemental insurance eliminates beneficiary cost-sharing. Christensen and Shinogle provide some new estimates of this effect using 1994 data from the National Health Interview Survey. The article expands upon past work by comparing inpatient and outpatient utilization among beneficiaries with three types of private insurance supplements—medigap plans, employment-based indemnity plans, and health maintenance organizations (HMOs)—and those with no supplemental insurance. These plan types often differ not only in the extent of their cost-sharing features, but also the range of benefits covered, and the extent to which managed care techniques are used to influence utilization. As expected, Christensen and Shinogle find higher utilization of both inpatient and outpatient services among beneficiaries with medigap and employment-based indemnity plans. Controlling for a variety of other factors, including some health status measures, the effect averages 28 percent for medigap and 17 percent for employment-based plans. For HMO members, utilization averages about 4 percent less than that for beneficiaries with no supplemental insurance.

The Medicare SELECT program represents an attempt by Congress to inject some managed care incentives into the Medicare FFS program. Congress expected that Medicare expenditures could be reduced by inducing Medicare beneficiaries to use selective provider networks established by medigap insurers. SELECT policies only pay full benefits when network providers are used. Lee, Garfinkel, Khandker, and Norton present findings from their evaluation of the SELECT program in 11 States during the initital demonstration period (1992-94). Under current law, the Medicare SELECT program has been extended to June 1998, at which time it will become permanent unless it is determined unsuccessful in meeting its objectives. As implemented, SELECT policies have incorporated weak managed care incentives. SELECT insurers have typically created hospital-only networks that discount or waive the Part A deductible and cover the Part B cost-sharing for any physician seen by the beneficiary. Where physician networks have been used, they typically have not employed gatekeepers.

Given the fact that SELECT policies are medigap policies, Lee et al. estimate separate medigap and SELECT impacts on Medicare utilization and cost. Their estimates of the medigap effect on cost per beneficiary range from +7.4 to +29.6 percent and are generally consistent with Christensen and Shinogle's estimates, as well as those of other studies. The effect of SELECT varies by State: increased cost in five States, decreased cost in three States, and no effect in three States. Cost decreases were sometimes attributable to increased use of less expensive hospitals, but more frequently were due to lower costs in physician offices or hospital outpatient departments. As to the unexpected cost-increasing impacts, Lee et al. suggest the following most likely potential explanations: Some providers may have increased their services billed to offset discounts negotiated with the SELECT insurer. SELECT insurers in some States may have unintentionally associated with relatively expensive physicians by virtue of the hospitals with whom they contracted.

In 1991, HCFA initiated a demonstration (the Medicare Participating Heart Bypass Center Demonstration) that paid a negotiated bundled price for the inpatient hospital and physician services associated with Medicare heart bypass surgery. By negotiating payment discounts and replacing FFS physician incentives with financial incentives that would encourage more efficient care delivery, the hope was that both individual Medicare beneficiaries and the program could benefit. The original phase of the demonstration covered two diagnosis-related groups (DRGs) (DRG 106 and DRG 107) and four hospitals, in Boston, Atlanta, Ann Arbor, and Columbus, Ohio. Subsequently, the demonstration was expanded to include three more hospitals, in Indianapolis, Houston, and Portland, Oregon. Based on the success of the heart bypass demonstration, new demonstrations are under development. The Medicare Participating Centers of Excellence demonstration would add more types of cardiovascular surgery cases and apply the concept to a specified number of orthopedic surgery cases. Hospitals could qualify for the demonstration as either cardiovascular or orthopedic centers, or both. Another demonstration, Medicare Provider Partnerships, would expand the bundled hospital-physician payment to all inpatient DRGs. The physician-hospital participants in this demonstration are located in New Jersey, New York, and Pennsylvania.

Cromwell, Dayhoff, and Thoumaian present results from their study of the early experience (1991-93) of the heart bypass demonstration. They found that cost savings were possible as a result of the altered physician incentives. Hospitals were already largely at risk under the DRG-based prospective payment system (PPS). Four physician specialists (thoracic surgeons, anesthesiologists, cardiologists, and radiologists) were assumed to be involved in all cases and were paid a fixed amount per case. All other consulting physicians were paid Medicare allowable fees from a consultant pool. The four key specialists were partially at risk for the consultant pool. Hence, the four key specialists stood to benefit from more efficient use of their own services and those of physician consultants. In two hospitals, physicians also received a share of any hospital savings from more efficient care.

Cromwell, Dayhoff, and Thoumaian found that, during the 2-½-year period of their study, the Medicare program and its beneficiaries together saved an average of $4,700 per bypass surgery case in the four participating institutions. Loss of savings due to shifts to outpatient care did not occur. Quality of care did not suffer. Although institutions differed in the degree of success in reducing costs and increasing market share, encouraging signs of significant cost savings were observed. For example, cost reductions were achieved in intensive care units, routine nursing, laboratory services, and pharmacy as surgeons assumed greater responsibility for managing patients throughout the course of their care.

The Balanced Budget Act of 1997 mandates as many as five demonstration projects of competitive bidding for Part B items and services, except physician services. Hoerger and Meadow discuss the issues that Medicare must confront in designing these demonstrations and report the results of their empirical investigation of these issues applied to clinical laboratory services. Recognizing the advantages of flexibility and dynamism that competitive bidding offers compared with administered fee setting, Hoerger and Meadow consider how the structure of bidding affects price and quality objectives. Medicare benefits from lower prices as long as quality is not jeopardized. They describe reasons why Medicare may wish to permit multiple winning bidders to supply services; conclude that a very high proportion of clinical laboratory expenditures could be captured within an acceptable scope for bidding; and identify factors that highlight the importance of careful selection of market areas for bidding.

Also required in the Balanced Budget Act of 1997 is the establishment of a PPS for rehabilitation hospital or unit services based on patient case-mix groups by October 1, 2000. Function-related groups based on the functional independence measure (FIM-FRGs) are one case-mix system that could potentially be used in a rehabilitation payment system. Stineman and Granger show that the basic construct of the FIM-FRGs can be used to develop patient groupings that explain variation in expected functional status at discharge and in the extent of improvement in functional status during the stay. They illustrate how, in the case of lower extremity fracture patients, the three distinct, but closely related, systems can be used for monitoring resource use and patient outcomes.

The Balanced Budget Act of 1997 mandates a demonstration to evaluate methods, such as case management and other models of coordinated care that improve the quality of care and reduce Medicare expenditures for beneficiaries with chronic illnesses in FFS Medicare. There is experience with case management in the delivery of long-term care services, and it has achieved some acceptance in the managed care sector. However, the findings about the effectiveness of case management have been inconsistent. Generally, research has failed to find any cost savings resulting from case management interventions. Nevertheless, it is possible that the lack of positive findings may relate to the methodologies used in these studies. Most studies have not attempted to differentiate the myriad of activities performed by case managers and have instead treated every case management contact as being equivalent. Using data obtained from the Medicare Alzheimer's Disease Demonstration, Newcomer, Arnsberger and Zhang unbundle the activities performed by case managers into the component parts and describe the critical elements of a case management intervention. Better understanding of when and how case management is differentiated could result in the design of case management programs that are more sensitive to the measurement of outcomes.