Notes All years referred to in this paper are calendar years. Numbers in the text and tables may not add up to totals because of rounding. All dollar amounts are expressed in 2005 dollars (having been converted using the GDP price deflator). Preface T otal Medicare spending exceeded $300 billion in 2004 and is expected to grow significantly in the coming decades. In response to those financial pressures, policymakers have considered a number of strategies for slowing the growth of Medicare spending, including increasing the share of spending paid by beneficiaries and enhancing competition in the provision of services. This Congressional Budget Office (CBO) paper, prepared at the request of the Senate Majority Leader, explores an additional approach: focusing on the small group of beneficiaries that accounts for a large share of the program's spending. Possible intervention strategies would aim to reduce that spending. The paper analyzes the concentration of Medicare spending in a given year and over time using data from Medicare claims and considers illustrative methods for identifying beneficiaries who are likely to incur high future costs. In keeping with CBO's mandate to provide objective, nonpartisan analysis, this paper makes no recommendations. Julie Lee and Todd Anderson of CBO's Health and Human Resources Division prepared the paper under the supervision of Bruce Vavrichek, James Baumgardner, Steve Lieberman, and Mark Miller. (Todd Anderson, Steve Lieberman, and Mark Miller have since left CBO). Susan Labovich provided computer programming support and coordinated CBO's access to the data used in this analysis. Tom Bradley, Philip Ellis, and Allison Percy, also of CBO, provided thoughtful comments on drafts, as did Amber Barnato of the University of Pittsburgh and Jeffrey Kelman of the Centers for Medicare and Medicaid Services. (The assistance of external reviewers implies no responsibility for the final product, which rests solely with CBO.) Dr. Barnato and Jennifer Bowman, formerly of CBO, helped to draft portions of an earlier version of the paper. Christine Bogusz edited the paper, and Loretta Lettner proofread it. Maureen Costantino prepared the paper for publication and designed the cover. Lenny Skutnik printed copies of the paper, and Annette Kalicki produced the electronic version for CBO's Web site (www.cbo.gov). Dongen Hz.Ere. Douglas Holtz-Eakin Director May 2005 CONTENTS Summary and Introduction 1 The Concentration of Medicare Expenditures 2 The Persistence of Medicare Expenditures 5 The Concentration of Spending Over a Five-Year Prospectively Identifying Future High-Cost Beneficiaries 8 Illustrative Strategies for Identifying High-Cost Beneficiaries 9 Approaches to Managing Care for High-Cost High-Cost Medicare Beneficiaries Summary and Introduction Medicare's rising demand for budgetary resources has been well documented, with spending for the program projected to grow significantly in the coming decades.' In response, policymakers are exploring ways to reduce that potential growth in spending, perhaps by increasing beneficiaries' cost sharing or promoting competition among service providers. Options that would constrain the program's costs without reducing the quality or availability of care, however, are limited. One potential avenue for reducing Medicare spending stems from the observation that a small fraction of Medicare beneficiaries accounts for a large share of the program's spending in a given year. That concentration of expenditures is characteristic of insurance programs in general. However, it also suggests the possibility of a policy alternative: identify the relatively small group of potentially high-cost beneficiaries and find effective intervention strategies to reduce their spending. If that approach was successful, even a small percentage reduction in the spending of that group of beneficiaries could lead to large savings for the Medicare program. suming that future high-cost beneficiaries can be identified, can effective strategies be devised to avert the anticipated high costs? This paper explores the first two questions in order to gauge the potential effectiveness of focusing on high-cost Medicare beneficiaries as a way to reduce the program's costs. To begin with, it documents the extent to which Medicare expenditures are concentrated: among relatively few beneficiaries. Analysts observe a significant degree of concentration in the spending of Medicare beneficiaries, both in a given year and over time. For example, highcost beneficiaries (those in the top 25 percent in terms of their spending) accounted for 85 percent of annual expenditures in 2001 and for 68 percent of five-year cumulative expenditures from 1997 to 2001. In addition, those high-cost beneficiaries, compared with beneficiaries in the bottom 75 percent in terms of their spending, were slightly older, more likely to suffer from chronic conditions, such as coronary artery disease and diabetes, and more likely to die in a given year. Finally, to determine whether beneficiaries with high future costs can be identified before those costs are incurred, the paper explores three simple observational methods for prospectively identifying beneficiaries who will incur extended periods of expensive medical care. The methods look at beneficiaries who were high cost in a certain year (in this case, 1997), who were admitted to a hospital that year, or who had multiple chronic conditions that year. The costs for all three groups are compared with those for a random sample of Medicare beneficiaries. Spending in 1997 for all three groups was more than twice as high as spending for the reference group, and it was four times as high for the group with a hospital admission. The next year, however, the previously hospitalized group had the largest decline in its share of spending, whereas the share of spending by beneficiaries with multiple chronic conditions barely fell at all. Over the next |