Page 1       TOP OF DOC
42–117 CC


before the


of the





FEBRUARY 25, 1997

Serial 105–4

Printed for the use of the Committee on Ways and Means
 Page 2       PREV PAGE       TOP OF DOC

BILL ARCHER, Texas, Chairman

BILL THOMAS, California
E. CLAY SHAW, Jr., Florida
NANCY L. JOHNSON, Connecticut
WALLY HERGER, California
JIM McCRERY, Louisiana
DAVE CAMP, Michigan
JIM RAMSTAD, Minnesota
PHILIP S. ENGLISH, Pennsylvania
J.D. HAYWORTH, Arizona
 Page 3       PREV PAGE       TOP OF DOC
ROBERT T. MATSUI, California
WILLIAM J. COYNE, Pennsylvania
JIM McDERMOTT, Washington
RICHARD E. NEAL, Massachusetts
JOHN S. TANNER, Tennessee

A.L. Singleton, Chief of Staff

Janice Mays, Minority Chief Counsel

Subcommittee on Health

BILL THOMAS, California, Chairman
 Page 4       PREV PAGE       TOP OF DOC
NANCY L. JOHNSON, Connecticut
JIM McCRERY, Louisiana

Pursuant to clause 2(e)(4) of Rule XI of the Rules of the House, public hearing records of the Committee on Ways and Means are published in electronic form. The printed hearing record remains the official version. Because electronic submissions are used to prepare both printed and electronic versions of the hearing record, the process of converting between various electronic formats may introduce unintentional errors or omissions. Such occurrences are inherent in the current publication process and should diminish as the process is further refined. The electronic version of the hearing record does not include materials which were not submitted in an electronic format. These materials are kept on file in the official Committee records.


    Advisory of February 12, 1997, announcing the hearing
 Page 5       PREV PAGE       TOP OF DOC

    Congressional Budget Office, Paul N. Van de Water, Ph.D., Assistant Director for Budget Analysis

    Prospective Payment Assessment Commission, Joseph P. Newhouse, Ph.D., Chairman; accompanied by Donald A. Young, M.D

    American Academy of Actuaries, Alice Rosenblatt

    Oxford Health Plans, Inc., Stephen F. Wiggins


    American Hospital Association, statement

House of Representatives,
Committee on Ways and Means,
Subcommittee on Health,
Washington, DC.

    The Subcommittee met, pursuant to notice, at 9:10 a.m., in room 1100, Longworth House Office Building, Hon. Bill Thomas (Chairman of the Subcommittee) presiding.
    [The advisory announcing the hearing follows:]

 Page 6       PREV PAGE       TOP OF DOC



CONTACT: (202) 225–3943


February 12, 1997

No. HL–2

Thomas Announces Hearing on

Medicare HMO Payment Policies

     Congressman Bill Thomas (R–CA), Chairman, Subcommittee on Health of the Committee on Ways and Means, today announced that the Subcommittee will hold a hearing to examine Medicare Health Maintenance Organization (HMO) enrollment growth and payment policies. The hearing will take place on Tuesday, February 25, 1997, in the main committee hearing room, 1100 Longworth House Office Building, beginning at 9:00 a.m.
 Page 7       PREV PAGE       TOP OF DOC
    In view of the limited time available to hear witnesses, oral testimony at this hearing will be heard from invited witnesses only. However, any individual or organization not scheduled for an oral appearance may submit a written statement for consideration by the Committee and for inclusion in the printed record of the hearing.
    In recent years, there has been a rapid increase in the number of Medicare beneficiaries choosing to enroll in Medicare risk HMOs. While only 4 percent of eligible beneficiaries were enrolled in 1991, by 1996 this number had climbed to 11 percent. The Congressional Budget Office estimates that over the next decade, this trend will continue. By the year 2007, more than one third of the Medicare beneficiaries are expected to be enrolled in HMOs.
    Under its risk contracting program, Medicare pays risk plans based on a county-level per capita monthly rate equal to 95 percent of the adjusted average per capita costs in the Medicare fee-for-service sector. Calculating the rates at the county level based on historical fee-for-service program spending results in wide variation in rates across counties. For example, the 1997 monthly per capita rates range from $221 in one county to $767 in another—an annual difference of $6,552.
    Medicare's county-level payment rates are based on the costs of treating the average beneficiary. The Health Care Financing Administration currently uses a limited set of demographic characteristics to adjust the rates, including gender, age, Medicaid eligibility, institutional status, and employment status (working or non-working aged). Unfortunately, these adjustments do not accurately account for much of the variation in costs. Better risk adjustment methods could reduce Medicare costs and increase the choices available to Medicare beneficiaries.
 Page 8       PREV PAGE       TOP OF DOC
    The President's fiscal year (FY) 1998 budget proposal fundamentally restructures the manner in which HMOs are paid by Medicare and reduces projected payments to HMOs by $34 billion from FY 1998 to FY 2002. The most significant policy change would be to sever the direct link between HMO payment rates and individual county-level fee-for-service spending. The proposal also would blend national and regional rates and institute a payment floor of $350 per month for payments in low-cost counties. Under the President's plan, the calculation of the HMO payment rates would exclude Medicare's special payments for direct and indirect medical education and to disproportionate share hospitals. Beginning in the year 2000, the President's plan would further reduce HMO payment rates by approximately 5 percent across-the-board to account for an ''overpayment'' which is alleged to occur due to risk selection.
    In announcing the hearing, Chairman Thomas stated: ''We are encouraged by the growing number of beneficiaries choosing to enroll in private health plans. Clearly, giving Medicare beneficiaries the choice of enrolling in HMOs and other health plans will be a fundamental element of any plan to save Medicare. I am happy to see that the administration has come to the table with a plan that recognizes that fundamental changes must be made in Medicare's payment system. I look forward to reviewing the details of the administration's plan.''
    This hearing will focus on the implications of the President's budget proposals on Medicare HMO enrollment and spending. These proposals will be assessed in light of the Medicare recommendations developed for the Congress by the Prospective Payment Assessment Commission and the Physician Payment Review Commission, as well as the policies contained in the Medicare Preservation Act of 1995 and the Balanced Budget Act of 1995.
 Page 9       PREV PAGE       TOP OF DOC
    Any person or organization wishing to submit a written statement for the printed record of the hearing should submit at least six (6) copies of their statement and a 3.5-inch diskette in WordPerfect or ASCII format, with their address and date of hearing noted, by the close of business, Tuesday, March 11, 1997, to A.L. Singleton, Chief of Staff, Committee on Ways and Means, U.S. House of Representatives, 1102 Longworth House Office Building, Washington, D.C. 20515. If those filing written statements wish to have their statements distributed to the press and interested public at the hearing, they may deliver 200 additional copies for this purpose to the Subcommittee on Health office, room 1136 Longworth House Office Building, at least one hour before the hearing begins.
    Each statement presented for printing to the Committee by a witness, any written statement or exhibit submitted for the printed record or any written comments in response to a request for written comments must conform to the guidelines listed below. Any statement or exhibit not in compliance with these guidelines will not be printed, but will be maintained in the Committee files for review and use by the Committee.
    1. All statements and any accompanying exhibits for printing must be typed in single space on legal-size paper and may not exceed a total of 10 pages including attachments. At the same time written statements are submitted to the Committee, witnesses are now requested to submit their statements on a 3.5-inch diskette in WordPerfect or ASCII format.
 Page 10       PREV PAGE       TOP OF DOC
    2. Copies of whole documents submitted as exhibit material will not be accepted for printing. Instead, exhibit material should be referenced and quoted or paraphrased. All exhibit material not meeting these specifications will be maintained in the Committee files for review and use by the Committee.
    3. A witness appearing at a public hearing, or submitting a statement for the record of a public hearing, or submitting written comments in response to a published request for comments by the Committee, must include on his statement or submission a list of all clients, persons, or organizations on whose behalf the witness appears.
    4. A supplemental sheet must accompany each statement listing the name, full address, a telephone number where the witness or the designated representative may be reached and a topical outline or summary of the comments and recommendations in the full statement. This supplemental sheet will not be included in the printed record.
    The above restrictions and limitations apply only to material being submitted for printing. Statements and exhibits or supplementary material submitted solely for distribution to the Members, the press and the public during the course of a public hearing may be submitted in other forms.

    Note: All Committee advisories and news releases are available on the World Wide Web at 'HTTP://WWW.HOUSE.GOV/WAYS_MEANS/'.
 Page 11       PREV PAGE       TOP OF DOC

    The Committee seeks to make its facilities accessible to persons with disabilities. If you are in need of special accommodations, please call 202–225–1721 or 202–225–1904 TTD/TTY in advance of the event (four business days notice is requested). Questions with regard to special accommodation needs in general (including availability of Committee materials in alternative formats) may be directed to the Committee as noted above.


    Chairman THOMAS. The Subcommittee will come to order. Today we are conducting the second hearing on the President's fiscal year 1998 budget proposal with a focus on the Medicare health maintenance organization payment policy. The Congressional Budget Office estimates that 12 percent of beneficiaries are enrolled in private HMO plans today and that that number will reach 23 percent by 2002 all other things being equal, and one of the things we are going to question is whether or not those numbers will be achieved since there are a number of folk who do not want to make sure that all other things are equal.
    CBO further projects that more than one-third of the beneficiaries will voluntarily choose private health plans over the fee-for-service traditional Medicare by 2007. The reasons for the rapid rise of enrollment are assumedly apparent: beneficiaries are finding that well-run private health plans can generally provide more benefits with less out-of-pocket costs than fee-for-service Medicare. By enrolling in risk contract HMOs, beneficiaries can reduce their Medigap premiums or forgo them all together and forgo paperwork hassles.
    We need to make sure that any changes we make in the Medicare HMO payment policy do not jeopardize the availability of private health plan options or inhibit the further expansion of beneficiary choice. This hearing will examine the concepts of the President's HMO payment proposal. I say examine the concepts because we have not yet gotten all of the legislative language to be comfortable with what the President is proposing in its entirety.
 Page 12       PREV PAGE       TOP OF DOC
    According to the Office of Management and Budget scoring, projected HMO payments would be reduced by $34 billion between fiscal year 1998 and 2002. The biggest decrease in health plan payments, $18 billion, would come from the indirect effect of lower fee-for-service spending. In addition, there would be a $10 billion reduction from removing teaching and disproportionate share payments from the HMO payment rates. Finally, the President's plan includes an across-the-board reduction to the rates in years 2000 and beyond to, they argue, account for risk selection. These are significant changes, and they need to be evaluated carefully.
    We have with us today three distinguished panel members to help us evaluate the President's proposal. I will introduce our witnesses in a moment. First, I will yield to my colleague and friend from California, the Ranking Member, Mr. Stark.
    Mr. STARK. Thank you, Mr. Chairman, for holding this hearing. Clearly, we need to make savings in this sector of Medicare, and like you I am not convinced the administration's proposals are the best way to achieve the savings. I hope today's witnesses can help us find a way to achieve savings in these HMOs, who have been in a very real sense cheating the program. They have been the ones signing up healthier patients and avoiding the care of sicker patients. By concentrating cuts on those types of HMOs, we can avoid hurting the good HMOs that have been efficiently doing the right thing.
    It was said by a former colleague of ours, something like pork roast and fatback come from the same animal, but there is a world of difference in the taste. Some HMOs are very good and some are just plain bad. For example, in today's panel, Puget Sound spends 96 cents of every premium dollar on health care while Oxford only spent 77 1/2 cents per dollar on medical care in 1995. Now, with nearly 20 cents a dollar difference, it is hard to imagine that the two HMOs provide the same level of quality and good outcomes. Similarly, a recent search of news articles indicates a litany of complaints about PacifiCare but great praise for Puget Sound. Now, clearly, there are major differences between HMOs and I hope we can explore how Medicare can reward the good plans and help consumers pick the plans that deliver health care rather than those that use up Medicare money in overhead and profits. I would like to enter into the hearing record recent data by the Sherlock Co. of Pennsylvania on HMO profitability and overhead costs along with a compilation of newsclips regarding the three HMO witnesses for today.
 Page 13       PREV PAGE       TOP OF DOC
    Chairman THOMAS. Thank the gentleman. Without objection, the papers will be entered in the record.
    Mr. STARK. Thank you, Mr. Chairman.
    [The following was subsequently received:]
    [The official Committee record contains additional material here.]


    Chairman THOMAS. Let us begin then. Our first panelist, Paul Van de Water, who is the Assistant Director for Budget Analysis, Congressional Budget Office, will give us as much of an update as he is able given the materials that we have; Gail Wilensky, Chair of the Physician Payment Review Commission, who is accompanied by Dr. LeRoy; and Dr. Newhouse is with us, who is the Chairman of the Prospective Payment Assessment Commission, ProPAC, accompanied as always by Dr. Donald Young.
    Let us begin with you, Paul, and then move across the panel.


    Mr. VAN DE WATER. Thank you, Mr. Chairman. Mr. Stark, Members of the Subcommittee, thank you for inviting me here this morning to discuss the role of health maintenance organizations in Medicare. Although the Congressional Budget Office has recently reduced its projection of spending for Medicare, the program's costs are still growing at rates that contribute substantially to the Federal deficit in the short term and are unsustainable in the long run. At the same time, CBO projects that enrollment in Medicare HMOs will continue to increase substantially.
 Page 14       PREV PAGE       TOP OF DOC
    These two trends then raise the question, How can Medicare's HMO Program be changed so as to achieve budgetary savings in the short term and to help secure Medicare's survival in the long run? The shift in enrollment toward risk-based HMOs will not slow Medicare spending unless the program can retain some of the savings possible from managed care plans. In the short run, that result could be achieved in a couple of ways. The simplest alternative would be to reduce Medicare's payment rate from 95 percent of fee-for-service costs to some lower rate. Another way is to break the link between payments to HMOs and costs in the fee-for-service sector.
    As fee-for-service enrollment shrinks, it makes less sense to base updates to HMO payments on fee-for-service spending in an area. Lower payment rates would achieve savings for Medicare, but at the same time they would slow the growth in HMO enrollment. In the longer run, a market-based strategy offers the most promising approach to slowing the growth of Medicare spending. Such a strategy would be based on a more competitive market and a defined contribution from the Federal Government. Although a complete restructuring of Medicare could require years to achieve, practical steps to begin that process could be undertaken now.
    In particular, taking steps to make risk-based plans more widely available to beneficiaries would be conducive to long-term reform even though those steps would not reduce Medicare spending in the short run. One easy step would be to introduce a coordinated open enrollment period, similar to that in the Federal Employees Health Benefits Program, in which beneficiaries could select from all health plans available in their area. Another possibility would be to reduce the wide disparities in Medicare payments among counties. A third idea would be to expand the array of risk-based plans to include a larger range of managed care and private fee-for-service options.
    The administration's proposal for Medicare managed care incorporates several of the foregoing ideas. The administration would continue to update HMO rates based on changes in Medicare spending, but it would change how those rates are calculated. Specifically, it would make five changes. One, it would reduce the percentage of Medicare's adjusted AAPCC, the adjusted average per capita cost, paid to plans from 95 to 90 percent starting in 2000. Two, it would remove payments for disproportionate share hospitals and graduate medical education from the AAPCC and return those funds directly to teaching and DSH hospitals based on the number of HMO enrollees they served. Three, it would narrow the gap between counties with high and low payment rates by phasing in a blend of local and price-adjusted national rates by 2002 and by setting a minimum payment rate of $350 per month. Four, it would ensure that no county's payment rate in 1998 and 1999 is reduced from the level in the previous year. That proposal includes a computation for budget neutrality intended to ensure that this provision and the $350 floor on payment rates would not increase HMO payments overall. And five, it would update the new payment rates by the growth in national Medicare spending per capita—with a minimum update of 2 percent a year—beginning in 2000.
 Page 15       PREV PAGE       TOP OF DOC
    The administration's proposal also contains several features intended to make HMO enrollment more attractive to beneficiaries. I will list three of them: One, allow contracting with additional types of plans, including preferred provider organizations and provider-sponsored networks; two, establish an open enrollment period and provide beneficiaries with standardized comparative information about plans; and three, guarantee that Medigap coverage would be available at community rates for beneficiaries choosing to disenroll from a Medicare HMO.
    The administration is still working out the details of some of its proposals, and CBO's analysis of the budget is not yet complete. However, we have reached some tentative conclusions about the effects of the administration's plan on HMOs. CBO estimates that the administration's proposal would not significantly increase or decrease enrollment in managed care plans. Reducing disparities between high- and low-cost areas, using a coordinated enrollment period, and contracting with additional types of plans, would tend to expand the managed care program. But enhancing the benefits package in fee-for-service Medicare and reducing HMO payments relative to those in the fee-for-service sector would tend to lower enrollment in managed care.
    Guarantee of Medigap coverage on disenrollment raises more complex issues. Such a guarantee could encourage HMO enrollment by easing beneficiaries' worries that they might be locked into a plan they did not like, but it would simultaneously encourage the disenrollment of sicker beneficiaries from HMOs, compounding selection problems and causing Medigap premiums to increase.
    CBO is currently estimating the effects of the administration's proposal on Medicare spending. The estimate of managed care savings depends upon both the managed care and the fee-for-service policies. We anticipate, however, that our estimate of savings in payments to managed care will fall somewhat short of the $34 billion that the administration projects. That ends my brief summary, Mr. Chairman.
    [The prepared statement and attachment follow:]
 Page 16       PREV PAGE       TOP OF DOC

Statement of Paul N. Van de Water, Ph.D., Assistant Director for Budget Analysis, Congressional Budget Office

    Mr. Chairman and Members of the Committee, thank you for inviting me to discuss the role of health maintenance organizations (HMOs) in Medicare. Although the Congressional Budget Office (CBO) has recently reduced its projections of spending for Medicare, the program's costs are still growing at rates that contribute substantially to the deficit in the short term and are unsustainable in the long run. At the same time, CBO projects that enrollment in Medicare HMOs will continue to increase substantially. Those two projections naturally prompt the question: can risk-based health plans become the foundation for a sustainable Medicare program?

Projections of Medicare Spending

    CBO projects that spending for Medicare primarily for medical benefits will increase from $194 billion in 1996 to $317 billion in 2002 and $469 billion by 2007, an average annual increase of more than 8 percent. Although growth in Medicare has slowed since the late 1980s and early 1990s, it will continue to outpace the growth in resources that finance the program (see Table 1). CBO projects that federal revenues will grow by only 5 percent a year about the same rate as the economy.
    Outlays for Hospital Insurance (HI) benefits will increase more rapidly than payroll tax revenues, depleting the HI trust fund by the end of 2001 (see Table 2). Moreover, because premiums for the Supplementary Medical Insurance (SMI) program may increase by no more than the Social Security cost-of-living adjustment after 1998, the share of costs covered by premiums will continue to shrink.
 Page 17       PREV PAGE       TOP OF DOC
    CBO's projections assume that the number of Medicare beneficiaries enrolled in HMOs will grow rapidly. Although most beneficiaries remain in the traditional fee-for-service plan, enrollment in risk-based HMOs jumped from 7 percent of the total in 1995 to 11 percent in early 1997. CBO projects that the fraction of beneficiaries in such plans will approach 25 percent by 2002 and 35 percent by 2007. With growth in overall enrollment in Medicare rising very slowly over that period, the number of beneficiaries in Medicare's fee-for-service sector will decline in absolute terms (see Figure 1).
"The Official Committee record contains additional material here."


    Several factors drive the strong growth in enrollment in risk-based plans that CBO projects over the next decade. First, because of the rapid shift in enrollment from fee-for-service to managed care plans in the private sector, an increasing proportion of people becoming eligible for Medicare at age 65 will already be HMO members. Second, rising premiums for Medigap coverage in the fee-for-service sector will make HMO enrollment relatively more attractive. Third, many employers are dropping or reducing the generosity of health insurance coverage for retirees.

Medicare's Payments to Health Maintenance Organizations

    As a result of the rapid increase in HMO enrollment, payments to managed care plans are the fastest growing element of Medicare spending. CBO projects that such payments will increase from $18 billion in 1996 to $73 billion in 2002 a 26 percent average annual rate of growth. In contrast to the private sector, however, where the increasing importance of HMOs and other health care plans has helped to slow the growth in health insurance premiums dramatically, under current Medicare payment policies the program incurs no financial benefit from higher HMO enrollment. In fact, the available evidence suggests that the growth in HMO enrollment actually works to Medicare's disadvantage.
 Page 18       PREV PAGE       TOP OF DOC

Current Medicare Payment Polices

    Medicare's current payment system for risk-based managed care plans is, by design, unrelated to their cost of doing business. Medicare pays risk-based HMOs 95 percent of the adjusted average per capita cost (AAPCC) for each beneficiary. The AAPCC is an estimate of what a similar beneficiary in the fee-for-service sector in the same county would cost the program. That payment mechanism was intended to allow Medicare to claim some of the savings expected from the more efficient practices of HMOs, while permitting any additional savings to be shared between the HMO and its Medicare enrollees.

Risk Selection

    The calculation of the AAPCC takes into account a number of characteristics of beneficiaries that influence Medicare's costs: their age, sex, institutional status, Medicaid eligibility, disability status, and whether they have primary health insurance coverage through an employer. Those adjustments for health risk are crude, however, and result in overpayments.
    Numerous studies suggest that Medicare's payment formula does not adequately adjust for differences in health status between HMO enrollees and fee-for-service beneficiaries. The Physician Payment Review Commission, for example, has estimated that new HMO enrollees incur below-average Medicare costs in the six months prior to their enrollment and that beneficiaries disenrolling from HMOs have above-average costs in their first six months back in the fee-for-service sector.
    The consensus of the literature is that Medicare pays about 5 percent more on behalf of HMO enrollees than it would have paid if they had remained in the fee-for-service sector. One should note, however, that the available analyses are all based on a period when very few Medicare beneficiaries were enrolled in HMOs. The extent of Medicare's overpayment would decline in the future if HMO enrollment becomes more representative of the entire Medicare population.
 Page 19       PREV PAGE       TOP OF DOC

Regional Variations

    Medicare's current payment system also results in large differences in payment rates among the nation's counties. The ratio of the highest AAPCC to the lowest is currently more than 3 to 1. Even adjusting for regional differences in input costs, that ratio is 1 1/2 to 1.
    The variation in price-adjusted payment rates stems from regional differences in patterns of medical practice in Medicare's fee-for-service sector, as well as from differences in the health status of fee-for-service beneficiaries. Regional variation in payment rates contributes to differences in the benefits that HMOs offer to Medicare beneficiaries. HMOs in counties with a high payment rate are able to offer more generous benefits to enrollees than HMOs in counties with a low payment rate. Regional differences in benefits in turn contribute to regional differences in rates of enrollment in HMOs.

Achieving Savings From Risk-Based Plans

    How can Medicare's HMO program be changed to achieve budgetary savings in the short term and help assure Medicare's survival in the longer run? The shift in enrollment toward risk-based HMOs will not slow Medicare spending unless the program can retain some of the savings possible from managed care plans. In the short run, that result could be achieved in a couple of ways.
    The simplest alternative would reduce Medicare's payment rate from 95 percent of fee-for-service costs to some lower percentage. That approach would save money for Medicare. However, it would also diminish the attractiveness of HMOs to beneficiaries, because HMOs would be less able to offer their current array of additional benefits. Yet in markets in which both the payment rate and enrollment in HMOs are high, those effects are likely to be small.
 Page 20       PREV PAGE       TOP OF DOC
    Another way of achieving savings is to break the link between payments to HMOs and costs in the fee-for-service sector. As fee-for-service enrollment shrinks, it makes less sense to base updates to HMO payments on fee-for-service spending in an area. One option is to set the rate of growth of risk-based payments so that it equals an external factor, such as the rate of growth of the economy. Lower updates would achieve savings for Medicare, but they would also slow the growth in HMO enrollment.
    In the longer run, a market-based strategy offers the most promising approach to slowing the growth of Medicare spending. Such a strategy would be based on a more competitive Medicare market and a defined contribution from the federal government. Although a complete restructuring of Medicare could require years to achieve, practical steps to begin that process could be adopted now. In particular, taking steps to make risk-based plans more widely available to beneficiaries would be conducive to long-term reform, even though those steps would not reduce Medicare spending in the short run.
    One easy step would be to overhaul Medicare's enrollment procedures. Although beneficiaries are currently given a list of risk-based plans in their area, no single, reliable source of information compares the features of those plans. Moreover, most beneficiaries are automatically enrolled in fee-for-service Medicare on first gaining eligibility. Instead, Medicare could institute a coordinated open-enrollment process similar to that of the Federal Employees Health Benefits Program in which beneficiaries could select from all health plans operating in their area. Beneficiaries would receive uniform information on all plans regarding benefits, costs, and access to providers.
    Another possibility would be to reduce the wide disparities in Medicare payments among counties. That move would provide an incentive for risk-based plans to serve areas that now have low payment rates. At the same time, reducing payment rates in some high-cost areas could bring payments more into line with the costs that plans in those areas incur in covering Medicare services.
 Page 21       PREV PAGE       TOP OF DOC
    A third idea would be to expand the array of risk-based plans to include a larger range of managed care and private fee-for-service options. Beneficiaries would be better able to find plans that suited their needs if the range of options was expanded, although doing so would also increase the possibilities for favorable selection. Offering a wider variety of plans could also raise a number of regulatory issues, such as solvency requirements, standards for quality of care, and antitrust considerations.

The Administration's Proposal

    The Administration's proposal for Medicare managed care plans incorporates several of the foregoing ideas. The Administration would continue to update HMO rates based on changes in fee-for-service spending. However, it would change how those rates are calculated, and it would encourage additional HMO enrollment by expanding the type of plans eligible to participate and changing the regulations governing Medigap insurance.

Payments to Health Maintenance Organizations

    The Administration's proposal would modify payments to HMOs in a number of ways. Specifically, it would:
    •  Reduce the percentage of the AAPCCs paid to plans from 95 percent to 90 percent starting in 2000.
    •  Phase in a reduction in the AAPCCs to reflect payments for dis-proportionate share (DSH) hospitals and graduate medical education. Removing those so-called special payments from the AAPCCs would reduce payment rates by about 5 percent. Those funds would be returned directly to teaching and DSH hospitals based on the number of HMO enrollees they served.
 Page 22       PREV PAGE       TOP OF DOC
    •  Narrow the gap between high- and low-AAPCC counties by phasing in a blend of 70 percent local and 30 percent national rates by 2002, and by setting a minimum AAPCC of $350 per month.
    •  Ensure that no county's AAPCC in 1998 and 1999 is reduced from its 1997 level. The proposal includes a computation for budget neutrality intended to ensure that this provision and the $350 floor on the AAPCCs would not increase HMO payments overall.
    •  Guarantee that the AAPCCs would be updated by a minimum 2 percent a year beginning in 2000.

Measures to Encourage Enrollment in Health Maintenance Organizations

    The Administration's proposal also contains several features intended to make HMO enrollment more attractive to beneficiaries. It would:
    •  Allow contracting with additional types of plans, including preferred provider organizations and provider-sponsored networks.
    •  Coordinate changes in HMO enrollment status through third-party brokers, provide beneficiaries with standardized comparative materials about eligible plans and Medigap policies, and establish an annual open-enrollment period.
    •  Guarantee that Medigap coverage would be available at community rates for beneficiaries choosing to disenroll from a Medicare HMO.

Impact of the Administration's Proposal

    The Administration is still working out the details of some of its proposals, and CBO's analysis of the budget is not yet complete. However, we have reached some tentative conclusions about the effects of the Administration's plan.
 Page 23       PREV PAGE       TOP OF DOC
    CBO estimates that the Administration's proposal would not significantly increase or decrease enrollment in managed care plans. Reducing disparities between high- and low-cost areas, using a coordinated enrollment period, and contracting with additional types of plans would tend to expand the managed care program. But enhancing the benefits package in fee-for-service Medicare and reducing HMO payments relative to those in the fee-for-service sector would lead to lower enrollment in managed care plans.
    The guarantee of Medigap coverage on disenrollment raises more complex issues. Such a guarantee could encourage HMO enrollment by easing beneficiaries' worries that they might be locked into a plan they did not like. But it would encourage the disenrollment of sicker beneficiaries from HMOs, compounding selection problems and causing Medigap premiums to increase.
    CBO is currently estimating the effects of the Administration's proposals on Medicare spending. The estimate of managed care savings depends on both managed care and fee-for-service policies. We anticipate, however, that our estimate of savings in payments to managed care plans will fall short of the $34 billion that the Administration projects.


Table 1

Table 2


    Chairman THOMAS. Somewhat short? Somewhat short? I was just commenting on the ''somewhat short.''
    Mr. VAN DE WATER. Oh, I am sorry.
    Chairman THOMAS. I do not know what ''somewhat short'' means, but we will pursue that.
 Page 24       PREV PAGE       TOP OF DOC
    Ms. Wilensky.


    Ms. WILENSKY. Thank you, Mr. Chairman. Thank you for inviting me here to present the views of the Physician Payment Review Commission on payment changes for HMO payment under Medicare. I would like to go through a few of the technical and policy issues that the PPRC has recently commented on and will be included in our 1997. Then I would like to talk about some of the policies that flow from those conclusions.
    There has been rapid growth in HMOs under Medicare with about 13 percent of the Medicare beneficiaries now enrolled in HMOs. Appropriately, there is a lot of concern about the method used to set payments to HMOs that is based on fee-for-service spending in the local area adjusted for the characteristics of enrollees. These concerns include wide geographic variation in payments, and also that the payments are volatile. There have been concerns raised that the current methods for risk adjustment are not adequate to the job. I am going to talk more about that in 1 minute. Another concern is the use of the earmarked funds, particularly for disproportionate share and graduate medical education. And finally there has been some concern in using the county as the basis for payment rates and whether a unit that is that small geographically is causing some of the volatility problems.
    In looking at ideas for changing payments to HMOs, there are several ways to go about it. One is to try to improve the basic payment now. That is the AAPCC. Another is to try to unlink the payment from fee-for-service. Now that could mean such things as has been proposed in the past as blending. That is, taking the national rate and the local rate and trying to not have wide disparities. It could mean something called trimming where you knock off the very low values now, some of the 250-dollar-per-member-per-month counties that now exist, and also some of the very highest rate values such as in Dade County where you have over 750-dollar-per-member-per-month payments. Or you could then find new ways to try to set those payments.
 Page 25       PREV PAGE       TOP OF DOC
    One of the issues I would like to raise—this will come up for a number of policies that you may be considering—is that if you break the link from the local area fee-for-service, you need to be concerned about driving a wedge in the payment level available for HMOs versus what is going on in the fee-for-service market. As much as we may worry about inequities of why someone under Medicare should be able to have $750 if they live in Florida per month and why they should have payments of $250 if they live in certain counties in Nebraska, if you start making the payment differential too great within a geographic area, you are inviting unintended adverse consequences. So while you may want to think about reforming the payment—I think you should and I think there are ways to do that—you need to worry about driving a wedge within geographic markets.
    In addition to unlinking from the fee-for-service fee, you could use market-driven competitive rates if you wish to do so. Probably, the most important single issue that you need to deal with is risk adjustment, and one of the recommendations of the Physician Payment Review Commission is that we need to start now. As you know, the President and administration has recommended going from 95 to 90 percent on the grounds that healthier people go to HMOs. This very crude blanket reduction will invite the kind of selection against the fee-for-service program that is now of concern. In fact, what we need to do is start doing better adjustment and to begin to do some now.
    Let me give some ideas about how you could do that. One is by using administrative data. Use some of the fee-for-service claims data for new enrollees. Use some of the hospital no-pay bills to look at mortality rates which would provide us with information in terms of how much movement we might make toward average payment exists. We can begin to make some modest payment changes now based on administrative data such as making lower payments for new enrollees. This would recognize the diversity among plans in the distribution of new enrollees. Some plans have more than 70 percent of their members who have been in the plan for more than 6 years. Rather than reduce the amount paid to all HMOs no matter what their own enrollees look like, you could begin to take account for some of the different risks without inviting the problem that you are trying to correct.
 Page 26       PREV PAGE       TOP OF DOC
    Finally, let me close on a point that has been raised both by you in the Subcommittee and by a number of commentators, and that is this notion as to whether the government ought to be looking to HMOs as a way to save Medicare money. I hope you will at least consider the notion that perhaps the government should try to be neutral with respect to where seniors seek to receive their care and to come up with payment policies that neither reward nor punish people for choosing various types of delivery systems. That means you must take account of risk selection as it occurs, but it would push the notion of looking at the proper payment irrespective of whether a senior chooses to receive traditional Medicare, to go to a network plan, or to go to an HMO. Your endpoint is somewhat different in terms of the role the government is attempting to carry out.
    Thank you very much, Mr. Chairman.
    [The prepared statement follows:]

Statement of Gail R. Wilensky, Ph.D., Chair, Physician Payment Review Commission

    Mr. Chairman and members of the Committee, I am pleased to present the Physician Payment Review Commission's views and recommendations on several issues related to payment under Medicare managed care. Expansion of managed care and introduction of new private health plan options for Medicare beneficiaries present both opportunities and challenges. The Commission has been working closely with congressional committees and staff to provide analysis and recommendations that can help inform your deliberations. Any policy changes should further the goals of ensuring Medicare's financial solvency and beneficiary access to timely, appropriate health care services. Accomplishing these goals, however, creates a tension between setting payments that are high enough to provide access but are also affordable.
    Over the past decade, there has been tremendous change in how Americans pay for and receive health care. Pressures to reduce growth in health care spending have created a new awareness among consumers, purchasers, and providers of the tradeoffs that arise when resources are finite. Managed care has grown in part because of purchasers and consumers' willingness to trade limits on access for lower health costs.
 Page 27       PREV PAGE       TOP OF DOC
    Medicare can learn from the experience of the private sector. In fact, as commercial managed-care penetration grows and managed-care enrollees age into Medicare, it is inevitable that more and more beneficiaries will select this option within Medicare. But it is important to keep in mind that Medicare differs in important ways. First, Medicare managed-care enrollment, while growing, still lags substantially behind commercial enrollment (Figure 1). Second, although managed-care growth in the private sector has been associated with reduced cost growth, under current policy, this is not the case for Medicare. In fact, some studies suggest that managed care growth increases program outlays. Third, while the private market encompasses a broader range of plan options than Medicare currently permits, most individuals with employer-based insurance have only a limited number of plans to choose from.
"The Official Committee record contains additional material here."


    The debate on Medicare managed care always eventually turns to payment. Changes in payment policy could serve several goals: reducing program spending, encouraging managed-care enrollment by making the program more attractive to plans in certain markets, and improving equity by reducing the variation in benefits offered by Medicare managed-care plans in different areas of the country. My testimony this morning focuses on these issues and the range of policy options that could be adopted. The challenge facing policymakers is to develop an approach to paying plans that is fair, reduces cost growth, and ensures that beneficiaries have access to appropriate care at a cost they can afford.
    My statement begins with some brief background information about Medicare managed care and the issues that will arise as managed care choices expand. I will then sketch out how Medicare now pays managed-care plans and the problems associated with current policy which the Commission and others have identified. Finally, I will talk about the different options that the Congress could take to address these problems (including those included in the Balanced Budget Act passed in the last Congress and the President's recent budget proposal) as well the Commission's recommendations concerning implementation of these options.
 Page 28       PREV PAGE       TOP OF DOC

Medicare Managed Care: Plan Participation and Beneficiary Enrollment

    As you know, Medicare managed care is growing. By the end of 1996, about 13 percent of Medicare beneficiaries were enrolled in some form of managed care, compared to 5 percent in 1990. Participation by beneficiaries varies widely, with over 20 percent of urban beneficiaries enrolled in managed care, compared to about 1 percent of rural beneficiaries. Although predominantly an urban phenomenon, enrollment rates differ across urban areas. Over half of beneficiaries in Riverside, CA, are in risk plans, for example, while virtually none are in Atlanta and Detroit (Figure 2).
    Most plans participate in Medicare through the risk-contracting program. Under a risk contract, plans commit to providing Medicare-covered services to beneficiaries for a fixed monthly payment from the program. There were 241 risk contracts in effect at the end of 1996; 17 more have been added in the last two months (Figure 3).
    The availability of risk plans varies widely across the nation. In most urban areas, beneficiaries can choose among several plans, while 80 percent of rural beneficiaries have no plan available. Overall, about two-thirds of beneficiaries are served by at least one risk plan; 25 percent have access to more than four plans (Figure 4).
"The Official Committee record contains additional material here."


"The Official Committee record contains additional material here."

 Page 29       PREV PAGE       TOP OF DOC
"The Official Committee record contains additional material here."


Current Policy Affecting Risk-Plan Payment, Benefits, and Premiums

    Now let's consider the current policies that determine how much risk plans are paid and the benefits and premiums that enrollees receive. Going over a few of the basics will be helpful in understanding the problems created by these policies.
    As a result of current policies and local competitive pressures, there is wide geographic variation in payments to plans, in the benefits available to beneficiaries, and in the premiums that they pay. For example, there is a three-fold difference between the lowest and highest county payment rates (Figure 5). Over 50 percent of 1997 county rates, however, are between $340 and $440. Currently, more than three-quarters of risk plans offer additional eye and ear care, and over half provide prescription drug coverage (Figure 6). By the end of 1996, two-thirds of plans provided benefits beyond those covered by Medicare at no additional charge to enrollees (Figure 7).
"The Official Committee record contains additional material here."


"The Official Committee record contains additional material here."

"The Official Committee record contains additional material here."
 Page 30       PREV PAGE       TOP OF DOC


Setting Payments and Benefits

    Payments, benefits, and premiums are the result of two separate administrative processes, as well as of local competitive pressures.
    Process for Setting Plan Payments. Payments are set to reflect local fee-for-service costs. This measure is referred to as the AAPCC. Actual per capita spending is adjusted for differences in the characteristics of local populations. The resulting AAPCC is the expected local cost of caring for a typical beneficiary. Each county's payment is set at 95 percent of the AAPCC. Plans are paid this rate with an adjustment for the characteristics of their enrollees.
    This two-step process of setting a local rate for a typical beneficiary in each county and then adjusting payments to plans based on actual enrollment was designed with two purposes. First, expected spending on managed care should equal that in fee for service less the 5 percent savings. Second, plans should be fairly compensated for the relative risks of their enrollees.
    In setting both the local rate and the payment to a plan, adjustments are made to reflect the characteristics that affect beneficiaries' use of health care. The same five risk adjusters are used in both steps: age, sex, welfare status, institutional status, and working status. Separate adjustments are made and AAPCCs calculated for the aged, disabled, and end-stage renal patient populations.
    Process for Establishing Required and Optional Benefits. The benefits and premiums that risk plans offer to beneficiaries are set in a second process. Plans submit adjusted community rate (ACR) proposals in which they estimate the cost of providing Medicare-covered services to enrollees based on the costs of serving their commercial population. If Medicare pays a plan more than these estimated costs, then the plan must return the difference to Medicare or to beneficiaries in the form of additional benefits. In practice, all plans opt to provide additional benefits to beneficiaries. The Commission estimates that in 1995, enrollees received additional benefits worth about $42 per month for which they paid no additional premium.
 Page 31       PREV PAGE       TOP OF DOC
    In response to local competition, plans may also choose to offer even more benefits. The ACR proposal establishes the maximum premium that plans can charge for these optional benefits, but plans can choose to waive all or part of this premium. In 1995, enrollees received optional benefits worth about $45 per month for which they paid an average of $18 per month.

Concerns About Current Policy

    The wide geographic variation and volatility in spending for traditional Medicare results in large differences in the AAPCC across counties. These differences in turn affect patterns of managed-care enrollment, premiums, and benefits across the country. They may contribute to the uneven pattern of Medicare managed-care enrollment that I described earlier. And they account, at least in part, for the wide and seemingly arbitrary variation in additional benefits that Medicare beneficiaries receive from risk plans in different markets.
    Several factors that could be addressed in legislation contribute to this geographic variation. The most important of these are:
    •  Inadequacies of current demographic risk adjusters. Inadequate risk adjustment results in increased Medicare spending in two distinct ways. First, local rates may overstate the likely cost of a typical beneficiary because the AAPCC reflects only beneficiaries who remain in fee-for-service and who have higher costs than managed-care enrollees (Figure 8). If these beneficiaries are less healthy than those in managed care and their poorer health is not captured by the current demographic adjusters, then expected fee-for-service payments are overstated. This is referred to as base-rate bias. Better adjusters would make the AAPCC a more accurate reflection of expected outlays for a typical beneficiary and would reduce some of the variation in payments.
 Page 32       PREV PAGE       TOP OF DOC
    Second, in addition to the local rate being too high, inadequate risk adjustment results in overpayments to plans for their particular enrollees. Current risk adjusters explain only a small portion of the variation in health care costs among Medicare beneficiaries. A more accurate set of risk adjusters would result in lower payments to plans reflecting their relatively healthier enrollment.
"The Official Committee record contains additional material here."


    As I will explain in a moment, the Commission plans to make a series of recommendations concerning risk adjustment in its 1997 annual report to the Congress due on March 31.
    •  Inclusion of earmarked funds. Medicare makes payments to hospitals for graduate medical education and for serving a disproportionate share of low-income patients. Including these special funds in AAPCC-based rates contributes to geographic variation in managed-care payments. It also raises the question of whether these payments should be passed along to all managed-care plans, since they are targeted to compensate specific hospitals for special circumstances beyond the costs of caring for Medicare patients.
    The Commission has recommended that these funds could be removed from the AAPCC. A related but separate issue is whether teaching and disproportionate share hospitals should receive additional compensation for seeing managed-care enrollees or whether managed-care plans should be compensated an additional amount for teaching or serving low-income patients. The Commission recommends that mechanisms be developed to ensure that hospitals and managed-care plans involved in training are paid fairly for these costs.
    •  Geographic basis of rates. Use of counties, which are relatively small geographic units, in setting payments leads to more geographic variation and volatility than may be appropriate. Variation and volatility reflect several factors, such as differences in practice patterns, difference in the health status of local populations, and, at least in some cases, small numbers of beneficiaries. Areas larger than counties would help address problems with the AAPCC and may be more consistent with the notion that managed-care plans serve markets, not counties. Using larger areas, however, loses information about the variation in health status at the county level that contributes to the accuracy of payment. For these reasons, any changes to geographic areas should be accompanied by implementation of better risk adjusters.
 Page 33       PREV PAGE       TOP OF DOC
    It is important to recognize that even if all of these technical issues were resolved, under current policy, savings from managed-care enrollment can not exceed 5 percent. Because managed-care payments increase in lock-step with Medicare fee-for-service expenditures, cost increases in fee for service drive cost increases throughout the program. Expanding managed-care without increasing outlays will require breaking the link between managed-care payments and fee-for-service expenditures.

Proposals for Change

    Over the past two years, the Congress and the Administration have been considering how to set Medicare capitated rates that are fair to plans and allow the program to benefit from managed-care efficiencies. Proposals to improve risk-plan payment policies were included in the Balanced Budget Act passed during the 104th Congress. Similar proposals were introduced by Senator Daschle and supported by the Administration last year and were more recently put forward in the President's fiscal year 1998 budget proposal. All of these proposals included provisions previously recommended by the Commission.
    There are basically three different ways to reduce the variation in risk-plan payment rates. These approaches could be implemented to achieve budget savings, or could be budget-neutral, focused solely on reallocating payments across areas.
    The first approach is to improve the AAPCC. Improving risk adjustment, removing earmarked hospital payments, and changing the geographic basis of the local rate would all result in better estimates of patient care costs, which would differ less across areas.
    A second approach is to unlink risk payments from local spending, using current rates as a starting point for setting new rates. A variety of strategies could be used to set rates which have less geographic variation than those now based on the AAPCC. These include blending current local rates with national rates, trimming rates through floors and ceilings, and setting new ways to update local rates. Since these approaches begin with the AAPCC, the Commission recommends that if they are adopted, that they be adopted in tandem with the improvements in the AAPCC that I just mentioned.
 Page 34       PREV PAGE       TOP OF DOC
    Finally, current policy could be discarded altogether in favor of market-driven competitive solutions. Under this approach, local market characteristics would be used to set rates, either through some form of competitive bidding or a defined federal contribution for both fee-for-service and risk beneficiaries. This approach would work only in markets with sufficient local competition. It could be adapted to markets with little managed-care penetration if payments are based on the experience of both managed-care and fee-for-service beneficiaries. The Commission has recommended that the Health Care Financing Administration (HCFA) test such alternative methods for setting payments, including competitive bidding and partial capitation.

The Importance of Risk Adjustment

    Regardless of how payment rates are set, as long as Medicare beneficiaries can choose among options, improved risk adjustment will be essential. Otherwise, plans will not be fairly paid for enrollees with better or worse-than-average health status (for example those with chronic conditions or functional disability). Without improvements in risk adjustment, plans will continue to have an incentive to avoid enrolling patients who will be expensive to care for.
    The Commission recommends that improved risk adjustment be implemented immediately. Although available approaches are not perfect, they would do a better job than the demographic factors currently used. As a first step, the Commission recommends that Medicare begin to phase-in risk-adjusted payment changes using administrative data. For example, our analyses and those of others would support an approach of paying less for new managed-care enrollees who have lower-than-average per capita costs. (New enrollees now account for 55 percent of Medicare managed-care enrollees, up from 43 percent in 1993.) Because there are substantial differences among plans in the proportion of new enrollees, this approach would be preferable to an across-the-board cut which would particularly hurt those plans with a large proportion of long-time enrollees (Figure 9). Since risk adjustment methods typically underpredict the true variation in costs and selection, improvements such as paying less for new enrollees do not risk over adjusting (that is paying too little) for individuals with certain characteristics.
 Page 35       PREV PAGE       TOP OF DOC
    Steps could also be taken immediately to improve the availability of data useful for risk adjustment. For example, hospitals are now required to submit ''no-pay'' bills to HCFA for hospitalized managed-care enrollees but many do not do so. The potential use of these data for risk adjustment increases the importance of enforcing this requirement.
    Use of administrative data for risk adjustment is only the first step. Over the longer term, the data and infrastructure required to support risk adjustment should be developed and implemented. This includes obtaining data that more accurately captures risk (such as those obtained from surveys of beneficiaries or encounter data collected by plans and their contracting providers), further development of risk adjustment models, and implementation of adjusted payment rates.
"The Official Committee record contains additional material here."


    The President's budget proposal includes a provision aimed at recovering some of the overpayments due to inadequate risk adjustment. It calls for setting local rates at 90 percent of the AAPCC, instead of the 95 percent under current policy. Although this would mitigate the budget impact of risk selection against the fee for service program, it would not adjust for risk selection among managed-care plans and so would not reduce plans' incentives to avoid enrolling costly beneficiaries.

Effects of Change

    The effect of any payment changes on total Medicare payments, plans, and beneficiaries will ultimately depend upon how they are implemented, how much payment levels change, and how plans and beneficiaries respond. The effect of payment floors, blended rates, and other approaches to reducing inappropriate variation in risk plan payments will differ, depending upon the exact combination of policies and the sequence in which they are calculated.
 Page 36       PREV PAGE       TOP OF DOC
    The effects of changes on plan participation and beneficiary enrollment are also uncertain. If plans and beneficiaries are sensitive to payment rates, then rate changes could lead to participation increases in areas with increased rates and declines in those where rates drop. But if plans and beneficiaries are relatively insensitive to risk-plan payment rates, then we might not see such effects.
    Unfortunately, there is little information that could guide us in predicting how plans and beneficiaries will react to payment changes. Researchers have been examining this question but their conclusions have been mixed. A staff analysis for the Prospective Payment Assessment Commission (ProPAC) indicates that plan entry into the risk program is highly sensitive to the local payment rate. Another recently published study found that beneficiary enrollment rates are much more sensitive to factors such as local managed-care penetration in the commercial market than to local Medicare rates.
    If risk payments differ from per capita fee-for-service outlays, then more detailed information about beneficiaries' enrollment behavior will be required in order to make accurate budget projections. In particular, it will be important to understand how beneficiaries of different risk categories select between managed care and fee for service.


    It is important to recognize that payment policy is only one of the factors that will determine the future of managed care within Medicare and its impact on the federal budget, beneficiaries, and providers. Realizing the potential of Medicare managed care will also require policy changes to minimize risk selection. Policies concerning information available about choices, the enrollment and disenrollment process, and enrollee grievance procedures must work together to allow plans to compete effectively and to protect beneficiaries. The Commission has made a variety of recommendations about these topics that I hope will provide the Congress some guidance.
 Page 37       PREV PAGE       TOP OF DOC
    I would also like to take the opportunity to mention that since the vast majority of Medicare beneficiaries remain in fee-for-service (and are likely to do so for the next decade), the Commission has also devoted some time to issues related to improving traditional Medicare's performance. I would be glad to provide information about these issues to the Committee.


    Chairman THOMAS. Thank you, Dr. Wilensky.
    Dr. Newhouse.


    Mr. NEWHOUSE. Thank you, Mr. Chairman, for inviting me here to testify. I am representing Prospective Payment Assessment Commission and have Donald Young with me. As you noted in your opening statement and as our chart 1 demonstrates, the risk program at one level has been quite a success because enrollment has grown strikingly. Since 1993, it has grown 32 percent a year. At another level, it has not been successful in achieving savings for the Medicare Program. There are many reasons for that that the previous witnesses have alluded to and, in fact, in our report that will be forthcoming to you next week, we recommend a number of modifications that are largely consistent with what you have heard from the prior two witnesses.
    Our recommendations focus on risk adjustment, on payment amounts, and on risk plan information. I will talk about risk adjustment and payment amounts this morning. The research that has been done on risk adjustment suggests that the Medicare payments for the at-risk plan enrollees are an estimated 5 to 7 percent too high. That is the basis for the administration recommendation. It means that Medicare is losing rather than saving money on the risk program. Those overpayments would be reduced if the payments for enrollees were adjusted to account for their likely use of services.
 Page 38       PREV PAGE       TOP OF DOC
    The overpayments are, of course, controversial, but I do not think it is so controversial that an adequate risk adjustment method should not be implemented that would reduce payments for any plans that had healthier enrollees and pay more to plans that had sicker beneficiaries—all plans are not alike. Research has gone along two different lines in developing better risk adjustors. One uses information on diagnosis. If a plan enrolled a woman with breast cancer, for example, it would pay that plan more. The other uses self-reported health and functional status. A risk adjustment method could use both kinds of information, but diagnosis information alone does about as well as both methods and that is what we would recommend.
    Even the best available risk adjustment method, however, does not fully offset any efforts by plans that are going to seek out healthier beneficiaries, which means that research on risk adjustment ought to continue. And in the meantime, a partial capitation method should be investigated to reduce risk selection. That would partially pay plans on the basis of actual utilization, which would presumably be less for plans that had healthier members.
    Our chart 3 documents the geographic variation that the previous witnesses have alluded to in the AAPCC. We would also like to reduce that. After you adjust for differences in input prices, you see that there is about a 200-dollar-a-month spread for urban and rural areas. In chart 4, you see the spread within the Washington, DC, metropolitan area going from about $600 a month in Prince George's County to around $400 a month in Fairfax. We recommend several changes that would reduce the geographic spread. We also recommend taking out the teaching and disproportionate share payments, accounting for use in military and veterans facilities, and making other changes in payments.
    We also agree with the previous witnesses that it would be good to break the link with the fee-for-service payment system, and that judgment could be used in updating the rates as it is on the fee-for-service side. In the longer run, we would like Medicare to investigate more competitive methods such as bidding or negotiations. Let me conclude by referring to chart 6 which shows that limiting payments and cutting the geographic spread will have the effect probably of reducing benefits to beneficiaries. What this chart shows is that areas in the high payment rate areas provide more extra benefits than plans in areas where the AAPCC is less, which is not terribly surprising. Since I see my time is up, let me stop, and I would be happy to answer your questions as best I can.
 Page 39       PREV PAGE       TOP OF DOC
    [The prepared statement follows:]

Statement of Joseph P. Newhouse, Ph.D., Chairman, Prospective Payment Assessment Commission

    Good morning, Mr. Chairman. I am Joseph Newhouse, Ph.D., Chairman of the Prospective Payment Assessment Commission (ProPAC). I am accompanied by Donald Young, M.D., the Commission's Executive Director. We are pleased to be here today to discuss improvements to Medicare's risk contracting program. During my testimony, I will refer to several charts. These charts are appended to the end of my written testimony.
    In 1985, Medicare implemented the risk contracting program. Under this program, participating health maintenance organizations (HMOs) receive a monthly capitation payment to provide the Medicare benefit package to each beneficiary they enroll. The risk program was created to allow Medicare to enjoy some of the advantages of capitation arrangements, such as predictable spending and savings. Beneficiaries who join risk plans also benefit because many plans provide additional services and have low cost sharing requirements.
    On one level, the risk program has been a success because more and more beneficiaries are choosing to receive services under these arrangements. Since 1993, enrollment has increased, on average, 32 percent each year. Today, 4.2 million beneficiaries, or 11 percent of the total Medicare population, have chosen this option for their health care coverage (see Chart 1).
    The risk program has yet to be successful, however, in its goal of achieving savings for the Medicare program. Capitated managed care arrangements have the potential to restrain Medicare expenditures because they create incentives to control the number of services furnished, as well as the cost of each unit of service. These arrangements have helped to curb spending in the private sector. To date, however, the risk program has not achieved the savings that the private sector experience suggests is possible. There are several reasons for this, most notably that Medicare payments to plans do not reflect their enrollees' lower-than-average probability of using health care services. Another reason is that the capitation rates are based on the spending experience of beneficiaries in the fee-for-service program, rather than the costs that would be expected under a managed care arrangement.
 Page 40       PREV PAGE       TOP OF DOC
"The Official Committee record contains additional material here."


    In H.R. 2491, the Balanced Budget Act of 1995, the Congress passed a number of reforms to improve the payment methodology for managed care plans. The President also has proposed a number of modifications in his recent budget proposal. ProPAC agrees with the Congress and the President on the need to better adjust risk payments and to move away from fee-for-service spending as a basis for setting rates.
    In our forthcoming Report and Recommendations to the Congress, the Commission will recommend a number of modifications that it believes are necessary to improve the risk program. We believe that, if adopted, these actions will benefit both the program and its beneficiaries. These recommendations focus on improvements in three areas: risk adjustment, payment amounts, and risk plan information. This morning, I would like to share with you the Commission's views. But first I will briefly summarize the current method for paying risk plans.

The Risk Payment Methodology

    As you know, Medicare pays risk plans a monthly payment for each Medicare enrollee to cover the program's share of costs for Medicare-covered services. This rate is based on 95 percent of projected fee-for-service Medicare program payments (the adjusted average per capita cost or AAPCC) in the county in which the enrollee resides. Separate rates are calculated for aged and disabled beneficiaries and for those who are eligible for Medicare because they have end-stage renal disease. The rates are adjusted by five factors to account for variations in enrollees' health care needs. They are the enrollee's age, sex, Medicaid status, institutionalized status, and whether the person has employer-based coverage. As I will discuss in a moment, these adjustments are not adequate to reflect enrollee spending patterns.
 Page 41       PREV PAGE       TOP OF DOC
    The Medicare program recognizes that risk plans are likely to keep their costs below their payments. While plans are permitted to return to the program any payments that exceed their projected costs, they also may use them to provide extra benefits to risk enrollees. Not surprisingly, most plans choose to offer extra benefits in the form of additional services, lower cost-sharing, or coverage of services from out-of-network providers. To further attract Medicare beneficiaries, plans may include even more benefits than they are required to provide.
    Almost every risk plan provides some type of extra benefits. In 1996, the vast majority of plans covered routine physicals and eye exams. Half of plans offered some type of pharmaceutical benefit and two-thirds charged no premium for their basic package. A ProPAC analysis estimated that in 1995, the average risk plan provided each enrollee with $43 in extra benefits each month. The amount of extra benefits varied tremendously across the country, however, even after adjusting them to reflect differences in local price levels (see Chart 2). In 1995, a tenth of plans offered extra benefits valued at over $100 per enrollee per month while another 10 percent offered less than $1. As I will describe later in my testimony, the level of extra benefits that risk plans provide is associated with the payment rates in the areas the plans serve. The variation in the value of extra benefits suggests that fee-for-service spending patterns are not good predictors of the costs plans might be expected to incur.
"The Official Committee record contains additional material here."


Improving Risk Payments

    Mr. Chairman, as both the Congress and Administration recognize, if managed care is to be a viable option under Medicare, the risk program must be modified. First, the program needs better risk adjustment methods. Second, Medicare must revise the risk payment methodology. Immediate changes would begin to break the link to fee-for-service spending and reduce the variation in payment rates across areas. Over the longer term, Medicare should consider new ways of setting risk payment rates. I would like to briefly address each of these issues.
 Page 42       PREV PAGE       TOP OF DOC

The Risk Adjustment Method

    In concept, the risk program should generate savings for Medicare because the payment rate is 5 percent less than the fee-for-service spending that would be expected for each beneficiary in an area. Instead, however, research has shown that Medicare payments for current risk enrollees are, on average, an estimated 5 to 7 percent greater than if these beneficiaries had remained in the fee-for-service option. Thus, Medicare is losing, rather than saving, money on the risk program.
    These overpayments would be reduced, and spending more appropriately distributed, if payments for enrollees were adjusted to account for their likely use of services. An adequate risk adjustment method would do this. It would reduce risk plan payments relative to fee-for-service spending to reflect the healthier population of risk plans. Further, it would increase payments to plans that serve sicker beneficiaries and reduce them to plans that have healthier enrollees.
    Researchers have been evaluating methods that could be used to better target risk payments. Two have been studied. One uses diagnosis information that accounts for prior health service use. The other is based on enrollee reports of their health and functional status, and past and present health conditions. While a risk adjustment method could be designed that would draw on both types of information, diagnosis information alone measures risk about as well as using both methods together. An outlier policy to address unusually costly enrollees would further improve payments to risk plans.
    An improved risk adjustment system would reduce overall risk plan payments as well as redistribute funds across plans and areas. Therefore, it may be appropriate to phase in a new system over time. Mr. Chairman, we know, however, that even the best available risk adjustment method will not fully offset efforts by plans that seek out healthier beneficiaries. Therefore, research needs to continue to seek further improvements in risk adjustment methods. This would help to ensure that Medicare payments to risk plans reflect the health care needs of their enrollees. In the meantime, a partial capitation method should be investigated as a means to reduce the effects of risk selection. This approach would pay plans partially on the basis of their enrollees' utilization, which would be lower for plans that had healthier members.
 Page 43       PREV PAGE       TOP OF DOC

Risk Plan Base Payments

    Another fundamental problem with the risk program is its reliance on fee-for-service spending to set risk payment rates. This approach has resulted in wide variations in risk payment rates. This year, for example, risk plan payments are based on rates that vary by as much as $500 per member per month depending upon the county they serve. Even after adjusting for differences in local input prices, per person payment rates can vary by as much as $200 per month across both urban and rural areas (see Chart 3).
    In addition, a plan offering services in neighboring counties may receive very different risk payments for enrollees living in those counties. For example, in the Washington, DC area, the 1997 monthly per person rates range from $401 in Fairfax county to $602 in Prince George's county—a 50 percent difference (see Chart 4).
    The current degree of payment variation across areas, particularly among plans within the same area, does not seem to be justified. There are areas where payments are such that risk plans can provide extra benefits. At the same time, payment rates may be too low in other areas, discouraging plans from participating in the program.
    In our upcoming report, the Commission recommends several changes to the current system that would result in more appropriate payment levels. These include removing special payments associated with teaching and disproportionate share hospitals, accounting for services provided in military and veterans' facilities, and making other changes that would increase minimum payment levels and further reduce payment variation. I would like to discuss each of these issues in turn.
    Removing Special Payments—Part of the variation in risk payment rates relates to Medicare fee-for-service payment policies that may not reflect the way managed care organizations operate. Because of the way they are determined, the capitation rates include special payments to hospitals that have graduate medical education programs or serve a disproportionate share of low-income patients. Risk plans, however, are not required to use these providers, or pass along these extra payments to them. Consequently, the capitation rates in these areas may be higher than risk plans' costs.
 Page 44       PREV PAGE       TOP OF DOC

"The Official Committee record contains additional material here."

"The Official Committee record contains additional material here."


    The Commission believes teaching and disproportionate share payments should be removed from the calculation of risk payments. In 1995, these special payments represented about 5.3 percent of total Medicare program spending, with wide variation at the county level. Among the 30 counties with the greatest risk enrollment in 1995, teaching and disproportionate share payments ranged from 1 percent of total fee-for-service spending to almost 20 percent.
    This change would reduce the rates the most in counties where fee-for-service spending is higher because of these special payments. In most counties, however, the amount of these payments is low so that risk payment rates would change only slightly. The Commission also believes a separate mechanism should be developed to make additional payments to teaching and disproportionate share hospitals for the Medicare risk plan enrollees they treat. This is necessary to preserve Medicare beneficiaries' access to care in these facilities and to continue Medicare's support for the special roles these institutions play in teaching, research, and serving the poor.
    Accounting for VA and DoD Services—Another source of variation is due to services received by Medicare beneficiaries in facilities operated by the Departments of Veterans Affairs and Defense that are not accounted for in Medicare's calculation of fee-for-service rates. In those areas where risk enrollees do not use these facilities to the same extent as beneficiaries in the fee-for-service system, risk payments may be too low. In areas with little risk enrollment, these lower rates might discourage risk plan participation. If payment rates were increased in these areas, adjustments might be needed for those risk enrollees that continue to use DoD or VA facilities.
 Page 45       PREV PAGE       TOP OF DOC
    Other Changes—Even with the modifications I have just mentioned, the Commission believes that other changes are necessary to improve capitation payments. In some areas, payment rates may need to be increased to a minimum level to provide adequate payment for the costs of providing Medicare services. This may be especially important in rural areas where sparse populations and less developed health care infrastructures add additional cost requirements for plans. Any increase in payments, however, should be offset either by reducing all payment rates above the minimum level or by lowering the highest rates.
    Overall variation in capitation rates could be constrained in several ways. One way would be to blend local amounts with the national average rate, bringing all payments closer to the average.

Updating Risk Payments

    In addition to recommending changes to risk plan base payment rates, the Commission believes that the method for updating payments must be replaced. Currently, risk payments change each year based on the spending experience in the fee-for-service sector. Because spending in many areas is quite variable, there can be profound changes in risk payments from year to year—especially in counties with few beneficiaries. For example, between 1996 and 1997, the payment rates for several counties jumped by 25 percent or more, while other counties experienced payment decreases of 10 percent or more (see Chart 5). Even for relatively large counties, the rates can vary substantially from year to year. Moreover, the problem can be compounded in areas where relatively healthy beneficiaries are choosing to enroll in risk plans. In those areas, risk plan payment increases are based on the higher spending patterns of sicker beneficiaries remaining in the fee-for-service system. Thus, the payment rates may become increasingly out of line with the costs of serving the risk enrollee population.
 Page 46       PREV PAGE       TOP OF DOC
    Mr. Chairman, this method of updating risk payments is flawed on several fronts. First, and perhaps most importantly, the method provides no way for Medicare to share in savings that occur when risk plan costs increase more slowly than the payment rate. Any difference between payments and costs goes towards extra benefits to enrollees. While extra benefits may be useful to attract beneficiaries, Medicare has no means for retaining any of the excess payment.
    Second, updating risk plan payments based on changes in fee-for-service spending may not reflect the performance of managed care plans in providing services to risk enrollees. The fee-for-service system is fundamentally different than managed care. Spending growth under fee-for-service is driven in large part by increases in the volume and intensity of services provided, which reflect fee-for-service payment incentives. A capitated system, by contrast, seeks to control the volume of services provided. In addition, unlike the fee-for-service system, risk plans can negotiate lower prices with providers and can sometimes shift patients from more expensive settings to less costly ones.
"The Official Committee record contains additional material here."


    The current system for updating risk payment rates should be discarded and replaced by a method that is analytically-based. A formula approach similar to one the Commission uses to recommend hospital payment increases should be implemented. Such a framework would consider factors that are likely to affect plan costs, such as inflation and industry productivity improvements. In this way, Medicare could break the link to fee-for-service spending and permit Medicare to share in the savings associated with any increase in efficiency.

Longer Term Changes to Risk Payments
 Page 47       PREV PAGE       TOP OF DOC

    The Commission believes that risk payments should be based on the costs that an efficiently run plan would be expected to incur in providing Medicare-covered services. As you know, however, this level is difficult to determine. While the Commission believes its recommended changes to the current system will improve the risk payment methodology, it also believes Medicare should begin looking at alternative ways for determining capitation rates. Market-based methods such as competitive bidding and third-party negotiations should be explored. These approaches also would break the link to fee-for-service spending and permit Medicare to take advantage of many of the same forces private sector purchasers have successfully relied on to reduce their health care costs.

Risk Plan Participation and Beneficiary Enrollment

    Changes to the risk payment methodology are necessary to ensure the success of the risk program. These changes have the potential to affect HMO participation in the program as well as beneficiary enrollment. Participation and enrollment could rise in areas where payment levels are increased but could fall in areas where rates are reduced. The likely impact of any changes, however, is difficult to quantify because participation and enrollment depend upon a number of factors. The changes we recommend, however, likely would differ little from changes that any prudent purchaser would impose given similar circumstances.
    A recent ProPAC analysis found that HMOs are more likely to participate in the risk program in urban areas with higher payment rates. At the same time, however, there are areas with relatively low payment rates where HMOs participate in the risk program and areas where there is minimal participation despite relatively high payment rates. This suggests that characteristics of the market as well as of the HMO itself also play a role in participation decisions. ProPAC analyses indicate that larger and older HMOs are more likely to participate. This may indicate that success in the commercial market is an important factor in an HMO's decision to enter the risk market. The extent of risk plan competition in an area also influences participation decisions; HMOs are less likely to enter a market where they would face a number of competitors.
 Page 48       PREV PAGE       TOP OF DOC
    Consequently, decisions to participate in the risk program involve a complex decisionmaking process, of which payment rates are only one factor. If faced with lower payment rates, participating plans can choose to not renew their contracts, but there are less drastic alternatives that plans might pursue. Plans could lower their costs through tightening administrative spending, accepting lower profits, or negotiating more stringent rates with providers. They also could reduce the level of extra benefits they offer beneficiaries.
    Raising payment rates in certain areas would encourage participation, but other factors may limit HMOs' responses. For example, provider shortages and sparse populations may have a greater influence on plan decisions in rural areas. I should point out that participation may increase in all areas if the Congress decides to expand the program to include additional entities, such as provider service organizations. This may be especially relevant in rural areas where providers who already serve Medicare beneficiaries may choose to develop these entities. Again, however, many factors are likely to come into play.
    The impact of rate changes on beneficiary participation also is unclear. A primary reason why Medicare beneficiaries join risk plans is because they can receive extra non-Medicare covered services at no additional costs. ProPAC analyses indicate that plans serving areas with higher payment rates tend to provide richer benefit packages (see Chart 5). But like plan participation decisions, the level of extra benefits offered by plans is influenced by other factors as well. For example, plans in more competitive areas tend to provide a higher level of extra benefits than plans that have little or no risk plan competition.
    Limiting payment rates may reduce the level of extra benefits that risk enrollees would receive. Given the relatively generous extra benefits in high payment areas, it is likely that beneficiaries in these areas would still receive some amount of extra benefits, regardless of any payment reductions. In addition, plans may have other incentives, such as competitive pressures, to forego a share of their profits to maintain a competitive benefit package. It also is possible that as more commercial managed care enrollees age into Medicare, they may choose to continue their coverage under a managed care arrangement, regardless of the level of extra benefits.
 Page 49       PREV PAGE       TOP OF DOC
"The Official Committee record contains additional material here."


Improving Risk Plan Information

    Mr. Chairman, as the risk program continues to expand, it is increasingly important that the program have sufficient information to ensure that risk payments are appropriate and that plans are delivering quality care. In addition, beneficiaries need to have comparative information to make informed choices between competing risk plans, or choosing between the risk option and remaining in fee-for-service.
    Currently, discussions about risk plan payments and costs are hindered because there are no data available on the actual costs risk plans incur to provide Medicare services. The only cost data available are from the adjusted community rate (ACR) proposals that plans annually submit to HCFA. These proposals, used to determine the level of extra benefits that risk plans are required to offer, set forth plans' projected costs in providing the Medicare-covered benefit package, including administrative outlays and profit.
    The process used to arrive at these projections is indirect. Plans estimate the monthly per enrollee costs needed to provide the Medicare benefit package to their commercial population and then adjust these estimates upward to reflect the higher usage rates of an older, sicker Medicare population. There is no mechanism to learn whether, and to what extent, risk plans' actual Medicare costs are above or below their projections.
    These cost projections may be particularly distorted because of the method plans use to calculate their Medicare administrative costs and profit estimate. Risk plans apply the share of their commercial costs that is devoted to administration and profit to their estimated Medicare patient care costs to obtain this estimate. Because Medicare's service-related costs are, on average, about triple those in the private sector, the amount of costs attributable to Medicare administration and profit is also about three times higher. According to ProPAC analyses of 1995 data, plans estimated they would receive, on average, about $20 per month to cover administrative costs and profit associated with each commercial enrollee. Because of the allocation formula, however, these items accounted for about $66 of risk plans' projected Medicare costs per member per month. I should note that these costs do not affect the payment that plans receive, but rather can alter the level of additional benefits that plans may be required to offer beneficiaries.
 Page 50       PREV PAGE       TOP OF DOC
    The Commission recommends that the Secretary require plans to provide information to assess the costs of furnishing services to Medicare enrollees. This information is needed to evaluate the appropriateness of plan payments as well as the relationship between payments and costs of care. This information could also be used to assess whether plans are returning appropriate amounts of excess payments to beneficiaries through extra benefits. This data collection would not need to be overly burdensome. It could be obtained through a process similar to that of preparing the current ACR proposal.
    Information to monitor and assess the quality of care provided by risk plans also is needed. In a managed care environment where there are incentives to provide less rather than more care, concerns about the quality of care are heightened. The Commission supports the Secretary's efforts to evaluate Medicare risk plans through the use of the Health Plan Employer Data and Information Set (HEDIS) and enrollee satisfaction surveys. While this is a good first step, the Commission believes that quality measurement tools should be evaluated continually and modified to improve the evaluation of plan performance.
    Finally, as the risk program expands, more and more beneficiaries will have the choice of enrolling in a risk plan, and choosing among risk plans. To date, beneficiaries have not had adequate information for making these choices. Information about the risk option furnished by Medicare has been general and provided only to new beneficiaries or those who request it. This year, HCFA will introduce a number of initiatives to improve the information beneficiaries can use to decide whether to join a risk plan. ProPAC believes that all beneficiaries should receive quality and satisfaction data about risk plans as well as the fee-for-service system. In this way, beneficiaries can make informed decisions about which option is better for them.

 Page 51       PREV PAGE       TOP OF DOC

    As Medicare managed care continues to expand, the growth in overall Medicare spending will depend increasingly on the performance of the risk program. This program has the potential to restrain Medicare spending, but only if problems with the payment methodology are addressed. Relying on the current risk adjustment methods and fee-for-service spending distorts risk plan payments. Changes that move towards breaking this link would permit Medicare to fulfill its role as a prudent purchaser of quality health care services for its beneficiaries.
    This concludes my formal statement, Mr. Chairman. I would be pleased to
answer any questions from you or other members of the Subcommittee.


    Chairman THOMAS. Thank you. I appreciate that. We have three panels and we will try to adhere to the lights a bit more than we did in the last Subcommittee hearing.
    Paul, you ended your statement by saying that the President's plan may fall somewhat short in the 5-year period of saving the $34 billion. My assumption is you have not yet honed your instruments enough to tell us how far short?
    Mr. VAN DE WATER. That is correct, Mr. Chairman.
    Chairman THOMAS. We await that information. The second thing the President does is under his argument say that he reduces it by $34 billion, but that enrollment will continue to increase in HMOs. Are you beginning to see that same pattern, which is the old model of losing on every model but making it up in volume? Is that actually going to happen, or can you tell us that yet?
    Mr. VAN DE WATER. As I think you indicated, Mr. Chairman, in our baseline, we assume there will be continued substantial growth in enrollment in Medicare risk plans. The current rate of enrollment in risk plans is about 11 percent, and we project that under current law—that is, without any of the changes contemplated by the President or any other proposals—enrollment in risk plans would reach about 25 percent of the total by 2002 and about 35 percent by 2007.
 Page 52       PREV PAGE       TOP OF DOC
    As I indicated in my statement, our tentative conclusion is that the President's proposal will not have much effect on the rate of growth of HMO enrollment that would otherwise occur. There are things in the President's proposal that would encourage HMO enrollment; there are also items that would tend to make HMO enrollment less attractive. We think, as a working assumption, that the pluses and minuses will roughly cancel out.
    Chairman THOMAS. Thank you. Dr. Wilensky, you talked about part of the President's proposal especially in trying to deal with the AAPCC and the enormous discrepancies that you indicated. It is a little bit like Yogi Berra's ''deja vu all over again'' because we wrestled with the same problem, and interestingly enough, when we finished our conference report, our floor was eventually to be $350. There was an attempt to try to reduce the more than 3-to-1 ratio. Unfortunately for us, today we lost that double digit increase that could have been used to deal with the disparity, but as I look at some of the President's proposals, they do not look a lot unlike what we were talking about.
    If he is approaching it somewhat similar, which I believe to be the case—would you agree—were there advantages to what we did or advantages to what the President is doing because although our goal is the same, reduce the spread, it is done differently, and what, if you have had a chance to analyze it, are the consequences of doing it the way the President is doing it versus the way we talked about doing it in the Balanced Budget Act?
    Ms. WILENSKY. As you indicated, blending rates between the local and the national, and trimming which gets rid of the very low and high values were in the Medicare Preservation Act that was passed. They are also in the President's proposal. So that in those areas, they are similar. There are two areas that are different. A third, removing disproportionate share and graduate medical education, was an issue that has been raised before.
    The area that I am particularly concerned about has to do with the reduction from 95 to 90 percent across the board, which was not dealt with last year. I do think it is possible to introduce some better risk adjustors now, but not one that is nearly so crude. But it exacerbates the other issue, which was handled last year but is not handled this year, and that is not bringing spending rates in an area in fee-for-service back in line with what goes on in capitation. As I indicated, I think there is a real problem that is brewing. Let us use the low spending county as an example. If in the fee-for-service area, in some counties of Nebraska, the spending is $250 per person per month approximately in fee-for-service, but in an HMO that amount will now be $350 because that is where the floor is, you are going to make it very difficult for fee-for-service medicine to continue in those counties. Similarly, in the very high areas, if your choice is being in fee-for-service medicine and having $750 spent on your behalf or being in an HMO and having $650 spent on your behalf, especially if you had slightly better risk adjustors, you are going to push people out of HMOs. So when you start making these adjustments to get rid of the very low and high values, it becomes very important that you bring spending in the fee-for-service area to that same level.
 Page 53       PREV PAGE       TOP OF DOC
    That happened more or less with the so-called fail-safe mechanisms which are now not being contemplated. Having the reduction from 95 to 90 percent is going to exacerbate some of the problems of risk selection, and I worry about that enormously. I think there is a real issue there, but this is not the way to handle it. We can do better. We can start now, and as we get better data, better methodologies, we can improve it. This is going to make the problem worse.
    Chairman THOMAS. My concern is if we wait around for the perfect risk adjustor, we will be waiting around a long, long time, and if we find better tools than we have now, plug them in and as we get better ones——
    Ms. WILENSKY. It is the very strong belief of the Physician Payment Review Commission that we can and should start now improving risk adjustment.
    Chairman THOMAS. Yes. Dr. Newhouse, in your testimony, you say that changes—at the top of the page—changes to the risk payment methodology are necessary to ''ensure the success of the risk program.'' What does ''ensure the success'' mean, and are we supposed to ensure the success? I do not know exactly what you mean by that.
    Mr. NEWHOUSE. I mean——
    Chairman THOMAS. If you are going to ensure the success of the risk program, leave it like it is. In fact, if we are overpaying, it is going to be very healthy. If our goal is to pay as little as possible while providing maximum quality, we may need to look at it. So I just do not understand what you mean when you say ''ensure the success of the risk program.''
    Mr. NEWHOUSE I think the intent was very similar to Dr. Wilensky's notion of neutrality. That if the risk program is unbalanced in either direction, it would not succeed for either the beneficiaries, on the one hand, possibly, or the government as a prudent purchaser, on the other hand.
    Chairman THOMAS. Or we have a teeter-totter.
 Page 54       PREV PAGE       TOP OF DOC
    Mr. NEWHOUSE. So in this sense, as a viable option.
    Chairman THOMAS. Do you believe that currently the teeter-totter tilts in the direction of paying more than we ought to, in a very general sense?
    Mr. NEWHOUSE. Yes.
    Chairman THOMAS. Do you believe the President's plan will tend to tilt us in the direction of not paying enough?
    Mr. NEWHOUSE. Well, my concern is similar to Dr. Wilensky's with the proposal to come down to 90 percent. That is to say that it would move the line, as it were, with respect to who was and who was not a profitable enrollee for the at-risk plan. So a person who was now unprofitable at 95 percent would become even more unprofitable. A person who was marginally profitable would become marginally unprofitable. And everybody who was more profitable, they would become somewhat less profitable. But I would expect, in other words, some short-run savings from this, but I would expect in the long run selection to reemerge.
    Chairman THOMAS. And, of course, the key would be to make the cuts where the cuts are appropriate rather than across the board, and that gets us back to the risk adjustors.
    Mr. NEWHOUSE. Well, the key, I think would be, as Dr. Wilensky said, to get better risk adjustment in place rather than just lop off a fixed percentage for every person.
    Chairman THOMAS. So on the whole, the President's plan may not be bad but the way in which it is applied probably is not the best possible way to apply it. Is that what you are saying?
    Mr. NEWHOUSE. Well, I would have preferred more of an effort to get risk adjustment in place on the capitated payment rather than just reduce the ratio. But, I do believe that in the short run, at least, there will be some short-run savings from it.
 Page 55       PREV PAGE       TOP OF DOC
    Chairman THOMAS. Thank you.
    Mr. NEWHOUSE. Although we will see what CBO thinks.
    Chairman THOMAS. Mr. Stark.
    Mr. NEWHOUSE. The official arbiter.
    Mr. STARK. I would like to ask Dr. Newhouse, Dr. Wilensky, and Dr. Van de Water, I hear everybody suggesting that we need some type of risk adjustment, and that makes sense, but we do not have it right here to do. Could you comment on the GAO's suggestion that we just improve the accuracy of the county payment rates? The suggestion involves recalculating the AAPCC by including HMO costs in the base and not just dealing with the fee-for-service costs. Can you comment on that, Joe? Have you see it before?
    Mr. NEWHOUSE. I have not thought about that. I have not seen the GAO's suggestion.
    Mr. STARK. OK.
    Mr. NEWHOUSE. I think the issue goes to how one would compute HMO costs.
    Mr. STARK. They have a system for doing that.
    Mr. NEWHOUSE. OK. I would have to look at that. In principle, that sounds like it would help. But it still leaves the incentives in place for the HMO to go after profitable people.
    Mr. STARK. Oh, yeah, but that we have to deal with.
    Mr. NEWHOUSE. Right. And that can only be remedied by, as has been said, getting into better risk adjustment. Now I do think there is better risk adjustment technology that is available and could be tried on a larger scale than it is now being tried. The risk adjustment, as I indicated, based on diagnosis, seems to me to be a substantial improvement on what we have. It is not perfect, but it is better and it should help.
 Page 56       PREV PAGE       TOP OF DOC
    Mr. STARK. Gail, did you see the——
    Ms. WILENSKY. I am looking right now at some of what they have suggested, and there are parts of it we agree with very much. That is make——
    Mr. STARK. It is mostly on page——
    Ms. WILENSKY. I am looking on page—at the notion of making an adjustment for new HMO enrollees. That is something we think can and should be done basically as soon as it can be put into legislation. And I think if you use beneficiary survey information along with encounter data and put more pressure to get the no-pay hospital bills from HMOs, that is also something that can be put into effect within 1 year's time.
    Mr. STARK. While these other things——
    Ms. WILENSKY. Exactly. So I think we need to keep working on this, but we could improve from where we are now right away.
    Mr. STARK. I hate to denigrate your testimony, but it is Dr. Van de Water who will tell us whether we can save any money doing that. Paul, how does this strike you at first blush?
    Mr. VAN DE WATER. I do not think I can give you a cost estimate off the top of my head. That would be——
    Mr. STARK. OK.
    Chairman THOMAS. Well, you could, but——
    Mr. VAN DE WATER But that would be very risky.
    Mr. STARK. All right. Thank you. Paul, in your testimony, you suggest that the real solution to the world is going to a defined contribution plan. Let me ask you, because it is a concern I have, if the defined contribution is structured, either because we have balanced budget amendments or whatever, so that it does not keep pace with medical inflation, say in the private sector, what happens? What is the effect then on the beneficiaries?
 Page 57       PREV PAGE       TOP OF DOC
    Mr. VAN DE WATER. I think you are honing in on the central issue here. The crucial question is, How is the growth of that defined contribution established?
    Mr. STARK. But if it does not keep pace, what happens to the beneficiary? What are the alternatives?
    Mr. VAN DE WATER. If it does not keep pace, then beneficiaries will bear an increasing share of the responsibility for paying for their own health care in their old age.
    Mr. STARK. I think that is a way to say increased costs; is it not?
    Mr. VAN DE WATER. Yes. It is a way of saying it increases costs for beneficiaries.
    Mr. STARK. Or they get less care?
    Mr. VAN DE WATER. Yes.
    Mr. STARK. Which would obviously be the less desirable result. OK. Thank you very much.
    Chairman THOMAS. Does the gentleman from New York wish to inquire?
    Mr. HOUGHTON. Dr. Wilensky, everybody, great to see you, thank you very much for being on the panel. I have got to follow up. You described the 95- to 90-percent reduction as a crude and a blunt tool. And let me just run you through a couple of figures. I come from a small county in upstate New York. The difference between that county and let us say the Bronx or Kings County or New York County is huge, far more than the difference in the cost of living. Believe me I have lived in both areas. Now we get 1997 part A and B AAPCC rating of about $358—OK—so now that will go down to 90 percent of that.
    Ms. WILENSKY. Correct.
 Page 58       PREV PAGE       TOP OF DOC
    Mr. HOUGHTON. OK. So that will be about $322, but now the President says you must have at least $350. So that will go back up to $350. So we end up at $8 less in effect than we are getting now, and furthermore we have one of the lowest rates of Medicare risk penetration in the whole State. So if the whole thrust is to move from fee-to-service to managed care, and we already have something like 0.2 percent comparing to an average of about 15 percent in the New York City counties, what incentive is there? What makes sense in terms of bringing us to the promised land?
    Ms. WILENSKY. You clearly have indicated a problem for those counties that are already receiving low payment. They will also lose an additional 5 percent, which may or may not be mitigated because of the floor that would be put in place. I would urge you to think about the thrust of allowing the seniors in your county to be able to choose an HMO plan if they want which is very difficult given the low payment rates that you now have rather than trying to push them in. I think it ought not to be the policy of the government to push seniors either to go into managed care or to stay in fee-for-service.
    But the disparities in payment across the country are astounding. There is a 3 to 1 difference between the payment in the lowest counties and the payment in the highest counties, and they can be brought closer together. Again, when you do that, you need to make sure you do not let the fee-for-service spending in those same areas continue as it did before, or you will either make it impossible for fee-for-service to continue or make it very unreasonable for someone to choose to leave fee-for-service and take such a large reduction in spending if they were to go into an HMO.
    So you need to think of all of these payments of interacting with each other. We have not talked about that issue yet, but the fact is there are many changes that either directly or indirectly will affect either HMOs or particular county areas, including what you do with graduate medical payments and disproportionate share spending, what you do for risk adjustment and what you do with these variations. They all interact with each other, and you have to think about them that way. But your area will not end up any better off even though there is now a floor because you will lose it by going down to 90 percent. And if you had a lot of sick people in an HMO, there is no way the HMO would receive any allowance for that. So you invite risk selection to worsen rather than trying to fix what we know is now a problem.
 Page 59       PREV PAGE       TOP OF DOC
    Mr. HOUGHTON. Can I ask you just one more question? I think we have a different light here. Should I be happy if I am an advocate of a provider service organization that this thing is happening now, that it makes it virtually impossible for an HMO to come into a rural community, and so ultimately the provider service organization is going to be able to be established competing with HMOs and do what we really want to and that is to keep money in the community? Should I be happy about that?
    Ms. WILENSKY. Not in the long run. In the long run, you ought to be happy to have choices available to the seniors, and if as a provider-sponsored organization you can offer a good package of services, all you should need is an even chance to be available to seniors. You do not want to rule out HMOs. It will come back to bite you if you try to shut them out.
    Mr. HOUGHTON. I sure do not want to be bitten. Thank you, Mr. Chairman.
    Chairman THOMAS. That shows you the difference between New York and California. I have some followup questions on the bite question.
    The gentleman from Maryland.
    Mr. CARDIN. Thank you, Mr. Chairman. I notice that none of you really spent too much time talking about the removal of the GME or disproportionate share from the AAPCC rate. Can I assume that your view is that it makes sense to take that payment out and pay it directly to the hospitals that are providing the services to the managed care companies rather than doing it generally through the rate?
    Ms. WILENSKY. Let me share with you PPRC's recommendation on that. Take it out, yes, but be very careful how you give it back. Our concern is that where it has been given in the past may not necessarily be the only places or the best places to give it back. If you take it out, you are forced to deal with the issue of which kinds of institutions or settings ought to be receiving graduate medical education financing—ambulatory care, managed care settings, academic health centers—under certain circumstances or all. So, yes, it makes sense to take it out.
 Page 60       PREV PAGE       TOP OF DOC
    Mr. CARDIN. But we have that problem generally in Medicare as to how we are handling GME costs.
    Ms. WILENSKY. Absolutely.
    Mr. CARDIN. But from the point of view of the payment to managed care companies, you would prefer to see it done directly to teaching hospitals rather than through the general AAPCC rate?
    Ms. WILENSKY. Yes.
    Mr. CARDIN. Is that generally the view of the other panelists?
    Mr. NEWHOUSE. Yes.
    Mr. CARDIN. No view from—Dr. Van de Water, there is one thing.
    Chairman THOMAS. Will the gentleman yield on that point, and I will make sure he has enough time? We are focusing on that which is built into the AAPCC with the argument that it is not passed through. You are talking about a cleaner mechanism, but on the fee-for-service side, as we continue to change delivery methods away from hospitals and into stand-alone clinics and even into outpatient structures, though certainly not to as great an extent, do you not have a problem on the fee-for-service side? I want to make sure the comments we are directing about the inability to get payments for graduate medical education certainly are more on the HMO side, but permeate the entire structure, and that what we offered in terms of a different type of funding, and I know you have been interested in all payers, makes sense not just on HMOs but on fee-for-service. Is that a fair statement? Everyone agree with that?
    Mr. CARDIN. I agree with you.
    Ms. WILENSKY. Yes.
    Mr. NEWHOUSE. Yes.
    Chairman THOMAS. OK. Thank the gentleman.
 Page 61       PREV PAGE       TOP OF DOC
    Mr. CARDIN. I think the Chairman makes a very good point. It is certainly a problem that goes beyond just the AAPCC rates that are paid to managed care companies. Dr. Van de Water, the point I was raising is that all of us are concerned about the risk issues of the enrollees in the HMOs. We know that individuals that tend to use more health care services tend to feel more comfortable in a fee-for-service environment. Therefore, despite the fact that we have some risk adjustors currently in the rates, the HMOs are taking a healthier population. You make an observation in your written statement that one of the President's proposals—which is very popular, that I personally hope will become law—that guarantees Medigap coverage on disenrollment and would allow people to come back into Medigap without being rated could, in fact, encourage HMOs to unload some of their higher risks, creating the same problem again.
    I am curious as to how serious a concern this could be. If we just did that, if we just changed the Medigap laws to allow people who drop out of their HMOs to get back into their Medigap plans, would we be exaggerating the problem we currently have with HMOs being overcompensated for the risks that they have?
    Mr. VAN DE WATER. I do not think anyone can be sure about that, and I for one would welcome Dr. Newhouse's and Dr. Wilensky's views on this. My guess is that, in the short run at least, the administration's proposal probably would tend to improve the situation since it would initially make HMOs more attractive to a wider range of people, because they would be less concerned about being locked into an HMO. The ease with which sicker beneficiaries could disenroll might not begin to cause problems until a bit later, but that is just speculation.
    Mr. CARDIN. But if a person enrolls who is healthy, wanting to protect their Medigap coverage if they get sick and want to come out of the HMO, it seems to me that you are encouraging more of the healthier people to leave. I could be wrong.
    Dr. Wilensky.
    Ms. WILENSKY. There is a second important issue the administration has raised, and that is annual enrollment. I think you probably would not in the end exacerbate the problem with the protection to go back into fee-for-service because of the annual enrollment. Right now you have a 30-day disenrollment. That strongly encourages individuals who think they may have some high expenditures who would like to be in an open-ended fee-for-service world to do so. If under current law you also gave the protection of Medigap insurance, you would exacerbate the situation even more than it currently stands. If you have annual enrollment but give protection at the end of the year that you can go back into a Medigap plan, I think on balance you might actually end up being better off from where you are now. So I think that is an extremely important provision, the annual enrollment.
 Page 62       PREV PAGE       TOP OF DOC
    Mr. VAN DE WATER. Mr. Cardin, if I may add one point? This is an area in which we are not entirely sure what the administration's proposal is. It is our staff's understanding that the annual open enrollment period would not replace the beneficiary's ability to disenroll in the next month.
    Ms. WILENSKY. Well, then strike my comment about having this mitigate the factor.
    Mr. CARDIN. I think this is something we are going to need to have more dialog on to make sure we do not just create a——
    Mr. VAN DE WATER. But certainly Dr. Wilensky is correct that if the administration's proposal were to substitute the annual open enrollment period for the existing disenrollment provisions, that would ameliorate the problem that you are concerned about.
    Ms. WILENSKY. Thank you.
    Mr. CARDIN. I think—Dr. Newhouse, did you want to comment?
    Mr. NEWHOUSE. I am, I think, somewhere between the two although I certainly agree on the annual enrollment. You might want to couple that with, say, a 3-month probationary period where you could disenroll in the first 3 months as a consumer protection measure.
    Rather than the image of the plan dumping the sick person out, I would have said it is clearly easier for a sick person to decide to disenroll if they can get their Medigap back at favorable terms than if they cannot. So I would expect some effect along those lines, but I think it is very uncertain just how big it would be.
    Mr. CARDIN. I think I have used my time and the Chairman's time.
    Thank you, Mr. Chairman.
    Chairman THOMAS. Thank you. And I want to try to keep some semblance of continuity on the questioning, and so periodically there may be a slight adjustment here. But did not the GAO report show that when people disenrolled, there was a 163-percent increase in the fee-for-service usage.
 Page 63       PREV PAGE       TOP OF DOC
    Ms. WILENSKY. Yes, that is really the point.
    Chairman THOMAS. And if you use that example, then it is absolutely critical we do not have, if you open up Medigap for coming back in on a monthly enrollment, you can have an enormous possibility for adverse risk selection, and so the length of time you must stay after, for example, as we did the last time, a window of opportunity for adjustment and then a length of stay, makes all the difference in the world——
    Ms. WILENSKY. You are correct.
    Chairman THOMAS [continuing]. In the cost payments. And so there was information in the GAO study that indicated that.
    Mrs. Johnson.
    Mrs. JOHNSON. I am interested in your support of carving out GME, IME, and DSH payments. Are you confident that that would leave a reimbursement that was sufficient to enable Medicare to compensate hospitals as well as managed care plans do through the private sector, because in New York when they carved out Medicaid payments, they were unable to. What was left did not allow the managed care plans to compensate the hospital, the teaching hospitals, at the same rate they were compensating other hospitals, and they actually had to increase their premiums to offset some of the loss. So while it appeared to save the government 12 percent in the drop in rates, actually when you added in the premium increase, it did not save all that. It cost shifted some of that across the board.
    So I personally fail to see how we will be able to compensate teaching hospitals well enough if we take out all of these payments to enable them to provide the needed care, because you are sending much more severely ill patients to the hospital, and while we understand DME and IME, I thought you guys were going to come up this year with some better understanding of DSH payments, what portion of DSH payments really relates to educational costs, and what portion really relates to uncompensated care. And until we understand that, taking those three out seems to me high risk for educational institutions.
 Page 64       PREV PAGE       TOP OF DOC
    Mr. NEWHOUSE. Mrs. Johnson, I think I will be testifying subsequently about our recommendations on both teaching and DSH payments, but our intent was that these moneys would be going back in some fashion. That it was not money taken back to the Treasury but that there would be a separate flow back to teaching institutions and institutions treating high numbers of low-income patients. Our concern here is that these moneys were put in for a special purpose, and now that they are flowing to managed care companies, there is no longer a fixed mechanism for assuring that these moneys achieve what they are trying to achieve. So that some other mechanism would have to be found to channel the money than going through the AAPCC and the managed care funds.
    Mrs. JOHNSON. Well, I appreciate the intent. I appreciate the theory. The fact is in practice that the art of rate setting has had some arbitrary components in it over the years, and so its accuracy across the country and its adequacy across the country varies. And I think before we take those three out, we have to be sure that what is left is equivalent to what HMOs are paying to other hospitals, and if we cannot assure ourselves of that, we are going to see happen across the country in teaching hospitals what happened in New York with teaching hospitals when they took this approach to Medicaid. That is what worries me.
    Then the second thing that really concerns me greatly is if we do this, and we go to a system, and we retain the system of capitated payments for HMOs, we absolutely have to have some risk adjustment. Now there are fewer hospital beds occupied for a shorter length of time with much more severely ill people. The lack of an outlier adjustment in Medicare or a realistic outlier adjustment is one of the things that is driving the debt of very small hospitals and was one of the key factors in our losing a small hospital in my district. It was a key factor, just one outlier, but a very big cost outlier. So unless you do something about outlier costs, capitation and the lowest level of reimbursement are going to combine to really underreimburse our hospitals, and I think that would be a very serious problem.
 Page 65       PREV PAGE       TOP OF DOC
    The other thing I wanted to ask Mr. Van de Water is if the AAPCC remains tied to the fee-for-service expenditures, it seems logical that all reductions in fee-for-service payments to providers will be reflected in the AAPCC. Now if we reduce the cost for providers, it should be reflected in the AAPCC, and if that is true, are you reflecting those costs savings in your estimates in regard to managed care savings because they should be the same?
    Mr. VAN DE WATER. Yes, they are, Mrs. Johnson.
    Mrs. JOHNSON. And they are reflected?
    Mr. VAN DE WATER. Yes.
    Mrs. JOHNSON. Thank you. My time is expired. Thank you.
    Chairman THOMAS. Mr. Becerra, do you wish to inquire?
    Mr. BECERRA. Yes. Thank you, Mr. Chairman. Thank you very much for being here this morning. If I could follow up on some of the questions having to do with the whole issue of annual enrollment and the 3-month probationary enrollment. In that vein, I am interested in talking more about the whole issue of providing more information to the consumer of the health services, and the whole idea of providing that consumer with the opportunity to make a decision, whether it is through the 3-month probationary period, an annual enrollment, or perhaps by providing on a periodic basis, annually, biyearly, or whatever the case may be, information to the consumer of the disenrollment rates for that particular provider.
    Comments from the panel on whether we should be providing that type of additional information to the consumer of health care services?
    Ms. WILENSKY. Yes.
    Mr. NEWHOUSE. Yes.
    Ms. WILENSKY. Absolutely. I agree with the notion of especially allowing first-time enrollees some period in which to disenroll. But if you are to help seniors choose the plan that best suits their needs, they need more information. They need more information about the Medigap plans in their area. They need more information about the capitated plans, including consumer satisfaction information and information on process and outcomes. HCFA is going to start requiring the so-called HEDIS 3.0 information be provided to HCFA. This ought to be distilled and provided back to consumers so that they can make wise choices including disenrollment information. It is a very important piece of consumers being able to vote with their feet.
 Page 66       PREV PAGE       TOP OF DOC
    Chairman THOMAS. Would the gentleman from California yield to me at this point? It will not come out of his time, the Chairman assures.
    Mr. BECERRA. That is fine.
    Mr. STARK. Just to follow up with Dr. Wilensky. I am concerned, and I think you have mentioned this, that when we sign up a lot of people, either in Medicare or Medicaid HMOs, we start paying them way before they ever see the people. I have suggested in one bill that perhaps there ought to be some requirement that before we start sending money, however much—$300, $600—a month, that there be a requirement that HMOs engage with this new patient, either a face-to-face interview or something other than just getting a long list and sending HMOs money. Is that something you have been concerned about?
    Ms. WILENSKY. I actually have not thought specifically about it. It is an interesting issue to have some kind of an enrollment, physical or contact, so that you know, the patient knows.
    Mr. STARK. Even the recognition that the enrollees know where to go—that they have been in contact and the HMO knows how much they weigh or what their blood pressure is.
    Thank you.
    Ms. WILENSKY. I think that is interesting and worth exploring.
    Mr. BECERRA. Dr. Newhouse, I think you wanted to respond to the question as well.
    Mr. NEWHOUSE. Yes. I just wanted to say that although our focus is Medicare, that putting information out there would help performance I think in the under-65 market as well.
    Mr. BECERRA. Yes. Responding to what Mr. Stark was asking, it seems to me that although it would be a good idea to have some type of activating event occur so that we can start providing the provider with money, it seems they would probably just initiate something very pro forma which would have them complete that particular task. I am not sure if we ultimately achieve what we wish unless we have very informed consumers who thereafter follow up from that initial episode and follow up with some type of active service or request for service.
 Page 67       PREV PAGE       TOP OF DOC
    Let me ask another question. It seems to me that, and I do not know how we can change this, but it seems to me we make providers, and in this case I am talking about insurance companies as well, as the gatekeeper not only for the services but also with the money. So they are holding everything, and they provide it as they see fit. It seems to me if we are able to make providers the gatekeepers of the services and the consumers the gatekeepers of the money, we would probably have a happier marriage between the service and the cost. How can we go further in trying to make sure that not only do we inform consumers, but we make them the actual gatekeepers of the money that will be paid for the service?
    Ms. WILENSKY. There are two different ways you can do it. One is that if you think about the Federal Employees Health Benefits Plan as a model, you at least give consumers a wide range set of choices with information so that they can choose which plan among many meets their needs. During November in Washington the airwaves are filled with advertisements, notices, and reminders to Federal employees. It is time to think about whether your plan is meeting your needs.
    The other option is one, of course, that was again raised last year and that is part of the demonstration under the Kassenbaum-Kennedy bill, and that is the notion of a medical savings account where you put a high deductible with a set-aside amount of money. That is the most direct way to give consumers control over money, and that is obviously an issue that has been raised and probably will be raised again.
    Mr. BECERRA. Dr. Wilensky, now to have an MSA, you either have to have the money in your pocket or the government has to give you the money so you can put it in your pocket to put in an MSA.
    Ms. WILENSKY. Correct. I was thinking more though in the case of Medicare, if you wanted to give direct control of moneys to consumers, aside from giving them a wide range of choices, the other is to have part of the Medicare moneys, again risk adjusted, be available and in an account for them to access.
 Page 68       PREV PAGE       TOP OF DOC
    Mr. BECERRA. Mr. Chairman, if I can be indulged for one last question. And returning to the whole issue of GME, DSH funds, if we are going toward a system that gives us a better read on how much we spend per Medicare enrollee so that we do not overcompensate a managed care provider, and if we are having some calculation that is giving us this ultimate amount of money that we are going to give to the providers, now better adjusted to reflect the costs, if we are going to have a carve-out, it seems to me we are going to have a smaller carve-out because the formula that is being used to ultimately trim the piece of the pie is smaller. That means we are probably going to have less money for teaching hospitals, less money for hospitals that take care of the poor, if we are just doing a simple carve-out.
    Do we not face the risk that we are going to jeopardize those institutions that are teaching institutions and those that are providing care to the indigent by saying we are going to rely on a carve-out that may actually become a shrinking part of the pie?
    Mr. NEWHOUSE. Well, how big this pot is is up to you as it is now. You have decided, for example, that you will pay around 7.7 percent extra for every 10 percent increase in the resident-to-bed ratio for indirect medical education, and that is what determines the size of the money. So that is entirely within your control.
    Mr. BECERRA. What if you removed the whole idea of funding DSH, IME, and GME, removed the whole idea of the carve-out and rather than that had a system where each provider or each institution would be obligated to provide some form of funding for a pot rather than rely on just those particular providers having a particular pot carved out for that particular purpose?
    Mr. NEWHOUSE. Well, I suspect the providers would say they do not have any magic money, that the money has to come from somewhere, and they would look to you for that money.
    Ms. WILENSKY. The carve-out really forces you to think more explicitly about who should receive the money. My only caution about taking it out of the AAPCC is now you must be prepared to say here is where the money should go. If you are not ready to do that, you may want to consider leaving it in. It is 5 percent of the total spending, but it is an issue that is very serious. The Congress will ultimately need to deal with how much and to whom should this money go.
 Page 69       PREV PAGE       TOP OF DOC
    Mr. BECERRA. Thank you. Thank you, Mr. Chairman.
    Chairman THOMAS. Dr. Wilensky just mentioned the 5 percent. I think in 1995, ProPAC indicated they thought the combined graduate medical expense out of both part A and part B was 5.3 percent. So that is generally accurate. Dr. Van de Water, do you believe that is currently an accurate estimate? Has it gone up or gone down? Or do you know?
    Mr. VAN DE WATER. We have no independent information. That is the figure we are using.
    Chairman THOMAS. And, therefore, you would not know a variation from county to county either. So it is about 5.3 percent roughly now; is that what we are saying?
    Ms. WILENSKY. Right. It varies from zero to 20, but it is about 5.3 percent.
    Mr. NEWHOUSE. Dr. Young has just given me a chart by county for the large counties from 1995 data, which we would be happy to share with you. I think your staff may have it.
    Chairman THOMAS. OK. Thank you. Does the gentleman from Texas, Mr. Johnson, wish to inquire?
    Mr. JOHNSON. Thank you, Mr. Chairman. Gail, I wonder if you could talk about the fact that one of the studies indicates there are only about 2.8 percent of the people that have disenrolled from HMOs and returned to fee-for-service. Can you tell me is that an accurate number, and do we really have to worry about disenrollment if it is that low?
    Ms. WILENSKY. It is a very small number in terms of the people who go back to fee-for-service. It is about 3 percent. They do tend to be sicker. So the answer is it is small, but they are not the typical enrollee. What you need to worry about is that averages can get misleading, and it could be a bigger problem for some places. But it is a much smaller number than most people think. It is a very small number of the people who go back to fee-for-service.
 Page 70       PREV PAGE       TOP OF DOC
    Mr. JOHNSON. But are you saying those are the really sick people or not?
    Ms. WILENSKY. There are sick people who stay, but for the small number who leave, they definitely tend to be high users.
    Mr. JOHNSON. OK. So the costs go up when they go back to fee-for-service; is that true?
    Ms. WILENSKY. Yes. And it is one of the reasons that some of the Medigap insurance companies are nervous about the provisions of being forced to take these people who disenroll back especially if it is at a community rate. They do tend to have high use. Our estimate is that these disenrollees spend 160 percent of the average expenditure, so they are substantially sicker or heavier users for whatever reason of health care.
    Mr. JOHNSON. Thank you. Dr. Newhouse, in your testimony, you talk about the administrative costs and profits accounting for about $66 per month of risk plans. This is higher, a lot higher, than your estimate of $20 a month to cover administrative costs. And, in fact, makes itself an annual difference of around 552. Can you tell me why this occurs? Is it because Medicare has to enroll or recruit enrollees—and maybe you could quantify the risk difference there?
    Mr. NEWHOUSE. Yes. If the issue goes to how these costs are accounted for and then reimbursed by Medicare, this is an example that illustrates the potential issue in trying to do the accounting.
    Mr. JOHNSON. Well, but are our administrative costs going up, I guess, is the real question?
    Mr. NEWHOUSE. I do not have any time series data, Mr. Johnson, or any data over time on it. The issue here is you can take your commercial data, and it may be different than your Medicare data, your Medicare true numbers. We do not really know very much about that. It is just pointing to a possible problem rather than trying to get a firm handle on how big the problem is.
 Page 71       PREV PAGE       TOP OF DOC
    Mr. JOHNSON. Well, do you mean we are unable to survey it or the companies cannot tell us?
    Mr. NEWHOUSE. Well, there is an inherently arbitrary nature you have. Let us say you have a marketing department in your HMO. How you decide how much of those costs of the marketing director should go to Medicare and how much should go to commercial business is somewhat an arbitrary decision.
    Mr. JOHNSON. Yeah. They probably would like to put it all on the Medicare, would they not?
    Mr. NEWHOUSE. Well, if you will pay it, I am sure they will come.
    Mr. JOHNSON. Thank you, sir. Thank you, Mr. Chairman.
    Chairman THOMAS. Thank you, Mr. Johnson. The gentleman from Wisconsin, Mr. Kleczka.
    Mr. KLECZKA. Thank you, Mr. Chairman. I have one quick question. I think I will direct it to Dr. Van de Water. The administration proposes a 5-percent reduction in the overall reimbursement. We are going to hear from panels this morning made up of HMO providers indicating that if, in fact, that would be adopted, not only will they have to curtail the additional benefits that are being offered but also the likelihood of charging additional out-of-pockets would probably exist. What is your reaction to that after looking at the administration proposal?
    Mr. VAN DE WATER. As I indicated in my summary, Mr. Kleczka, the administration has some proposals that would encourage enrollment in HMOs, and other proposals that would discourage HMO enrollment. Taken by itself, the proposal to reduce payments from 95 to 90 percent certainly falls in the discouraging category.
    Mr. KLECZKA. OK. Do you have any recommendations on options from going that route?
 Page 72       PREV PAGE       TOP OF DOC
    Mr. VAN DE WATER. No, the recommendations we leave to Drs. Wilensky and Newhouse.
    Mr. KLECZKA. Gail.
    Ms. WILENSKY. We think there are better ways to do risk adjustment. There is indication there are overpayments in general, but reducing the payment from 95 to 90 percent will exacerbate that. We need to start making better adjustments using encounter data, taking advantage of the fact that new enrollees appear to be healthier, making lower payments for new enrollees. The President's proposal is a very crude, blunt instrument. We can do better and should do better. Some HMOs are predominantly made up of individuals who have been there for 6 or more years. That is a very different HMO than one where 80 or 90 percent are new enrollees.
    Mr. KLECZKA. Thank you very much. Thank you, Mr. Chairman.
    Chairman THOMAS. Thank you. I will let folks know when I ask loaded questions so this is a loaded question. And the only appropriate person to direct it to, I guess, is Dr. Wilensky. On February 13, we had the administrator of HCFA in front of us, and we asked for, since they obviously had this plan which reduced HMOs by the various structures we have been talking about, to indicate to us what the impact was county by county. He advised us that they had not done a thorough county by county analysis, but assured us that by this last Friday we would be provided with the materials so that we could have a better discussion on this particular panel.
    As you might guess, we have not been provided with the materials. It is our understanding that as they actually ran it county by county, they came up with some problems that they had not anticipated. During your tenure, would you have ever made a multimillion dollar decision without a detailed impact analysis under the structure county by county?
    Ms. WILENSKY. I cannot imagine that we would have.
 Page 73       PREV PAGE       TOP OF DOC
    Chairman THOMAS. Is that a yes? That is a no?
    Ms. WILENSKY. We would not have—no, we would not have made such an analysis. I am surprised the information was not immediately available in the sense that that is how you analyze the information of lowering the high payment rates and raising the low payment rates, and when you are doing as much as taking $34 billion out of one sector, looking at how it will be impacted on a county-by-county basis is a first order question.
    Chairman THOMAS. It just amazed me, but we are looking forward to the corrections, and, in fact, our belief is that it will lead to policy adjustments in the larger picture based upon what they have now discovered to be reality at the county level. And my concern, as you might imagine, is a degree of concern over the entire package if, in fact, that is the type of analysis that they have used to put it together. Thank you for participating in that needed discussion of a loaded question.
    Basically, this will be based on Dr. Newhouse's testimony, but obviously anyone who wants to respond or follow up can. The gentleman from California, Mr. Stark, made some comments, and I am sure he will get into a discussion with the specific HMO plans, and, in fact, I believe he has introduced a bill dealing with profit and not-for-profit issues here. My concern is in your testimony, you say a ProPAC analysis estimated that in 1995 the average risk plan provided each enrollee with $43 in extra benefits each month. That is part of the cost savings, and HCFA requires them either to return it to HCFA or provide extra benefits, and the average was $43.
    My question is, Have you ever broken down the return in benefits, and that is, of course, the average, and it will be different in different areas, have you ever broken it down in terms of for-profit versus not-for-profit benefits that have been returned?
    Mr. NEWHOUSE. Yes.
    Chairman THOMAS. Oh, you have.
    Mr. NEWHOUSE. Yes.
 Page 74       PREV PAGE       TOP OF DOC
    Chairman THOMAS. Oh, good.
    Mr. NEWHOUSE. These are plan weighted averages.
    Chairman THOMAS. I understand.
    Mr. NEWHOUSE. Not member weighted or enrollment weighted, but the for-profits provide $48 per member per month, and the not-for-profits provide $32.
    Chairman THOMAS. So with an average of $43, the for-profits are above the average and the not-for-profits are below the average?
    Mr. NEWHOUSE. Yes.
    Chairman THOMAS. OK.
    Mr. NEWHOUSE. This does not control where the for-profits and not-for-profits are. Although we have not done that analysis, it could be the for-profits are disproportionately in the high AAPCC area. So if you saw in my last chart, the amount of benefits is higher there but the numbers are 48 and 32.
    Chairman THOMAS. I think you will find that may be one of the reasons HCFA has not provided us with the county by county. When you begin breaking it down, it gets very complex with a number of factors impinging upon how you deal with the amount and who gets what, not just in terms of geographic area but in a competition area and a number of other factors.
    Mr. NEWHOUSE. You are exactly right, Mr. Chairman.
    Chairman THOMAS. Yes. In your testimony, when you talk about accounting for Veterans Administration and DOD services, we currently are entering into negotiations in which there will be a subvention program with the Department of Defense in their TRICARE Program, and the Veterans Administration is interested in moving forward although I believe their so-called VISN Plan is perhaps not as mature, and it, of course, involves a larger number than the military retirees. And I just want to begin exploring this area because it seems to me the adjustments that are being made in VA and DOD are primarily because there is no World War II veteran today not eligible for Medicare, and the problems these veterans and military retirees are facing are geriatric problems, are aging problems, not necessarily related to military service. Do you believe we should be looking at the merging, consolidation, and/or appropriate removal of duplication between the VA, the DOD and the Medicare system?
 Page 75       PREV PAGE       TOP OF DOC
    Mr. NEWHOUSE. Well, I have to be in favor of removing duplication, but let me be clear about the concern here. The concern here was that the AAPCC as you know is based on the average fee-for-service use in the county. Now if there is a big VA or DOD facility, that will not be counted in the fee-for-service use, and, therefore, relative to a county where there is not such a facility, the rates may be depressed. That is the first order effect we were trying to get at.
    Chairman THOMAS. I understand. I am trying to jump you to the larger question.
    Mr. NEWHOUSE. Right.
    Chairman THOMAS. Notwithstanding the need to adjust for those facilities being available and affecting the fee-for-service, why not look at the larger fundamental question?
    Mr. NEWHOUSE. Correct.
    Chairman THOMAS. That the services they are utilizing at the VA and the DOD tend to be more related to age than to service.
    Mr. NEWHOUSE. Right.
    Chairman THOMAS. Do you believe that to be the case?
    Mr. NEWHOUSE. Oh, I have no data but let us suppose it is true. I do not know if it is true. Could be.
    Chairman THOMAS. Anyone else have any information on that? This is an area I think we need to begin to explore.
    Mr. STARK. Mr. Chairman.
    Chairman THOMAS. Yes.
    Mr. STARK. Are we not paying twice, in a sense, with these? In other words, if we spend Medicare dollars at the local VA hospital or for the Department of Defense, and we are also subsidizing the medical school for the Department of Defense and building the hospitals out of some other pocket than this Subcommittee's, but as opposed to those same people in CHAMPUS going to the local hospital who needs the——
 Page 76       PREV PAGE       TOP OF DOC
    Chairman THOMAS. And there are those who will argue that the Department of Defense is delivering services that would be more rightly paid under Medicare and that, in fact, they are not getting reimbursed for it, and that is why we are looking at trying to integrate the whole structure.
    Dr. Wilensky.
    Ms. WILENSKY. Mr. Thomas, in terms of your previous question about who is using the facilities, an analysis that I had done some time ago indicated that the users of the VA are primarily those who are uninsured, with very low incomes, or have specific service-related disabilities that are best treated at particular VAs. But, in general, seniors or other individuals who have alternative coverage—seniors through Medicare and under 65 through private insurance—tend not to use the VA system and use either their private insurance or Medicare except if they have specific service-related disabilities that are best treated by the VA.
    Chairman THOMAS. Thank you. Dr. Newhouse, in your testimony you indicate the fee-for-service system is fundamentally different than managed care. I think that statement goes unchallenged, and that is why we are looking at trying to break away the payment structure to managed care from fee-for-service. Is it so fundamentally different that it does not make a lot of sense to take the benefit package of the fee-for-service side and require exact duplication within a managed care structure and then allow for add-ons that may or may not be justified by virtue of the money, or is it worth our while to examine the fundamental question of what the actuarial equivalent would be for the delivery of those benefits under the fee-for-service for managed care, that is maximizing the opportunity for the fundamentally different managed care structure to deliver product for price?
    Mr. NEWHOUSE. I would think in the spirit of trying to be neutral, one would want to try to have the same stated benefit package. The only point I would make is that it is very hard to define what is a truly equivalent benefit package in a managed care program. In the fee-for-service program, the benefit is put in the package, and it is assumed that the beneficiaries and providers will just use it. On the managed care side, the managed care plan may have a different philosophy of how to treat a given illness. And may treat it quite differently so that how people think about a benefit package may not be sufficient to ensure equality in practice as opposed to on paper.
 Page 77       PREV PAGE       TOP OF DOC
    Chairman THOMAS. Anyone else wish to react to that?
    Ms. WILENSKY. Let me continue the thought that Dr. Newhouse was raising. That is to some extent going to an actuarial notion would require very clear information to the senior when they are making a choice. That is, it may be very difficult to literally think about having the same benefit packages when you talk about fee-for-service and network plans because of their different orientation. You need to be very clear about the benefits that are available and how they are accessed and any restrictions that might exist in terms of seeing a specialist or accessing new technology and satisfaction. If this information were available along with clear measures of actuarial equivalents, you actually may be able to provide better information to a senior than by trying to keep with the literal same benefit definition that now exists, which may not be a realistic definition.
    Mr. VAN DE WATER. If I could just carry that one step further still. HMOs by design are generally structured so as to have low copayments and deductibles. Most people would probably not suggest you try to force HMOs to move back in the direction of fee-for-service with its large cost-sharing amounts. A second complexity is that under the current pricing system, even though most beneficiaries have some gap coverage—either through private insurance or Medicaid—and that gap coverage induces further demand for Medicare services, the induced demand for the basic services is not reflected in the price of the gap coverage but is included in the cost of basic Medicare. So how to make all this stuff comparable is indeed challenging.
    I should note one of the administration's proposals is to help standardize the benefit packages offered by HMOs and to make the choice process easier in that regard. That might possibly be a step forward.
    Chairman THOMAS. Along those lines, Dr. Newhouse, you talk about the various changes and concerns on adjustments in risk payments, where various approaches would ''break the link to fee-for-service spending and permit Medicare to take advantage of many of the same forces private sector purchasers have successfully relied on to reduce their health care costs.'' And I think we have got a conceptual problem there because one of the reasons I believe the private sector has moved more into the managed care concept was because they were comparing what they were getting for price against the old fee-for-service, and it was relatively a free market for determining that price.
 Page 78       PREV PAGE       TOP OF DOC
    When you move that concept into Medicare, you have such a heavily subsidized structure that, in fact, we may, as we have been hearing, be paying more for the managed care than for the fee-for-service, but no one talks about the amount of subsidy that we are giving to folk. Paul just mentioned the problem of the Medigap insurance not reflecting an actual price, let alone the 75 cents on every dollar being subsidized to everyone including millionaires on the part B side. So how can you get a real feel for the cost savings of a different kind of a structure if you do not have the people who would be making the choice; i.e., a managed care versus the fee-for-service, inside the system appreciate the increased costs? As long as we buffer that, as long as we subsidize it, people are not going to see the true relative costs of the two programs.
    And what I am concerned about is we are going to ratchet down heavily the HMO side and assume we will not have a negative effect on the number of people in enrolling, and my only argument would be that if we reflected the true costs a little bit more; i.e., premium changes or other changes, removing the buffer, so that people could feel the actual costs on the other side, you would get movement across from a fee-for-service to a managed care. But my concern is that as long as we have this artificial lid on what people actually have to pay versus the services they are getting, you cannot get this comparison that you present in your testimony as an example of how we have made savings in managed care in the private sector carrying it over to Medicare. Is that a fair statement?
    Mr. NEWHOUSE. Yes, I think that is a fair statement. We do reflect some of that back in the extra benefits that are provided, but that is not quite the same as you are speaking about, so, yes, I agree.
    Chairman THOMAS. I was talking about a far larger assessment of the actual costs represented by the part B subsidy, by the Medigap in essence subsidy, and then the mix of benefits, some of which are the buydown on the premium or otherwise costly provisions like paid prescriptions or the rest. Clearly, if we standardized the benefit package, and then government said first you add A, then you add B, then you add C, if you have extra, you would be able to make a more realistic comparison, but I think you would then really have to make sure that the payment structure conforms to a true marketplace because you are not allowing these folks to offer a competitive package in a particular area which may be different than a competitive package in another area, and that concerns me a lot.
 Page 79       PREV PAGE       TOP OF DOC
    Mr. NEWHOUSE. Yes, I agree, Mr. Chairman.
    Chairman THOMAS. One last question before we move on. Because it is technical, I want to go through it very carefully because as we are trying to understand what the President has proposed, there is a relationship within his proposals that I think I am clear on now, but I want your, Paul, and Gail, if you can, reaction to what I am talking about. When you look at the President's proposal to reduce the direct, indirect, and disproportionate share from the AAPCC, you would think you would look at where the high teaching costs are there would be a greater reduction, and that where there are low teaching costs, you would reflect a lower reduction in the AAPCC adjustment. That is it would reflect since it is county by county the costs in that structure.
    My understanding is that that is not the case. That the President's proposal actually removes those teaching payment items from the larger aggregate payment, not looking at the particular use profile in a given area. I'll give you an example so that everyone understands what I am talking about. In 1995 the monthly payment rate in Kings County, New York, was $647. I understand you estimate that $124 of this payment in this particular area, 19.2 percent, was due to direct, indirect, and disproportionate share. That is in comparison to the figure we mentioned earlier of 5.3 on the average. So clearly it is a much higher cost for teaching in that particular county.
    Now, the President's proposal would not take that $647 AAPCC payment and subtract the $124 from it; is that correct? They do not take it directly out of the cost for the county. Instead, with the hold-harmless provision in the President's plan, Kings County would get a zero update for 1998 and 1999. So the cost of removing those payments comes from the aggregate which means relatively those at the lower end who get the blended rate would get less than they otherwise would get. Is that accurate?
    Ms. WILENSKY. It is the hold-harmless, and, yes, once again the interactive effects are raising themselves.
 Page 80       PREV PAGE       TOP OF DOC
    Chairman THOMAS. Now for a minute or two, What is the advantage or the disadvantage of taking it from the aggregate rather than from those high cost areas where there is a relative benefit by leaving it into the basic structure?
    Mr. VAN DE WATER. Mr. Chairman, if I could say another word about the previous question before we get into the pros and cons. As Dr. Wilensky said, this is an illustration——
    Chairman THOMAS. I can see my time has not expired so you can go ahead.
    Mr. VAN DE WATER. As you correctly said, there are interactions between these proposals. Figuring out the interactions is, as we understand it, one of the reasons the administration has not been able to come up with those county-by-county numbers that we are all seeking. But, as you said, there is a tendency for the reduction—to the extent that the reduction in the payment is due to pulling out the teaching and DSH payments—to be undone by the hold-harmless provision. We are not sure to what extent it is undone, but it will surely happen in some cases, as you cited. It may not happen in others because of the phase-in the administration has. The administration is not taking out the payments all at once. Their initial proposal was to take out 40 percent of the payments the first year and 60 percent the second. Our understanding is they are now considering doing it 50–50 or in some other way, but the phasing out of these payments has to be tied in with the workings of the hold-harmless provision.
    Chairman THOMAS. But the fundamental point I am making, I believe, has not changed, and that is it would not be a subtraction where these costs are incurred, but rather a burden across the board where people would get relatively less than they would otherwise get because it is taken out of the aggregate and not the county-specific.
    Mr. VAN DE WATER. The answer is that it would in some cases but not in others.
 Page 81       PREV PAGE       TOP OF DOC
    Ms. WILENSKY. The other issue you need to think about is whether you are going to unlink the payments from fee-for-service. If you were to unlink the payments for HMOs from what goes on in fee-for-service, you would have now what is now 5 percent as an overall average payment become a smaller and smaller piece of the whole for the HMO payment because over time it would grow based on some other factors. So whether or not this is a serious issue is not so much where you are now but where again there are not very many seniors in HMOs, it will become increasingly more serious over time.
    If you unlink HMO payments from what goes on fee-for-service, it is not a very serious problem in most parts of the country although there are places in New York where it represents a lot of payment. If you continue this linkage over time with fee-for-service and HMO growth goes as CBO projects it will, it will become an extremely important issue. So if you are going to take it out, it is much more important if you are going to do that and keep the linkage there, but if you take the payment out, you are going to have to decide what to do with that money very explicitly, and it will not necessarily come from the places where the spending is high because of the hold-harmless.
    Chairman THOMAS. Is that it? Thank you very much. The gentleman from California.
    Mr. STARK. Joe made some statements about who spends what on extra benefits, not on basic benefits. But my question is mostly concerned with the outliers. The overheads are in a range of maybe 10 percent from 95 or 90 to 85 or 80. What would be the harmful effect—I can think of what the benefits would be—of establishing a loss ratio? Eighty percent or eighty-five percent of everything we pay you has to be spent in medical care. Anything wrong with that?
    Mr. NEWHOUSE. I think the problem may be in actually trying to enforce it. That is——
 Page 82       PREV PAGE       TOP OF DOC
    Mr. STARK. Pretty easy.
    Mr. NEWHOUSE [continuing]. It is cleaner in a fee-for-service plan where there is a clear break between the insurer and the provider. For example, if I am in an at-risk plan, and I have a medical director on staff who negotiates some of my rates with my physicians but also provides some services, how do I allocate that person's salary?
    Mr. STARK. The same way you would the person sitting at the 800 number who does the same work the medical director does in the other plan. That is a cost accounting problem that does not seem to me to be insurmountable. We have got accounting standards for all kinds of things. Certainly, the plans do it among themselves when they file their SEC reports. We have got some auditor who audits it and tells you what their costs are.
    Ms. WILENSKY. You do that in Medigap, of course.
    Mr. STARK. Yeah.
    Ms. WILENSKY. You have some requirements. The other option——
    Mr. STARK. I think the Medigap payout is set so low that everybody has jumped the hurdle, but there might be somebody trying to do 50 percent and arguably that ain't enough. I do not know. Why could we not do the same thing in——
    Ms. WILENSKY. You could, but you alternatively could require information be provided to seniors about the benefits that were offered, the disenrollment, the satisfaction, and let them decide for themselves.
    Mr. STARK. I would trust you and Dr. LeRoy to figure it out. My mother could not. That is the difference.
    Ms. WILENSKY. I suspect you would help her.
    Mr. STARK. Reading annual reports is beyond the grasp of a lot of seniors. I am perfectly willing to let them decide whether they want eyeglasses or hearing aids, the stuff they can decide that affects them, but I am not sure it is not a little too difficult for them to figure out loss ratios. Joe has already said he is not sure he can figure out what the cost ratio would be.
 Page 83       PREV PAGE       TOP OF DOC
    Mr. VAN DE WATER. I have nothing to add on that.
    Mr. STARK. OK. Thank you, Mr. Chairman.
    Chairman THOMAS. Does the gentlewoman from Connecticut wish to inquire?
    Mrs. JOHNSON. Thank you. There have been studies, among them the study done by Gerald Riley at HCFA, that indicate the first year HMO enrollees are less expensive, the second year they are about average, in the third year they are more expensive. HMOs are just an insurance vehicle, and why would you want to cut reimbursements the first year when frankly the principle of insurance means that you are going to get more money for some because others are going to cost you more than they are paying in. So will you not weaken the whole HMO option if you now try to cut reimbursement for the first year people who are apparently a lower cost when after all we are trying to get people to stay in these plans the rest of their lives?
    Ms. WILENSKY. What you are trying to do is to make some adjustment for significant differences in health risks. While I believe that over time plans are likely to have a pretty typical mix of health risk, it does appear that there is a big enough difference in new enrollees that you could pay a lower payment for first year enrollees and not have an incentive to disregard these individuals. PPRC would like to see better risk adjustments basically available throughout. This is just an easy one to come up with as we try to get a little better diagnosis data and encounter data.
    Mrs. JOHNSON. I agree it is the easiest one, but it is very destructive to the overall system because we have not been successful. Look at our outlier policy. That is gross. It is a gross cut because we did not know how to do it better. And until we can guarantee the world that we can do risk, severity adjustment looking at a company's whole pool more effectively in years four, five, six, and seven, I do not really see the merit of that. Anyway I just wanted to bring that up.
 Page 84       PREV PAGE       TOP OF DOC
    The other thing I want to bring up because the lights do travel for me, the other thing I want to bring up is, in a sense, the shortsightedness of our whole discussion here. Last year we tried to track DME and IME into a trust fund and supplement it outright with taxpayer dollars because we know on this Subcommittee probably better than any other Members of Congress that the Medicare benefit plan is outdated. We have a very hard time doing anything about it. It is truly ludicrous that we gave women mammograms every other year, and I do not know whether you know this or not, but my office spent days fighting with HCFA because a woman needed a mammogram, she had a lump, but it was the wrong year. Now we never intended to deny her a diagnostic mammogram because it was the wrong year. But that is what we are into.
    I do not know whether you guys are aware that HCFA has recently changed its policy so it does not cover annual exams as it used to when people want to come in, need renewal of their hormone program. There is usually a reason to look at what is going on overall in that woman's body. But they are saving money by withdrawing reimbursements for procedures they used to automatically reimburse, and I would just say I think we need this direction, this movement, into allowing private sector plans to offer a larger benefit plan, a more integrated comprehensive plan, and I think if we try to micromanage them the way we have fee-for-service medicine, we are going to get in exactly the same problem. And fee-for-service medicine in Medicare is quietly denying payment for things that we thought they were paying for and so on and so forth because of budget pressures.
    So, this whole idea now of trying to micromanage HMO payments, we certainly have to slow their growth. It is much simpler to just say we are going to slow your growth, guys, because we are going to let Medicare only grow at this amount, and reflect that slowing of growth in the fee-for-service payment structures, which is basically what we tried to do last year, than it is to get into we are going to take away money the first year when we do not know how to hand it back to you in years five, six, and seven. So I have the feeling we are discussing the wrong issue here. If we want quality and more comprehensive programs, we have to get the AAPCC up there at a decent level so we get that, and we slow its growth, and that is how we reform Medicare.
 Page 85       PREV PAGE       TOP OF DOC
    So, on the gross level, do we not get better reform if we slow the premium growth? I would have to tell you I think we are grossly underestimating the migration. I just had a company in this morning that is deluged with applications. Why? Because Medicare is an outdated benefit program. They cannot handle all the applications, and they add a premium. They are not for free, but they are better benefits. So, I would just wonder what you all think about spending your time now beginning to think about a standard HMO package and micromanaging the reimbursement rates in the HMO sector when that whole approach is now quietly collapsing under us in the fee-for-service area? It is just a statement. I would like your feedback.
    Ms. WILENSKY. Micromanagement for a $200 billion program is a very difficult task. I would not want to encourage you to go into that more than Congress has already done.
    Mrs. JOHNSON. Thank you. My time has expired.
    Chairman THOMAS. Does the gentleman from Nevada wish to inquire?
    Mr. ENSIGN. Thank you, Mr. Chairman. I would like actually any of you to respond to this. As people leave managed care, go back into fee-for-service, Do we have any evidence of, first of all, why they are leaving; second of all, What percentage of people are leaving; and third of all, the reflection between the various plans, Do we have an assessment of if plans offer more of the services that fee-for-service does not offer, is that an incentive for them to stay in the plans, and do more of those people stay in those plans?
    Ms. WILENSKY. We do have some information. PPRC had a survey done of managed care enrollees and disenrollees. That was completed last year. About 3 percent of the people in an HMO disenroll to go back into fee-for-service. There is another 3 or 4 percent that choose a different HMO. They usually go into the HMO because of the availability of a wide range of services, prescription drugs, vision, dental, other benefits, that are not available under Medicare, and very frequently at no premium cost. Although as Mrs. Johnson indicated, there are some plans that will have small premium charges.
 Page 86       PREV PAGE       TOP OF DOC
    The reason these very small numbers of people disenroll to go back into fee-for-service are mixed. Almost one-third of them go back for reasons having to do with something going on with the plan, but the biggest other reason has to do with problems with physicians. About one-fourth—again, of 3 percent—say they had some problem with the physicians. A small number had some problems with the location or with being able to get whatever service they felt that they needed.
    As I mentioned earlier, we know the people who disenroll tend to use a lot of services after they disenroll. This may either be deliberate on their part, they knew they needed some major health care and decided to go in the open-ended fee-for-service, which they can do in 30 days, or they may have felt they were not getting what they thought they needed. So it is a little hard to disentangle, but again the number is very small, about 3 percent. But those that disenroll in the first year tend to use a lot of health care.
    Mr. ENSIGN. What is your overall opinion on the effect of risk selection with this disenrollment? How much do you think it affects the overall system?
    Ms. WILENSKY. It is a small number. I do not think the disenrollment per se has a major impact because of the number disenrolling. But it is a part of a bigger risk selection problem that we need to deal with for HMOs. To the extent we can adjust for it, we ought to do so.
    Mr. ENSIGN. How often or how long into the future do most of them disenroll? Is it after 1 year being in managed care or is it 2 years, 3 years?
    Ms. WILENSKY. We do not have a lot of good data, but as I recall, there was an initial period in which people have a higher probability of disenrolling. Then after they have been in for awhile, it is not very common. And there are plans around that have 70-plus percent of their enrollees that have been around for more than 6 years.
    Mr. ENSIGN. In southern Nevada right now, we have over 50 percent of our senior population enrolled in managed care, and it went in 2 years from a little over 30 percent to now over 50 percent, and when you start approaching numbers like that, do you think you are still going to end up with the same adverse risk selection?
 Page 87       PREV PAGE       TOP OF DOC
    Ms. WILENSKY. It is a different problem. It is less likely or it may not be a problem in terms of fee-for-service versus managed care. You can have some plans that attract poor risks because of the doctors they have or where they are located. If they do not have a way to have an adjustment in payment, they are either going to go out of business or try to find a way to shun very sick enrollees. Doing some risk adjustment is something to make sure that sick people can get into HMOs as well as to make sure the government does not overpay HMOs with healthy people. There are a lot of reasons for selection; just as some hospitals may get very sick patients, some HMOs by bad luck or where they are located, could get some very sick enrollees. You do not want them to find ways to get rid of them. You want to encourage them to be able to provide the care for sick people.
    Mr. NEWHOUSE. I would like to make two quick points, if I could, on this discussion. One is in terms of the disenrollees, I believe it is the case that most disenrollees actually enroll in another at-risk plan rather than go back to fee-for-service.
    Ms. WILENSKY. Half and half.
    Mr. NEWHOUSE. Half and half. The other point I wanted to make in terms of the argument that as the percentage in at-risk plans goes up to 50 percent or higher, the selection problem changes. This reminds me of the old story about the person that transferred from Harvard to Yale and raised the average at both places. It potentially raises the cost at the HMO but it also may raise the costs remaining in fee-for-service even further. So most of the evidence I have seen suggests it is the latter, that is to say it is not going to solve the problem. Obviously, the 100 percent solves the problem.
    Mr. ENSIGN. OK. Thank you, Mr. Chairman.
    Chairman THOMAS. Mr. Becerra, do you wish to inquire?
    Mr. BECERRA. Thank you, Mr. Chairman. This may have been answered to some degree, but could I go back to the whole issue of the disparity in profit or costs, whichever way you wish to look at it, among the various HMO providers and how some are making a great deal of money and profit, others are barely making a profit and providing most of the money they get from Medicare in services. Would it be too disciplinarian to impose some minimum ratio on HMO providers so that we know that the vast majority of the dollars that they are getting to provide a service to each Medicare beneficiary is going into services?
 Page 88       PREV PAGE       TOP OF DOC
    Ms. WILENSKY. Well, you do that in some ways with the so-called ACR, the adjusted commercial rate, that is in calculating——
    Mr. BECERRA. Right, but we found that in many cases underestimates by 12 to 35 percent what they are likely to make in profit.
    Ms. WILENSKY. Well, you have three choices, it seems to me. The first is you can use this adjusted commercial rate, but it is my impression from actuaries that there is a lot of fuzziness with that number. A second possibility is the one Mr. Stark raised earlier. You can impose a certain medical loss ratio as occurs with Medigap plans for group plans and individual policies. And the third is you can require better information about the benefits offered, the health outcomes, the satisfaction and the disenrollment, as well as all the quality monitoring and performance measures that HCFA can or should put in place, and let people judge for themselves what plans, either fee-for-service, a network, or HMOs make sense for them. That seems to me the basic three choices you have available.
    Mr. BECERRA. Any opinion on whether a minimum payment ratio would cause inefficiency in that some HMOs would just spend more to meet the——
    Ms. WILENSKY. I think that definitely would happen. It would ensure you spend no more than a certain amount on administration and profit, but plans with very low nonmedical spending are not necessarily the most efficient. It just means that most of the money is going in that direction, and those are different issues.
    Mr. BECERRA. Any of you like to give me two or three bits of data or useful information that we do not currently use to try to determine what a beneficiary should know about a particular plan in order to make a wise decision? As we try to make consumers become more informed consumers of health services, are there certain things that we do not already give them that we could easily——
    Ms. WILENSKY. We give them almost nothing. I would be glad to submit for the record some information that I think would be helpful. If you are a senior trying to decide between various Medigap plans and risk plans in your community, the aging association or your church or seniors group may have some information. There is very little formal information available about what is out there, what you know about the doctors or the health care plans or the hospitals, enrollment, disenrollment, satisfaction, or outcomes. We do not do a good job at all.
 Page 89       PREV PAGE       TOP OF DOC
    Mr. BECERRA. If you can provide the Subcommittee with that, if you could do us a favor, please be sure you disaggregate and let us know what can be implemented immediately, what requires legislation, what you think requires also some further changes in AAPCC or whatever it might be, but give us some sense of what we can do immediately, what takes a little longer, what really requires further reform, so we know what we can act on depending on what we are able to get done.
    Ms. WILENSKY. Yes.
    [The following was subsequently received:]
    [The official Committee record contains additional material here.]


    Mr. BECERRA. Appreciate that. Thank you. Thank you, Mr. Chairman.
    Chairman THOMAS. Thank you. And I would tell you that one of the things we are concerned about is that when Medicare beneficiaries are initially enrolled, they are automatically enrolled into the fee-for-service program. That helps the next panel, GAO, to work out procedures for determining relationships between the two, but it does not necessarily benefit us nor does it benefit the beneficiary because they do not get the information in a timely fashion. Your point is very well taken, and we look forward to any materials you might present to the Subcommittee.
    Does the gentleman from Nebraska wish to inquire?
    Mr. CHRISTENSEN. Thanks, Mr. Chairman, just briefly. During Dr. Wilensky's testimony, I heard you say that 86 of Nebraska's 93 counties are below the floor of $350. Matter of fact, if you go down to Arthur County and Banner County, $220, $221.
 Page 90       PREV PAGE       TOP OF DOC
    Ms. WILENSKY. The lowest in the country.
    Mr. CHRISTENSEN. The lowest in the country. And I guess one of the questions I have is even if we get some of those 86 of the 93 counties, and our top county being Douglas County in my district, $432, which is still fairly low, if we took those lower counties, I do not see HMOs jumping into that market and taking it up to $350. What other barriers are we looking at for an HMO other than just cost in terms of breaking into a market, such as western Nebraska, a rural area? While it may be an appropriate question for the final panel, I would like to hear your thoughts on that, Dr. Wilensky.
    Ms. WILENSKY. You certainly should ask the HMOs directly, but you have indicated some of the other barriers: the population density, the facility availability. I suspect if you increase the payment substantially, there will be a mix of other plans if you allow them to be offered. If it were possible to have a network plan, a preferred provider organization, or a PSO network or other type of organization in addition to the traditional HMO, and the payment were somewhat higher, I suspect you would see the physicians and hospitals who are now providing services to your seniors organize in a risk group.
    I have been out to Kearney, Nebraska, and I spoke to the hospitals and physicians there. They had indicated it would make a very large difference if the payment were different, but that is a larger area, and I think there will still be barriers in some places. But more flexibility in terms of the kinds of plans that were made available to seniors would help break down some of the difficulties that now exist.
    Mr. CHRISTENSEN. Thank you. Thank you, Mr. Chairman.
    Chairman THOMAS. I tell you, Jon, one of the problems is that in this AAPCC component, it is the medical usage and that in some of the high areas, people get up in the morning, look in the mirror, look better, feel better, and decide to go to the doctor to find out why. And at the other extreme, if you have no pulse, take two aspirin and call me in the morning is probably not the appropriate response. The difficulty in looking at these averages is to figure out, one——
 Page 91       PREV PAGE       TOP OF DOC
    Mr. CHRISTENSEN. Mr. Chairman, we have a lot of hardy farmers out in western Nebraska that have virtually no pulse and take two aspirins in the morning and go right back to work.
    Chairman THOMAS. Well, there is a degree of the type of person who is willing to deal with infirmities as part of life versus those who are not, and it is unfair to lock that structure in, and what we have been talking about is trying to figure out a way to break it away from those rather archaic determiners.
    Does the gentleman from Wisconsin wish to inquire?
    Mr. KLECZKA. No, Mr. Chairman.
    Chairman THOMAS. I want to thank the panel very much. I look forward to your returning with more specific proposals and Mr. Van de Water with a clear understanding of what the President's proposals look like through the CBO process.
    Thank you very much.
    The next panel is designed to give us a look at the aggregation of information, what works and what does not, and what may be available in terms of a slightly different way of looking at those options. The panel members are Dr. William Scanlon, who is the Director of the Health Systems Issues of the General Accounting Office, and Alice Rosenblatt, who is a member of the Board of Directors of the American Academy of Actuaries, Health Practice Council. Dr. Scanlon, we have your written testimony. It will be a part of the record, and if you will address us in any way you see fit in the time you have, we look forward to your testimony.


 Page 92       PREV PAGE       TOP OF DOC
    Mr. SCANLON. Thank you very much, Mr. Chairman and Members of the Subcommittee. I am very pleased to be here today as you consider HMO payment rates. As you know, 10 years of research——
    Chairman THOMAS. Bill, if you would turn that mike directly toward you and speak directly into it. They are very unidirectional and not very good.
    Mr. SCANLON As you know, 10 years of research shows that the Medicare payments method for HMOs can produce excess payments as a result of favorable selection, or the tendency of HMOs to attract Medicare beneficiaries who are healthier than those remaining in fee-for-service. As we have heard today, there have been a number of proposals made to help alleviate Medicare's HMO payment problems, both proposals discussed today as well as changes within the Balanced Budget Act of 1995. At your request, Mr. Chairman, we have reviewed HCFA's method for setting HMO rates to identify feasible options for promptly reducing the amount of excess payments.
    A comprehensive discussion of our findings will be published in a forthcoming report. Today, what I would like to do is focus my comments on a modification that we have identified to Medicare's current HMO rate-setting method that could help immediately reduce excess payments. It would also reduce excess payments under the administration's proposed method, as we understand it, or other methods that rely on fee-for-service costs to initially set or update HMO payment rates. Most importantly, it targets the adjustments where excess payments are occurring rather than reducing payments across the board unfairly affecting some HMOs and beneficiaries. Central to the current rate-setting method and proposals for change is an estimate of the average cost county by county of serving Medicare beneficiaries under fee-for-service. The actual rates paid are set by adjusting the county averages up or down based on each individual's risk of incurring higher or lower costs.
    Considerable attention has been focused on the failure of the current risk adjustment methodology to adequately account for favorable selection. Our work centers instead on improving the estimates of the average cost of serving a county's beneficiaries, a problem that has largely been ignored. HCFA's method for determining the county average costs or AAPCC exclude HMO's enrollees' expected costs in estimating their per beneficiary average cost. The result is that in counties experiencing favorable selection, HCFA's method will overstate the average cost of all Medicare beneficiaries and lead to overpayments.
 Page 93       PREV PAGE       TOP OF DOC
    Our proposed modification is designed to fix that problem. We have developed a way to estimate HMO enrollees' expected fee-for-service cost using information available to HCFA. Our approach produces a county rate that more accurately represents the cost of all Medicare beneficiaries. To illustrate the effect of our modification, we analyzed data for counties with different shares of beneficiaries enrolled in HMOs. We found that the application of our approach could have reduced excess payments by about 25 percent. Substantially better risk adjustment, which appears to be years away from implementation, would target the remaining 75 percent of excess payments.
    For the counties that we analyzed, which contain 30 percent of all Medicare HMO enrollees nationwide, our modification would have reduced payments by $276 million in 1995. Despite the size of these savings, our modification would have produced relatively small changes ranging from $3 to $38 in the county's monthly per beneficiary capitation payments. It is sometimes suggested that the excess payment problem will be mitigated as more beneficiaries enroll in Medicare managed care and HMOs contain a more expensive mix of beneficiaries. Our analysis does not support this hypothesis. Our data, which included counties with up to 36 percent HMO penetration in 1995, indicated that the overstatement of county average costs is very often larger in counties with higher Medicare HMO penetration. Indeed, there is a tendency for the overstatement to increase as HMO enrollment increases.
    In conclusion, many fear that enrolling more beneficiaries in managed care could increase rather than lower Medicare spending unless Medicare's method for paying HMOs is revised. Our approach to improving the estimate of county rate average costs could lessen the risk of overpayment and could lower Medicare expenditures by several hundred million dollars each year. Moreover, it is a targeted approach. It reduces payments more for HMOs in counties where excess payments are higher and less for HMOs in counties where excess payments are lower.
    Our method would improve payment accuracy whether the Congress adopts any proposal that uses county averages cost as a baseline, whether HCFA develops and adopts better risk adjustors, or Medicare's current rate-setting process remains in place. We believe it is one of the better tools that you requested to improve payments now.
 Page 94       PREV PAGE       TOP OF DOC
    Thank you, Mr. Chairman. I would be happy to answer any questions you or Members of the Subcommittee may have.
    [The prepared statement follows:]

Statement of William J. Scanlon, Ph.D., Director, Health Financing and Systems, Health, Education, and Human Services Division, U.S. General Accounting Office

    Mr. Chairman and Members of the Subcommittee:
    We are pleased to be here today to discuss the rates Medicare pays health maintenance organizations (HMO) in its risk contract program, Medicare's principal managed care option. As you know, Medicare's method for paying risk contract HMOs was designed to save the program 5 percent of the costs for beneficiaries who enroll in HMOs. However, 10 years of research on Medicare's costs under HMOs has found that the program's rate-setting method results in excess payments to HMOs because HMO enrollees would have cost Medicare less if they had stayed in the fee-for-service sector. Recently, the Physician Payment Review Commission (PPRC) estimated that annual excess payments to HMOs nationwide could total $2 billion.
    A number of proposals have been made recently to help alleviate Medicare's HMO payment problems. For example, the proposed Balanced Budget Act of 1995 called for, among other things, mechanisms to lessen rate disparities across geographic areas and to decouple annual HMO rate increases from annual fee-for-service spending increases. The administration's current budget proposal adopts several provisions from the proposed Balanced Budget Act but also adds new twists—such as an across-the-board reduction in Medicare's HMO payments that would lower the payments from 95 percent to 90 percent of estimated fee-for-service costs. Under the auspices of the Health Care Financing Administration (HCFA), which administers the Medicare program, several demonstration projects are planned or under way, including efforts to improve risk adjustment and using a process of competitive bidding to set rates.
 Page 95       PREV PAGE       TOP OF DOC
    At the request of the Subcommittee's Chairman, we reviewed HCFA's method for setting HMO rates to identify feasible options for promptly reducing the amount of excess payments. A comprehensive discussion of our work is included in a forthcoming report. In conducting our study, we reviewed previous research on Medicare's HMO rate-setting method, analyzed available HCFA data, and had our findings reviewed by experts on HMO payment issues, including staff at PPRC and HCFA.
    Today, I would like to focus my comments on our proposed modification to HCFA's HMO rate-setting method. We believe this modification could help reduce excess HMO payments under Medicare's current payment method, the administration's method, or other methods that rely on fee-for-service costs to set initial HMO rates or update those rates. Central to the current method and proposals for setting HMO rates is an estimate of the average cost of serving Medicare beneficiaries under fee-for-service in defined geographic areas (currently, counties). The actual rates paid HMOs for an enrollee are set by adjusting these averages up or down based on the enrollee's ''risk'' of incurring higher or lower costs. Considerable attention has focused on the failure of current risk adjustment methodology to adequately account for favorable selection, the term used to describe the tendency of HMOs to attract a population of Medicare seniors whose health costs are generally lower than those of the average beneficiary. Our work centers on the estimate of average cost of serving a county's beneficiaries: the county rate.
    In summary, we found that HCFA's current method of determining the county rate produces excess payments. Because HCFA's method excludes HMO enrollees' costs from estimates of the per-beneficiary average cost, it bases county payment rates on the average per-beneficiary cost of only those beneficiaries that remain in the fee-for-service sector and ignores the costs HMO enrollees would have incurred if they had remained in fee-for-service. Research has shown the costs of those remaining in fee-for-service to be higher on average than the likely costs of HMO enrollees. A difficulty in correcting the problem is that HCFA cannot directly observe the costs HMO enrollees would have incurred if they had remained in the fee-for-service sector. Our proposed modification is designed to fix that problem. We developed a way to estimate HMO enrollees' expected fee-for-service costs using information available to HCFA. Our approach produces a county rate that represents the costs of all Medicare beneficiaries and could result in hundreds of millions of dollars in savings to Medicare.
 Page 96       PREV PAGE       TOP OF DOC

How Medicare Determines an HMO's Payment Rate

    Essentially, HCFA's calculation of its per-enrollee (capitation) rate can be expressed as follows:

    Capitation rate = average per-beneficiary cost × .95 × risk adjustment factor

    Medicare pays risk HMOs a fixed amount per enrollee—a capitation rate—regardless of what each enrollee's care actually costs. Medicare law stipulates that the capitation rate be set at 95 percent of the costs Medicare would have incurred for HMO enrollees if they had remained in fee-for-service. In implementing the law's rate-setting provisions, HCFA estimates a county's average per-beneficiary cost and multiplies the result by 0.95. The product is the county adjusted average per capita cost rate.
    HCFA then applies a risk-adjustment factor to the county rate. Under HCFA's risk-adjustment system, beneficiaries are sorted into groups according to their demographic traits (age; sex; and Medicaid, institutional, and working status). HCFA calculates a risk factor for each group—the group's average cost in relation to the cost of all beneficiaries nationwide. For example, in 1995 the risk factor for younger seniors (65- to 70-year-old males) was .85, whereas for older seniors (85-year-old or older males) it was 1.3. HCFA uses the risk factor to adjust the county rate, thereby raising or lowering Medicare's per capita payment for each HMO enrollee, depending on the individual's demographic characteristics.

Medicare's HMO Rate-Setting Method Has Led to Excess Payments
 Page 97       PREV PAGE       TOP OF DOC

    One reason the HMO rate-setting method overstates the expected fee-for-service costs of HMO enrollees is that it uses only the cost experience of fee-for-service beneficiaries. If the health status of the mix of beneficiaries enrolled by HMOs were the same as the health status of those in fee-for-service, using fee-for-service beneficiaries to estimate the expected fee-for-service costs of HMO enrollees would be an appropriate method. However, because research has shown that HMOs have in general enrolled healthier-than-average beneficiaries, the beneficiaries remaining in fee-for-service represent a sicker-than-average population. This, in turn, means that using data on fee-for-service beneficiaries exclusively produces HMO payment rates higher than envisioned when the current rate setting provisions were enacted.
    Medicare's risk adjustors explain about 3 percent of the variation in individual-level health care costs and are thus not adequate to account for the cost differences among beneficiaries. The difficulty is that, within the same demographic group, HMO enrollees are healthier than fee-for-service beneficiaries; for example, 70-year-old males in HMOs are, on average, healthier than 70-year-old males in fee-for-service. Medicare's risk adjustor is said to be inadequate because, while it makes broad distinctions among beneficiaries of different age, sex, and other demographic characteristics, it does not account for the significant health differences among demographically identical beneficiaries. The cost implications of health status differences can be dramatic—for two demographically alike beneficiaries, one may experience occasional minor ailments while the other may suffer from a serious chronic condition.

Including HMO Enrollees' Costs in County Average Improves Accuracy of County Rates

    Independent of improved risk adjustment, modifying the method for calculating the county rate would help reduce Medicare's excess HMO payments. In setting county rates, HCFA currently estimates the average Medicare costs of a county's beneficiaries using the costs of only those beneficiaries in Medicare's fee-for-service sector. This method would be appropriate if the average health cost of fee-for-service beneficiaries were the same as that of demographically comparable HMO enrollees. However, in counties where there are cost disparities between Medicare's fee-for-service and HMO enrollee populations, this method can either overstate the average costs of all Medicare beneficiaries and lead to overpayment or understate average costs and lead to underpayment. Correcting this problem is difficult because it is impossible to observe the costs HMO enrollees would have incurred if they had remained in the fee-for-service sector. Therefore, we developed a method to estimate HMO enrollees' expected fee-for-service costs using information available to HCFA. Our method consists of two main steps:
 Page 98       PREV PAGE       TOP OF DOC
    •  First, we compute the average cost of demographically similar new HMO enrollees during the year before they enrolled—that is, while they were still in fee-for-service Medicare. These fee-for-service costs are available through HCFA's claims data.
    •  Next, we adjust this amount to reflect the expectation that a new enrollee's use of health services will, over time, rise.
    Having completed these steps, we combine the result with an estimate of the average cost of fee-for-service beneficiaries. This new average produces a county rate that reflects the costs of all Medicare beneficiaries.

Selected 1995 County Rates Produced Substantial Excess Payments

    To illustrate the effect of our approach, we analyzed data for counties with different shares of beneficiaries enrolled in HMOs. We chose counties within a single state to eliminate variations attributable to state differences. We selected California because it covers 36 percent of all Medicare HMO enrollees and includes counties that in 1995 had the nation's highest HMO penetration rates. We found that our method could have reduced excess payments by more than 25 percent. Although better risk adjustors could further reduce the large remainder of excess payments, improvements to risk adjustment require developing direct measures of health status, which is a complex effort that may take years.
    The following key points also emerged from our analysis:
    •  First, for the counties that we analyzed, we estimate that total excess payments in 1995 amounted to about $1 billion (of about $6 billion in total Medicare payments to risk HMOs in the state). Of that amount, applying our method for setting county rates would have reduced the excess by about $276 million.
    •  Second, the excess payments attributable to inflated county rates were concentrated in 12 counties with large HMO enrollment and ranged from less than 1 percent to 6.6 percent of the counties' total HMO payments, representing between $200,000 and $135.3 million. Despite the size of these amounts, the application of our method would have produced relatively small changes in the monthly, per-beneficiary capitation payments, ranging from $3 to $38.
 Page 99       PREV PAGE       TOP OF DOC
    •  Third, our analysis did not support the hypothesis, put forward by the HMO industry and others, that the excess payment problem will be mitigated as more beneficiaries enroll in Medicare managed care and HMOs progressively enroll a more expensive mix of beneficiaries. Our data—which include counties with up to a 39-percent HMO penetration in 1995—indicated that the disparity between Medicare rates and our rates is larger in counties with higher Medicare penetration. For example, the four counties with the highest rates of excess payment, ranging from 5.1 to 6.6 percent, were also among the counties with the highest enrollment rates in 1995.

Data Are Available to Enable HCFA to Promptly Adjust County Rates

    Because the data we used to estimate HMO enrollees' costs come from data that HCFA compiles to update HMO rates each year, our method has two important advantages. First, HCFA's implementation of our proposal could be achieved in a relatively short time. The time element is important, because the prompt implementation of our method would avoid locking in a current methodological flaw that would persist in any adopted changes to Medicare's HMO payment method that continued to use current county rates as a baseline or fee-for-service costs to set future rates. Second, the availability of the data would also make our proposal economical: we believe that the savings to be achieved from reducing county-rate excess payments would be much greater than the administrative costs of implementing the process.


    Medicare's HMO rate-setting problems have prevented it from realizing the savings that were anticipated from enrolling beneficiaries in capitated managed care plans. In fact, enrolling more beneficiaries in managed care could increase rather than lower Medicare spending—unless Medicare's method of setting HMO rates is revised. Our method of calculating the county rate would have the effect of reducing payments more for HMOs in counties with higher excess payments and less for HMOs in counties with lower excess payments. In this way, our method represents a targeted approach to reducing excess payments and could lower Medicare expenditures by at least several hundred million dollars each year.
 Page 100       PREV PAGE       TOP OF DOC
    Furthermore, our approach is useful under several possible scenarios, including whether (1) the Congress adopts any proposal that uses current county rates as a baseline, (2) HCFA develops and adopts improved risk adjustors, or (3) the Congress takes no action and preserves Medicare's current rate-setting process.
    Mr. Chairman, this concludes my prepared statement. I would be pleased to answer any questions.

For more information on this testimony, please call Jonathan Ratner, Associate Director, on (202) 512–7107 or James C. Cosgrove, Assistant Director, on (202) 512–7029. The analysis was conducted by Scott L. Smith, Project Director, and Richard M. Lipinski, Project Manager. Other major contributors to this statement included Thomas Dowdal and Hannah F. Fein.

Related GAO Products

    Medicare HMOs: Rapid Enrollment Growth Concentrated in Selected States (GAO/HEHS–96–63, Jan. 18, 1996).
    Medicare Managed Care: Growing Enrollment Adds Urgency to Fixing HMO Payment Problem (GAO/HEHS–96–21, Nov. 8, 1995).
    Medicare: Changes to HMO Rate Setting Method Are Needed to Reduce Program Costs (GAO/HEHS–94–119, Sept. 2, 1994).


    Chairman THOMAS. Thank you very much, Dr. Scanlon.
    Ms. Rosenblatt.
 Page 101       PREV PAGE       TOP OF DOC


    Ms. ROSENBLATT. Thank you, Mr. Chairman. My name is Alice Rosenblatt, and I am here today representing the American Academy of Actuaries. I currently serve on the board of that organization, and I have previously chaired the Risk Adjustment Work Group of that organization. I will be talking about risk assessment, risk adjustment, the relation of these to the current Medicare risk contracting program, and I will also be commenting on how to improve the current program, and I will be making comments on some of the earlier discussion.
    Examples of ways of doing risk adjustment include demographics like age and sex and health status mechanisms like self-reported health status, diagnoses and prior use. Popular prior use models include something called the ACGs, or ambulatory care groups, which use age/sex and ICD–9 classifications and also DCGs or diagnostic cost groups which use different groupings of the ICD–9. There has been a lot of discussion this morning about the classification system which Medicare currently uses which is the AAPCC, but what is also going on in the Medicare system is something called the ACR or the adjusted community rate.
    If this ACR is less than the AAPCC for a given health plan, then the plan can either add benefits or return the excess to HCFA. And many health plans do add benefits such as prescription drugs.
    There have been several studies that conclude that the AAPCC methodology increases overall costs. These studies base this conclusion on the fact that they find risk plans cover the healthier segments of the population. To cite an example, in an extreme case where the risk plans were covering people who had no claims, then every dollar of the prepayment reimbursement to those plans is 100 percent in excess of what the fee-for-service payment would have been.
 Page 102       PREV PAGE       TOP OF DOC
    In general, we believe the criteria for a risk assessment method should include accuracy, should be practical and understandable and, therefore, low cost. Timeliness is required. Predictability is required for solvency concerns on the part of the health plans. And whatever method is used should not be subject to a health plan's ability to manipulate it or game it.
    Some of the advantages of the current AAPCC system include that it is a simplified form of risk assessment. It is based on clear, auditable classifications like age, sex, and institutional status. Therefore, it is administratively easy and probably low cost, and it is not susceptible to gaming. However, the ACR modification could produce a problem.
    Some of the limitations of the current methodology are that the county variations cause HMOs to compete heavily in high payment areas and few HMOs offer coverage in low payment areas. If the healthier people join the HMOs, then the fee-for-service payments due to the linkage will escalate due to adverse selection, and this will tend to continue the escalation of the AAPCC and cause further overpayment.
    For the plans that may attract sicker members of the population, due to perhaps broader provider networks or strong specialty networks, those plans currently receive no additional compensation for the higher cost care of those people. The Society of Actuaries sponsored research on risk adjustment recently, and the findings were that even with the best risk assessment methods, there were still problems and opportunities for risk selection.
    Some suggestions from the academy include reducing the wide variations in the geographic region values because they currently are not an accurate reflection of geographic driven costs; use of a credibility approach or blended approach similar to that proposed by Joe Newhouse; the carve-out of particular disease populations; modification of the ACR methodology; basing the AAPCC on something other than fee-for-service payments; and also the combination of studies and demonstration projects on risk assessment.
 Page 103       PREV PAGE       TOP OF DOC
    We do believe that across-the-board adjustments could be a problem for the plans that currently enroll the higher cost populations. Also, the administration's plans to allow provider-sponsored organizations concern us in that we hope that those organizations will be subject to the same solvency standards as insurance carriers.
    I would like to comment on the loss ratio issue, although I see that my time is up. There are problems with what was mentioned by Mr. Stark on suggesting a loss ratio methodology. As much as risk adjustment is a problem to statisticians and actuaries, the determination of incurred claims which goes into the medical loss ratio is also an actuarial art rather than a science.
    There are also problems due to catastrophic claims where a given large claim in any 1 year can cause the loss ratio to go up. There are even problems with the definition of what is a claim versus what are administrative costs. For example, if a provider does utilization management or the health plan does utilization management, you may end up with two different definitions. Different benefit plans have different administrative requirements, and there are also different issues for staff model HMOs versus other models.
    I would also like to comment on the comments made about the small percentage of people disenrolling from HMOs, if I may. I have done studies of the population under 65, and a statistic that one encounters there is that 5 percent of the claimants generate 50 percent of the claim costs. So although we are talking about a small percent of people disenrolling; i.e., 3 percent, the impact of that from a cost point of view may be great.
    Thank you very much, and I would like to offer the Academy's assistance in helping to solve this very important problem.
    [The prepared statement follows:]

Statement of Alice Rosenblatt, Board of Directors; and Chairperson, Risk Adjustment Work Group, American Academy of Actuaries
 Page 104       PREV PAGE       TOP OF DOC

    The American Academy of Actuaries is the public policy organization for actuaries of all specialties within the United States. In addition to setting qualification standards and standards of actuarial practice, a major purpose of the Academy is to act as the public information organization for the profession. The Academy is nonpartisan and assists the public policy process through the presentation of clear, objective analysis. The Academy regularly prepares testimony for Congress, provides information to senior federal elected officials and congressional staff, comments on proposed federal regulations, and works closely with state officials on issues related to insurance.
    The American Academy of Actuaries has analyzed numerous elements involved with health care reform initiatives, especially risk adjustment. The actuarial profession is uniquely qualified to examine the various alternatives to reform Medicare due to the profession's extensive practical experience designing efficient health care plans for employers and other organizations in the private sector. Additionally, the actuarial profession has provided actuarial input on the Medicare program and other social insurance programs.
    The Subcommittee on Health of the Committee on Ways and Means asked the Academy to discuss risk assessment and risk adjustment, how it relates to the current Medicare risk contracting program, and comment on the ways to improve the current risk contracting program as well as provide comments on provisions within the Administration's 1998 budget proposal regarding HMO payments.

Risk Assessment and Risk Adjustment

    Risk assessment and risk adjustment are methods intended by policymakers to promote competition on the basis of medical and administrative efficiency, rather than risk selection. This is done in a two-step process: measuring the expected relative health care costs of the individuals enrolled by a plan (risk assessment) and adjusting payments to plans to reflect the relative risks (risk adjustment).
 Page 105       PREV PAGE       TOP OF DOC
    Health risk assessment measures the actual or expected deviation of each individual's cost from the average cost. The assessment objectively determines the relative risks of individuals or groups of individuals by using a classification system: individuals are classified by objective criteria into one of several categories. Risk classification is the process of grouping individuals or groups with similar risk characteristics, so that differences in expected costs can be appropriately recognized. Each classification is assigned a numerical value, so that a weighted average value can be determined and used to compare the relative risk of one population versus another.
    Risk adjustment is a method used to compensate health plans for differences in risks. The risk adjustment process uses the results of risk assessment to determine the compensation to health plans.
    It is important to distinguish between prospective and retrospective risk adjustment methods. Retrospective adjustment is based on the actual experience (type and amount of claims) during the period, resulting in an adjustment following the end of the period. Prospective risk adjustment methods apply an amount to the premiums (prepayments) that adjust for the different expected risks of the insureds. Retrospective risk adjustment methods develop health-plan payments in a settlement process, such as at the end of a calendar year, and could be combined with an interim payment methodology. Prospective adjustments are applied at the beginning, predicting the future risk based on prior history.

Goals of Risk Adjustment

    Risk adjustment methods can be designed to help accomplish several goals:
    •  Help reduce the effects of either inadvertent or intentional risk selection, so that health plans in a competitive market can compete on the basis of medical and administrative efficiency and quality of service and care rather than on the ability to select good risks.
 Page 106       PREV PAGE       TOP OF DOC
    •  Compensate health plans fairly and equitably for risks they assume.
    •  Maintain consumer choice from among multiple health plans based on premiums and benefits that reflect relative medical and administrative efficiencies.
    •  Protect the financial soundness of the system.

Examples of Risk Assessment

    The objective of risk assessment is to measure the expected costs of individuals or groups so it can be used to assign levels of relative risk to these individuals or groups. Existing models use two general types of data to predict the expected costs: demographic variables and health status. Demographic variables include age, sex, family status, location, and welfare status. Measures of health status can range from self-reported health, to diagnoses, to prior use of health care services.
    Some of the prior use risk assessment methods generally referenced today include:
    •  Ambulatory Care Groups (ACGs): This is a model that uses age/sex, and ICD–9 diagnoses to classify risks. It was developed by Jonathan Weiner and others at Johns Hopkins University.
    •  Diagnostic Cost Groups (DCGs): This is a prior-history model that uses a different grouping of ICD–9 diagnoses to classify risks. This model was developed by Arlene Ash and Randy Ellis of Boston University.

Current Medicare Risk Assessment Method

 Page 107       PREV PAGE       TOP OF DOC
    Medicare pays per-capita payments in advance to eligible managed care organizations, such as health maintenance organizations (HMOs). The Health Care Financing Administration (HCFA) has a detailed method to determine payment for each person in the managed care plan. These managed care plans are paid monthly, in advance, for each enrollee at a rate equal to 95 percent of the Average Adjusted Per Capita Cost (AAPCC). The AAPCC is considered a method of risk assessment.
    The AAPCC methodology places beneficiaries into risk categories that account for age, sex, welfare status, institutional status, and basis for Medicare eligibility (age, disability, or end-stage renal disease). The Medicare HMO payment methodology is tied to its fee-for-service data. Payments are based on 95 percent of the average of a geographic county's costs to Medicare for its fee-for-service (traditional) program. For each geographic county, there are payment amounts for each of these risk categories. The payment to a particular health plan is based on summing these results for each covered person.
    There is also an adjustment, called the ACR (Adjusted Community Rate) that theoretically adjusts for the health plan's cost structure. If the ACR is less than the AAPCC, then the at-risk plan may add benefits or return the excess funds to HCFA. This results in many health plans adding benefits, such as prescription drugs, to the basic benefit package.
    The geographic variation in the AAPCC for 1997 is illustrated by the following values:

    National Average = $467

    Some high areas are:

    Richmond, N.Y.$767
 Page 108       PREV PAGE       TOP OF DOC
    Dade, Fla.$748
    Bronx, N.Y.$728

    Some low areas are:

    Holmes, Ohio$225
    Banner, Neb.$221
    Arthur, Neb.$221

    There have been studies on the impact of this payment methodology on Medicare's overall costs. Some of these studies conclude that the AAPCC methodology increases the overall costs. These studies claim that the Medicare risk plans have covered the healthier segments of the population, and that payment based on 95 percent of fee-for-service payments overcompensates these health plans. In an extreme case, if the Medicare risk plans received prepayments for individuals with zero claim costs, then 100 percent of the prepayment is in excess of what the fee-for-service payment would have been.
    An analysis done by the Physician Payment Review Commission examined enrollees in at-risk plans between 1989 and 1994 and compared their spending with a controlled group of individuals who remained in the traditional fee-for-service Medicare environment. The spending comparison was done when both groups were enrolled in the traditional fee-for-service Medicare. The analysis showed that those enrolled in at-risk plans spent 63 percent as much as the average Medicare beneficiary during the six months preceding enrollment when both groups were enrolled in traditional Medicare. Persons who disenrolled and returned to traditional fee-for-service Medicare spent 160 percent of the average beneficiary in the six months following disenrollment.
 Page 109       PREV PAGE       TOP OF DOC
    The study by PPRC, as well as observations by other studies, are evidence that Medicare at-risk plans are benefiting from favorable selection. Namely, these plans are attracting the healthier people in the Medicare population leaving Medicare fee-for-service with the sicker population. The result is that the overall costs of the program are increased due to the prepayment methodology for the risk plans.

Criteria for Risk Assessment and Risk Adjustment

    The following criteria can be used in analyzing the pros and cons of a particular risk adjustment method.
    •  Accuracy: Because financial payments to health plans will be determined based on the risk adjustment mechanism, accuracy and avoidance of statistical bias is critical.
    •  Practical and Understandable, Low Cost: The risk-adjustment mechanism cannot be so complex that implementation is extremely cumbersome, thereby adding additional cost to the system.
    •  Timeliness and Predictability: Health plans need time to develop necessary marketing and enrollment material, and need predictability to avoid solvency concerns.
    •  No Manipulation: The risk-adjustment mechanism should aim to make it impossible for specific health plans to benefit financially by ''gaming'' the mechanism.

Analysis of AAPCC Methodology

    The current method HCFA uses to determine risk assessment has some advantages:
 Page 110       PREV PAGE       TOP OF DOC
    •  The current AAPCC technique is a simplified form of risk adjustment based on age, sex, welfare status, institutional status, and basis for Medicare eligibility (age, disability, or end-stage renal disease).
    •  The entire AAPCC methodology requires little administrative cost. The classifications of each individual (age, sex, institutional status, etc.) are easy to identify.
    •  Because of the classifications used, the AAPCC methodology is not susceptible to gaming and can be audited. However, since AAPCC payments are modified by the ACR, there is a possibility that the rigor of the well-defined classifications would be lost.
    However, this Committee and Congress should consider the following limitations in the methodology:
    •  The county variations caused by fee-for-service payment linkage cause HMOs to compete heavily in the high-payment areas, with few HMOs offering coverage in low-payment areas.
    •  If healthier segments of the population join the HMOs, then the fee-for-service payments will escalate due to adverse selection. (Adverse selection is when individuals are motivated, either directly or indirectly, to take advantage of a risk classification system. For example, individuals may enroll in the traditional fee-for-service plan when they know they need specific care for an illness). What this means is that the average risk profile of those remaining in the traditional fee-for-service plan will continue to deteriorate, causing higher and higher fee-for-service payments and more excessive AAPCC amounts due to the fee-for-service linkage.
    •  Although evidence suggests that Medicare risk plans attract the healthier people in the Medicare population, there may be some plans that tend to attract the sicker members of the population, due perhaps to broader provider networks, or strong specialty networks. These plans are most disadvantaged by the current AAPCC methodology due to the fact that such plans receive no additional compensation for the higher-cost care these sicker members will require.
 Page 111       PREV PAGE       TOP OF DOC

Analysis of Risk Assessment Methods

    Last year, the Society of Actuaries sponsored a research project entitled ''A Comparative Analysis of Methods of Health Risk Assessment.'' This project was undertaken by analysts at Coopers and Lybrand, LLP and Harvard University. The study compared the predictive accuracy of different risk-assessment methods, compared the different risk-assessment methods based on other criteria including those criteria listed above (such as administrative practicality and ability to resist manipulation by insurers), and explored the potential for risk adjustment using a list of high-cost conditions. The study concluded that all models of risk assessment perform well for large random groups, including the age and sex model. If enrollees distribute themselves randomly across plans, then the current risk-assessment methods are sufficient. Unfortunately, evidence shows that this is not the case for the Medicare program. Therefore, it is important that the risk assessment method predict well for individuals and non-random groups. The findings of the report indicate that opportunities for risk selection and inequities in payments remain even with the best current risk-assessment models. Although no current risk-assessment method can completely remove incentives for risk-selecting behavior, some methods produce better results than others.
    The age/sex model, which like the current AAPCC methodology is based on demographic criteria, had the lowest predictive accuracy of the eight models tested. However, the age-sex model is easy to administer, resistant to manipulation, and provides no incentives for unnecessary care.
    All the diagnosis methods might tend to cause up coding—coding diagnoses as worse than they were to increase carrier payments. Another disadvantage of diagnosis-based models is that it takes a long time to collect and analyze the necessary data. The Society of Actuaries research study also showed that there are differences in the reporting and coding of diagnoses between health plans. Due to these differences, methods like the ACG and DCG may reward or penalize plans based on how well they record and capture diagnosis information. It should be pointed out that record keeping would improve if payments were based on such methods.
 Page 112       PREV PAGE       TOP OF DOC
    It is important that any risk assessment method based on prior usage not reward plans based on over-treatment. For example, if the prior-usage models were based on actual costs, rather than diagnoses, plans selecting more-expensive treatment protocols would be financially rewarded.

Possible Improvements to Current AAPCC Method

    Concern has been raised over the inadequacies in the current AAPCC method. Below are a few suggestions to correct or minimize the current problems in the methodology.
    •  Reduce the wide variations in the geographic region values of the AAPCC since the variation is not an accurate reflection of the geographic driven costs.
    •  Use a ''credibility'' approach, similar to the approach recommended by Joseph P. Newhouse. This approach would base payments to health plans on a combination of prepayments and a plan's actual claims experience. However, the problem with this approach is that, as noted above, the portion of the payment based on actual experience may reward the plans that have more-expensive treatment protocols.
    •  Consider ''carving out'' particular disease populations (such as diabetes) or particular procedures (such as coronary artery bypass graft surgery). This adjustment would separate disease populations and/or high cost procedures from the base, and thereby limit the variation in costs due to risk differentials. This approach applies the same logic as the end-stage renal disease program.
    •  Modify the ACR methodology because it is extremely complex. The American Academy of Actuaries can provide detailed comments on redesigning the ACR, if desired.
    •  Base the AAPCC methodology on something other than fee-for-service payments. One alternative is to consider the use of a competitive bidding process.
 Page 113       PREV PAGE       TOP OF DOC
    •  Continue studies and demonstration projects on risk assessment and adjustment methods to improve on those currently in use.

Comments on Administration's Budget Proposal

    The Administration's fiscal year 1998 budget proposal includes a phased-in reduction in HMO payment rates from the current 95 percent of the fee-for-service payments to 90 percent beginning in the year 2000. The plan would also narrow the payment disparities between urban and rural areas by blending local and national rates. The Academy was asked by this Committee to discuss alternatives to an across-the-board cut, such as modifying the risk-assessment method or modifying the current payment method which links HMO payments to fee-for service payments. As noted above, both appear to be alternatives worthy of further discussion and consideration by Congress. Across-the-board adjustments such as the change from 95 percent to 90 percent will not address the inequities of the current system. Plans with higher-than-average cost populations will face payment cuts the same as plans with lower-than-average cost populations.
    The Administration's budget plan would remove the current payments on graduate medical education and disproportionate-share hospital payments from the HMO payment formula and pay the amount directly to academic health centers and HMOs with residency programs. This change would improve the current HMO payment formula by reflecting the actual costs for providing care to the enrollees.
    Another item related to HMOs in the Administration's budget plan is allowing provider-sponsored organizations (PSOs and PPOs) to participate in Medicare under minimum federal standards that apply to current Medicare contractors, with states imposing more stringent standards after four years. If Congress is concerned about a level playing field for those participating in Medicare, it will be necessary to ensure that PSOs are subject to similar regulatory requirements as HMOs and traditional insurers. Specifically, the American Academy of Actuaries is concerned about adequate solvency standards for new health entities, such as PSOs. For a detailed analysis on this issue, refer to the current work by the American Academy of Actuaries on solvency regulation and risk-based capital for health organizations.
 Page 114       PREV PAGE       TOP OF DOC


    The American Academy of Actuaries is available to provide assistance to Congress and the Administration as steps to improve the current Medicare system are being considered. We offer our help in analyzing potential solutions and in helping elected officials and their staff understand the potential implications of any short-term actions.


    Chairman THOMAS. Thank you very much, both of you.
    Ms. Rosenblatt, your last statement is evidence throughout the health care field, and that is in terms of the groups generating most of the costs, it is always a relatively small number. I think that created the concept of insurance and pooled risk and a number of other issues that lead us to our current position.
    Bill, in your testimony, you talk about the structure that you want to offer as an option, looking at folks before they go in and come out. As I mentioned earlier, there is a concern if, in fact, they go directly into HMOs, we will not have that front snapshot, but at least given the current structure, it is something that makes some sense to put in. My concern is that HCFA will argue that making some of these changes might take some time. From your analysis, do you have any estimate of how long it would take to put in some of these risk adjustors that you have suggested, for example, or others that were mentioned? My feeling is that this is not something that would take years to execute.
    Mr. SCANLON. We agree with you completely. We undertook this analysis and had the challenge of becoming familiar with HCFA's data systems and understanding HCFA's claims. We also had the challenge of exploring a variety of alternatives and yet two people working for less than 1 year managed to develop this method, to test it, and to prepare a report on it.
 Page 115       PREV PAGE       TOP OF DOC
    Chairman THOMAS. And to underscore the point, that is data that is already available to HCFA for them to create an internal comparison?
    Mr. SCANLON. That is correct.
    Chairman THOMAS. Yes.
    Mr. SCANLON. And we believe there is going to be an initial startup cost and then some time, but we are talking, I think, months as opposed to years, and that if you think about the savings potential that would be associated with this type of an adjustment or the equity potential in terms of locking in rates across different areas of the country and then trend them forward in the future, there is a real benefit to be gained from making that investment and fixing the rates today.
    Chairman THOMAS. Now, we also talked about the fact that there are some risk adjustors within the AAPCC applied to the use basis of age, gender, where they are in terms of income level, on the Medicaid status, institutional status, and so on. What percentage of the total cost is attributable to the adjustment of those risk factors? And Ms. Rosenblatt, if you have any information on this, that would help as well.
    Mr. SCANLON. When studies have looked at the variation in cost across individuals, those factors only account for less than 5 percent of the variation in cost across individuals. So they really are not up to the task of adjusting for differences in cost of individuals going into HMOs.
    Chairman THOMAS. So are you saying then that applying those adjustments, you have got a 3- to 5-percent impact on the dollar amount, that inside that base amount, there are other costs that could be pulled out if we use different kinds of risk assessments, or is that just inherent in the base as to where the geography is, as my colleague from Nebraska indicated the way in which people receive medical services there or believe they need them?
 Page 116       PREV PAGE       TOP OF DOC
    Mr. SCANLON. The method we have worked on is designed at improving the average rate that will be received by HMOs in the county. The risk adjustors that are currently being used then account for differences in the demographic mix of people entering HMOs in every county, and risk adjustors of the type that Drs. Newhouse and Wilensky talked about in terms of using diagnostic information or administrative information on service use would further improve the rates being paid to HMOs in a given county. These things are all compatible to be used jointly and will ultimately end up targeting rates to be more appropriate for the mix that an HMO has.
    Chairman THOMAS. As best you are able, given the state of the art or the science, if we had a really aggressive approach to putting in known risk adjustment factors, what percentage of the amount currently paid would be attributable to risk adjustment factors above the current 3 to 5 percent, as you indicated? How high could we go in a realistic risk adjustment model?
    Mr. SCANLON. Well, I think that in terms of our method, we have identified that about 25 percent of the excess payment, which works out to be about 4 1/2 percent of overall payments, would be affected. If we use some of the risk adjustors that Drs. Wilensky and Newhouse talked about, I think we would add a couple of percentage points perhaps to the amount of reduction in rates, but a significant share, maybe 25 to 40 percent of the excess payment, might also be eliminated. I have to say that I am now out on a limb, that I do not know if they have more precise estimates of that, and I would certainly defer to them for those estimates.
    Chairman THOMAS. It would be fair to say, though, that if pushed the science, we could look at a one-fourth to a one-third adjustment factor between areas reflecting true risk adjustment factors in those areas rather than the 3 to 5 percent that we have now?
 Page 117       PREV PAGE       TOP OF DOC
    Mr. SCANLON. I think we could definitely reduce the excess payments by that amount, and one of the things we have tried to do is be conservative about this adjustment method.
    Chairman THOMAS. Yes.
    Mr. SCANLON. So that we are not trying to take it to the extreme and be at the risk of underpaying HMOs. The other thing about our method that——
    Chairman THOMAS. Nor would we want to push the envelope. But we also want to be smart buyers and pay for what we are getting.
    Mr. SCANLON. That is correct.
    Chairman THOMAS. In your testimony, you looked at those 12 counties when you were creating your model with large HMO enrollment, and you estimated that the excess payments were in a range of less than 1 to 6.6 percent. Well, if you have got that swing in the sample that you looked at, how comfortable are you with the President's proposal that you move to an across-the-board percentage cut?
    Mr. SCANLON. We are not comfortable at all with the idea of an across-the-board cut because what this work indicates is that the payment differences due to adverse selection are quite different across counties. The bulk of counties in this country would not be affected at all by the method that we propose because their HMO enrollment at this point is low and their rates are not biased by the type of phenomenon that we have identified.
    Chairman THOMAS. So you would go to where the action is, in essence?
    Mr. SCANLON. We believe it is much better to target these kinds of adjustments.
    Chairman THOMAS. Ms. Rosenblatt, we tried to put together, as you know, in the 104th Congress, a restructuring of the AAPCC. A number of us think that frankly our work product was better than the President's, and we are going to try to improve on the President's. One of the things that surprised me, though, was that everyone knows the problem with the county-based structure, but when we tried to move to a metropolitan statistical area, we saw aberrations in different directions. Number one, the President basically utilizes this blending structure national to county. And what we are trying to do is move away from the county but using an aggregate of the county, and is there anything other than the county that you think might be an appropriate way to approach this or is the way we have been looking at it and the President's plan now begins to focus on it, trying to move individual counties toward an aggregate county structure? You understand my question? Because it is still based upon that county unit that really is not a service area or a market area. It is just a handy political unit.
 Page 118       PREV PAGE       TOP OF DOC
    Ms. ROSENBLATT. I can comment on that because actuaries are constantly running into similar problems when they do rating. When you set premium rates, the issue is on what geographic entity are you setting a premium rate on. Some insurance companies have gone to a zip code methodology for that, but even that is not entirely accurate. A blending approach is an approach used by many actuaries. We call it a credibility approach, and often where we have an entity that is not large enough or does not have enough people in it to have what we call statistical reliability, we will blend that small entity with something larger.
    Chairman THOMAS. So blending for now seems to be an appropriate approach failing to get a better measuring unit?
    Ms. ROSENBLATT. It could be if the problem is caused by statistical aberrations in counties insuring small numbers of people.
    Chairman THOMAS. Hopefully, this is a slow pitch over the plate for an actuary now. The President has talked about reducing 34 billion and projects increased enrollments in the HMO plans notwithstanding the fact that he is reducing them. Does that seem reasonable to you? Is that an appropriate analysis or should we sharpen our pencils and be concerned about those kinds of reductions impacting on the argument for continued increase enrollment in HMOs?
    Ms. ROSENBLATT. Unfortunately, I cannot comment on that. I have not seen any of the underlying assumptions or methodology in the 34 billion.
    Chairman THOMAS. Neither have we, and we are anxiously looking for it so that we can get some comfort level about what is going on. I appreciate your willingness to indicate that until you get the facts you are not going to make a statement. That rarely bothers those of us in this business, but we would very much like to take a look at what the President's plan is rather than what the President's plan is according to folks telling us what the President's plan is. So, I look forward with you to getting the data and then I will ask that question of you. Does the gentleman from California wish to inquire?
 Page 119       PREV PAGE       TOP OF DOC
    Mr. STARK. Is it true, Mr. Chairman, that in the marriage vows, actuaries are allowed to respond ''I might''? [Laughter.]
    Chairman THOMAS. That is better than auditors. They start in on a long critique.
    Mr. STARK. Dr. Scanlon, are you telling us that basically out of $6 billion that we spent in California in 1995 on risk HMOs, $1 billion of that $6 billion was excessive? I do not want to say wasted, but more than we should have paid?
    Mr. SCANLON. That is the estimate of total excess payments including both concern about a county rate, or county average, that is being biased, as well as the risk selection that is going on in terms of healthier people on average enrolling in the HMOs, and that proportion is consistent with what other studies have shown to be the amount of the favorable selection effect.
    Mr. STARK. Now, your plan would not get all that back. We might get $200 or $300 million of that back.
    Mr. SCANLON. That is correct because there is a risk that given what we know about how to adjust for health status in a risk adjustment process, that being too aggressive, you will underpay certain HMOs. As Dr. Wilensky indicated, an HMO might attract sicker people under some circumstances. We have encountered that in State Medicaid Programs where a Medicaid Program will have people with disabilities moving into managed care, and particular HMOs with certain specialists are attracting disproportionate numbers of people with quite significant problems.
    Mr. STARK. We are going to hear from HMOs concerned that the way to solve this problem, of course, if you are losing money, is to increase volume. Well, maybe you would like to comment on, as the enrollment grows, Does the problem get better or worse?
    Mr. SCANLON. Well, as we have seen in the range of information available to us, the problem gets worse. The largest enrollment penetration was a county with 36 percent of the Medicare beneficiaries in HMOs, and that county had one of the highest excess payments due to a county rate that was overestimated. It is a bit like——
 Page 120       PREV PAGE       TOP OF DOC
    Mr. STARK. Was that in the Chairman's district or mine?
    Mr. SCANLON. No. It was in neither of your districts fortunately.
    Mr. STARK. OK. Do you know what it is in Alameda County just out of curiosity?
    Mr. SCANLON. My recollection is that the overpayment due to the county rate was about 2 percent, a little bit less than 2 percent.
    Mr. STARK. In Alameda?
    Mr. SCANLON. Right.
    Mr. STARK. What was the worst county in California, highest, I guess, not worst?
    Mr. SCANLON. Well, the highest is in Los Angeles County where the penetration is the highest.
    Mr. STARK. There is a question of the regression to the mean. After a number of years, what happened to the usage by HMO enrollees over time?
    Mr. SCANLON. That has been a concern about this research that has looked at people's expenditures before they join an HMO and then using that as the conclusion to say that the HMO enrollees are healthier. HMOs rightly argue that when people join the HMO they might be healthier, but that there is a tendency for people to get sick over time, and that many of them remain in the HMO and are being served there. In order to take that into account, we examined a method of identifying how much regression to the mean there is or how much increase there is in cost of enrollees over time and discovered about one-half of the gap between prior use costs by HMO enrollees and fee-for-service costs was removed in the first year, and then in subsequent years the HMO enrollees' costs also increased but did not, in our time period that we were able to examine, reach the level of fee-for-service enrollees' costs.
 Page 121       PREV PAGE       TOP OF DOC
    Mr. STARK. You recently released a different report on fraud and administrative problems in Medicare, and you made a number of comments which I find enlightening and I agree with. HCFA's failure to better administer this program has perhaps been costing us a lot of money. Could you just review very briefly for the Subcommittee what you found in HCFA's administration of these HMOs and their willingness to investigate them? Can you give 2 minute's summary on that?
    Mr. SCANLON. Yes, I can. In that work, and that is a summary of work that has been done by the GAO over a number of years, we have had concerns along two dimensions. The first being HCFA's role as the certifier of participating HMOs and the process that involves HCFA annually reviewing the operations of HMO to certify that it is fit to be a part of the Medicare Program. One concern in that area has been that these reviews have tended to be only paper reviews. They have asked the HMOs to demonstrate that they do have a plan to conduct quality assurance and other essential activities. However, they do not look beyond that plan to see actually how it is being implemented, whether medical reviews are being done appropriately, whether the right samples of patients are being identified, and what actions are being taken when unfavorable findings are noted.
    The other aspect of this in terms of HCFA's oversight responsibilities is that when they do actually find problems, the mechanisms and the aggressiveness in terms of requiring corrections have been quite lacking. We found a particular HMO that had a repeated history of violations, serious violations of the conditions of participation, and despite that fact, the only thing that was ever done was that each time the HMO was asked to submit a corrective action plan. And once that plan was submitted and accepted, it was allowed to operate again, but then at the next inspection HCFA found that there were problems again.
    The other dimension that we report on in this report is the issue that Dr. Wilensky brought up where information can be a very powerful force in assuring quality and promoting quality. And what we found is that HCFA, unlike many private sector employers, has failed to provide information to beneficiaries about the HMO choices available in their area. As you are aware, the Federal Employee Benefit Program as well as many private sector employees give you the basics of every plan and some go further in giving you quality information about plans. HCFA does none of this.
 Page 122       PREV PAGE       TOP OF DOC
    We found that HCFA's own region 9 office actually put together some of this information comparing benefits as well as comparing disenrollment rates across different plans, and made that available to people that requested it. It is a feasible task. We did it, too. We looked at the information that was available in different markets, both in terms of benefit structures as well as disenrollments, and found remarkable differences. We found that disenrollment rates in a single market could vary from 4 percent for one plan to 30 percent for another plan.
    The HMOs themselves and the private employers recognize that disenrollment does not tell you the true and full story about quality, but they all think it is a pretty strong indicator, and we found that HMOs used disenrollment rates in their advertising when the rates were favorable to them. So we think they recognize that it says something about the quality of care being offered.
    Mr. STARK. Region 9 is in San Francisco; right?
    Mr. SCANLON. Right.
    Mr. STARK. It is my district. Would it be easy for a Member of Congress to take that data for his or her district and send it out in our newsletters? Heaven knows they are boring enough for most of our constituents to read anyway. Might as well load them up with something useful. Would it be a difficult task for us to get that information and put it into a meaningful format for our own constituents?
    Mr. SCANLON. I think that the data can be put into a meaningful format. The problem at this point is that they would have to be assembled first by HCFA. They were assembled for region 9 by the regional office. Then we assembled them from some markets to do some comparisons across the country, but they are not at this point readily available.
    Mr. STARK. Would it be a budget-buster to do that?
    Mr. SCANLON. No, it is not. Again, we devoted some resources to this activity, but we did not devote a huge amount of resources to this activity. It is something that I think provides so much benefit, potential benefit, that the expenditures of resources would be justified.
 Page 123       PREV PAGE       TOP OF DOC
    Mr. STARK. I thank you for indulging me, Mr. Chairman. Thank you very much.
    Chairman THOMAS. The Chairman wants to announce that the gentleman's views about newsletters are his own and do not necessarily reflect the management. [Laughter.]
    I guess we could assume that HCFA is part of a giant conspiracy and there are payoffs and therefore they do not do a good job of followup, management, and policing of HMOs. And some might jump to that conclusion. My own belief tends more to the direction reflected in HCFA's current reorganization plan, which is, in essence, bringing managed care within HCFA out of the cold and into the same building and realizing that they are going to have to move more in terms of hands-on management structure. Do you have any information and data confirming the conspiracy theory versus the neglected theory, and is the reorganization, if you are familiar with it, something that looks like it might bear fruit?
    Mr. SCANLON. I have no information about the conspiracy theory or the neglect theory.
    Chairman THOMAS. Are you denying you have information or are you truly not knowledgeable? See when you go down this path, it gets very, very murky.
    Mr. SCANLON. Right. I am truly not knowledgeable about the conspiracy theory.
    Chairman THOMAS. OK.
    Mr. SCANLON. In fairness to HCFA, I think we are dealing with a world in which the health care system is changing rapidly, and we do need to move rapidly to try and become prudent purchasers or improve our ability as prudent purchasers, and that there is a need on HCFA's part to devote considerable effort to that and to see where they can improve. In terms of the current reorganization, what we do know is that it does seem to bring managed care in and integrate it fully with the rest of HCFA.
 Page 124       PREV PAGE       TOP OF DOC
    We obtained some information about the reorganization, and we understand it is very much a work in progress, that while the major structure has been defined, the critical elements beneath that major structure have yet to be defined, and those are key to how effectively the integration of managed care will work as well as how, overall, HCFA will be able to perform in the future.
    Chairman THOMAS. But fundamental to that, it would seem to me that if Dr. Newhouse is correct, that managed care structure is fundamentally different than the fee-for-service, then you need fundamentally different measuring tools to determine whether or not they are performing their job, and my concern is that HCFA has not adopted measuring tools that seem to be fairly evident, nor have they been creative, as you folks have, in coming up with tools that within their own database would seem to provide at least some ability to be discernable beyond what they now have, and your job as usual, if you choose to accept the assignment, will be to continue to oversee what they do and how they do it. I appreciate your statement. I believe the gentlewoman from Connecticut would like to inquire.
    Mrs. JOHNSON. Thanks. I just have a couple of brief questions. When you say you could save 25 percent of overpayments and 4 percent across the board, are you assuming you will pay less for lower risk patients but pay more for higher risk patients?
    Mr. SCANLON. Our method would not involve an adjustment on a patient-by-patient basis. It would involve an adjustment in each county and so that if a county's HMOs have enrolled a disproportionate number of healthier or lower risk beneficiaries, then that county would have a lower rate being paid to it.
    Mrs. JOHNSON. In other words, you do not adjust for severity. You assume that is reflected in the cost figures for fee-for-service and the cost figures for HMO patients; correct?
    Mr. SCANLON. We do happen to adjust for severity by comparing the enrollees of HMOs and looking at their prior utilization costs, and then adjusting those costs for increases, and by looking at the HMO enrollees and looking at them in quite great detail in terms of their prior use, we feel we have a very good control for differences in health status and differences in expenditures that result from health status.
 Page 125       PREV PAGE       TOP OF DOC
    Mrs. JOHNSON. So do you convert HMO enrollees back into a fee-for-service system for purposes of analyzing their costs? You look at what services they have.
    Mr. SCANLON. We are not able to look at the services that an HMO enrollee receives after they actually join an HMO, and it would be inappropriate for us to use that as the basis for paying the HMO because after all we are asking or we are hoping is that people will join HMOs because the delivery of care is more efficient, and the amount of service, unnecessary service that is provided, is reduced. Our method, consistent with HCFA's current method, is designed to pay what Medicare pays for a comparable person in the fee-for-service sector, even though the amount of service in the HMO may be lower.
    Mrs. JOHNSON. In other words, comparable from the point of view of diagnosis or age group or, you know, comparable in what regard?
    Mr. SCANLON. We developed our method using prior utilization in terms of the amount of spending in the prior year. What that does is it combines a large variety——
    Mrs. JOHNSON. Prior year spending when they were in the fee-for-service system?
    Mr. SCANLON. When they were in the fee-for-service sector.
    Mrs. JOHNSON. So the old system looked only at the cost in fee-for-service. You are looking at the cost in fee-for-service plus a first year HMO enrollees based on their recent year in fee-for-service system?
    Mr. SCANLON. And then we matched those people with other people of similar characteristics and similar health use patterns in the prior year, and we followed them through time because we are very sensitive to the notion that people do get sicker over time, people joining HMOs may have postponed some of their service use, as sometimes happens in the fee-for-service sector as well.
 Page 126       PREV PAGE       TOP OF DOC
    Mrs. JOHNSON. But you look at the pattern of care of an individual in an HMO and his preceding year in fee-for-service?
    Mr. SCANLON. Right.
    Mrs. JOHNSON. And then you match him to someone who stays in fee-for-service?
    Mr. SCANLON. Right.
    Mrs. JOHNSON. And attribute the increased costs of the fee-for-service person to the HMO person?
    Mr. SCANLON. We ascribe that cost to the HMO person but that is because what we are doing is we are looking for how much that HMO enrollee would have cost us in the fee-for-service sector, and doing that is actually protecting the HMO, not penalizing them for being more efficient about the delivery of service that they are getting.
    Mrs. JOHNSON. Then basically the functional effect of your method is to treat everyone as if they were in fee-for-service because you look back to the last fee-for-service year, you pair them with someone in fee-for-service, and you assume their increased cost based on the fee-for-service person. Do I follow you?
    Mr. SCANLON. That is correct, but the underlying basis for HMO rates is what a person would have cost in the fee-for-service sector. When a person enrolls in the HMO, we know that their service utilization is going to change, and we think it is going to change in a positive direction.
    Mrs. JOHNSON. Yeah. Let me just try to get this clear before the red light goes off.
    Mr. SCANLON. Sure. OK.
    Mrs. JOHNSON. Because I thought this was a simple question, and it has turned out not be so simple. I thought the difference between your approach and HCFA's approach was that HCFA based the AAPCC on the cost of the fee-for-service group, and that you are including not only costs in fee-for-service but costs in HMO; is that the difference?
 Page 127       PREV PAGE       TOP OF DOC
    Mr. SCANLON. We are including an estimate of the cost of people in HMOs that they would have incurred if they had remained in fee-for-service.
    Mrs. JOHNSON. Basically your estimate is also fee-for-service based. Why is it so much lower? I assumed what you meant was as in prevailing wage. One of the problems in prevailing wage is you tend to get the wage of the unionized organizations and then nonunion people will not tell you. So you do not have a real view of what the average wage is. And what I heard you saying is we know what the costs are in fee-for-service. HMOs are supposed to save money. So you were going to average in the capitated rate for the HMO person, bringing it down to what must be the average cost. But if you are simply adjusting the HMO people so basically they are fee-for-service people, why do you get a 25-percent saving or a 4-percent saving across the board?
    Mr. SCANLON. Because what happens when healthier people enroll in the HMOs, the payment is based on the average person in the county joining an HMO, and what we have done is we have adjusted that average so that instead of computing the average with only the less healthy people who remain in fee-for-service, we include some of the healthier people that joined the HMO, and that lowers the average cost for the county and therefore reduces the payment that will be received by the HMO.
    Mrs. JOHNSON. Now in an elderly county, presumably, a county with a lot of over 85 seniors, presumably your costs would be higher then, and the AAPCC—have you done any study of some of the rural counties where you have a lot of elderly people who actually are not getting very good care because they do not have access?
    Mr. SCANLON. That was not something.
    Mrs. JOHNSON. Showing the AAPCC is too low?
    Mr. SCANLON. Right. That was not something that we looked at here. In fact, the method that we have developed would not affect the rates being paid in those rural counties because the problem that we are dealing with only occurs when there is a significant number of people enrolling in HMOs and the net effect when you use only fee-for-service people to estimate the average is that you are potentially biased either upward or downward. What we think of as an advantage of the method that we have developed is that if the HMOs in the county have adverse selection, that is they are getting sicker people on average, then our method would raise the rate being paid those HMOs over what HCFA's current method does for those HMOs.
 Page 128       PREV PAGE       TOP OF DOC
    Mrs. JOHNSON. Well, I think if you are not able to take into account where services are not being provided, then I do not think we are getting a true picture of savings. I gather that you do not take into account the fact that HMO people have access to more medical services than non-Medicare people? Are you taking that into account?
    Mr. SCANLON. In terms of the first part of your question, we have not attempted to deal with the question of the underservice that may be in particular counties. That is definitely another part of the problem, and we recognize the proposals to raise the floor rate deal with both that and with some of the difficulties of operating HMOs in less populated areas, and dealing with both of those issues is something we think is worth considering. How much money is required to be able to overcome those problems, we have not done the work to identify that.
    Mrs. JOHNSON. Just last because my time has expired, but have you done any study of Medicare HMO payments versus premiums for a comparable benefit plan adjusted for age? Do we know anything about whether HMO premiums are high or low for the benefit plan offered in the market they are in adjusting it for age. Of course, the uniqueness for the Medicare HMOs is that you are all over 65.
    Mr. SCANLON. Right.
    Mrs. JOHNSON. Do we know anything about these premium comparisons?
    Mr. SCANLON. We have not done any work like that, and I am not aware of any work that has done that.
    Mrs. JOHNSON. Are you capable of doing that type of work?
    Mr. SCANLON. We would need to think about how to approach that work. As you noted, the important difference between people receiving insurance in the under 65 population versus the Medicare Program is the large increase in utilization associated with age. This would be an interesting task to undertake. We will certainly consider it.
 Page 129       PREV PAGE       TOP OF DOC
    Mrs. JOHNSON. I am very interested in that. If you would get back to me on that.
    Mr. SCANLON. We will.
    Mrs. JOHNSON. I think it is extremely important especially now with some of the HMO plans offering the comprehensive coverage that often employer provided plans do cover including some prescription drugs. So it would be very useful for us to know is this plan that looks like the average employer plan in this market, when adjusted for age, costing us more or less because if we are going to keep track of whether reimbursements are honest or dishonest or reasonable or unreasonable, we need to at least know how they compare to the private sector where the competition is intense and the bargaining is very intense, and that substitutes for price controls and has substituted for price controls more effectively and done a more effective job than price controls in public sector. So I would like you to think that over and see if we could develop a plan to do that kind of study.
    Mr. SCANLON. We definitely will and we will contact your staff.
    Mrs. JOHNSON. Thank you.
    Mr. SCANLON. Thank you.
    Chairman THOMAS. Just so I think I understand what you did based upon the discussion with the gentlewoman from Connecticut, part of the concern, I think, was that we were talking about trying to compare quality versus price, and you were just focusing on price, and your concern was that since HMOs are reimbursed on the adjusted average per capita cost and if there is an argument that HMOs for whatever reason wind up with a selection process that removes the more healthy folk from that larger universe, that when you take the fee-for-service on the adjusted average per capita cost, you have a lower health profile than you otherwise would. And so what you have tried to do is figure out what it would look like for a true adjusted average by creating a methodology which would, in essence, include those who went into the HMO Program.
 Page 130       PREV PAGE       TOP OF DOC
    The way you have done that is take a look at them before they go in, take a look at them after they go in, find someone who matches them had they stayed in the fee-for-service and blend that out?
    Mr. SCANLON. That is correct.
    Chairman THOMAS. And what you have come up with is that on average the adjusted average—and obviously to the degree more people go into managed care programs, the more skewed the adjusted average becomes, and it is offset slightly by an area which would have a higher HMO enrollment. You would tend to have, I would think, a degree of competition among HMO and that would have to in part work itself out, but the data you have used to indicate that there is a clear potential for a downward adjustment of payment to HMOs is data that is already available to HCFA. They simply have not looked at the before or after or maybe even understood the fundamental methodological flaw of assuming the adjusted average was a true average, notwithstanding the fact they argue because they want to reduce the payments to HMOs that HMOs are pulling out of that group, healthier people.
    Mr. SCANLON. Yeah. When HCFA began this program, there were few beneficiaries in HMOs to speak of and so the average was not distorted by a very great amount if at all. As time has gone on and there have been counties that have had much greater penetration of HMOs, then what we see is more significant distortions in the average, and so this is what we are trying to correct for at this point.
    Chairman THOMAS. And you can choose to answer this question or not. Do you believe that HCFA is engaged in a conspiracy of not informing people about their choices of moving into HMOs to allow them to maintain a model which more accurately reflects the universe because they do not have folks signing up, or is this just a failure to do what ordinary people would do in terms of letting people know their options and realizing that as people choose options, it changes the mix, as it was before, and they need to get better risk adjustors to be able to make the changes? A or B?
 Page 131       PREV PAGE       TOP OF DOC
    Mr. SCANLON. Do I have a choice about answering the question?
    Chairman THOMAS. I wound up giving you a compound choice within that structure. And the reason I am continuing to do this is that there are frankly people who believe A frequently is the reason. And what we need to do, especially with the President's basic structure, is to with whatever tools you are able to provide us with, as soon as possible, begin making these internal adjustments because no one wants to pay more than what we should for the services being given. It is absolutely legitimate to discuss whether or not that amount if fair, but it certainly is not if we are paying more than we should. And I want to thank both of you, and we will obviously fall back on your resources as we move forward and we may fall back on your resources right now.
    The gentlewoman from Connecticut.
    Mrs. JOHNSON. The other question that I did not get to that I really wanted to ask you guys—I am not going to ask you to answer it, but maybe you can provide me some information—the preceding panel and your panel have assumed that HMOs tend to select the healthier people. When HCFA did an extensive study of Medicare Select and who went into Medicare Select Programs, which was the first way that seniors could get into HMOs, and they dragged the study out because they did not support my bill and they did not want me to pass it, and the study finally came out showing that they do not select the healthier people, that, in fact, there was no bias. And when I talked to my seniors about who is looking at HMOs, it is the ones who need prescription medications. They are the first group that want to look at these new plans because Medicare does not cover HMOs.
    So it seems to me that the people with the most illness tend to be the first to be interested in plans that offer some prescription coverage and a little more comprehensive approach. So logically, it does not make sense to me that they would select the healthiest, and yet that is the assumption of everyone, and the Medicare Select study demonstrated, at least in that instance, it was not the healthier ones. So if you would get back to me on that, I would appreciate it.
 Page 132       PREV PAGE       TOP OF DOC
    Mr. SCANLON. We would be happy to.
    [The following was subsequently received:]

    Response: Several features of HMOs should attract less healthy beneficiaries and, other things equal, result in ''adverse selection''—the tendency of enrollees to be less healthy than the average beneficiary. HMOs' prescription drug benefits would be expected to attract those beneficiaries who need prescription drugs and who often have a chronic condition. Similarly appealing features of HMOs—coordination of care, low copayments, and the absence of deductibles—would also be expected to attract beneficiaries who anticipate needing substantial amounts of medical care.
    Factors discouraging less healthy beneficiaries or attracting healthier ones, however, may offset such factors as drug benefits that attract less healthy beneficiaries to HMOs. Less healthy beneficiaries may be reluctant to join an HMO for several reasons: they may have formed a close relationship with a physician who is not part of the HMO network, prefer having greater choice in selecting a physician or hospital, or fear that the HMO might not authorize payment or provision of certain medical services or procedures. In contrast, beneficiaries in good health may be attracted to HMOs because they value more the monetary benefits that HMOs often provide—zero premiums and low deductibles and copayments—as well as extra benefits ranging from eyeglasses to health club memberships.
    Although a HCFA-sponsored study of Medicare SELECT and a few studies of Medicare HMOs have found that managed care enrollees are not healthier than the average beneficiary, the weight of evidence indicates that the factors attracting the less healthy to Medicare HMOs are weaker than those discouraging them from joining or encouraging the more healthy to enroll.(see footnote 1) In reviewing the research on beneficiaries' health status, we found that Medicare HMO enrollees are healthier, on average, than beneficiaries who remain in fee-for-service. The better health status of the typical HMO enrollee is reflected in estimates of the cost differences between HMO and fee-for-service beneficiaries:
 Page 133       PREV PAGE       TOP OF DOC

    •  Compared to the average beneficiary, new HMO enrollees have much lower costs when measured by their Medicare costs while still in fee-for-service.
    Since HMOs' management of care aims to reduce utilization by eliminating unnecessary services, cost comparisons between beneficiaries in an HMO and those in fee-for-service would not measure differences in health status. Consequently, researchers have looked at the amount of Medicare spending per beneficiary in the year before he or she joined an HMO. Many studies have found that enrollees' prior-year costs were much lower than comparable fee-for-service beneficiaries' costs. For example, the Physician Payment Review Commission found that new enrollees' prior-year costs were 63 percent of the average fee-for-service beneficiary's costs.(see footnote 2) Similarly, we found that new Medicare HMO enrollees in California had prior-year costs that were roughly 62 percent of comparable fee-for-service beneficiaries.(see footnote 3)

    •   All enrollees' costs—although higher than new enrollees' costs—are lower than the costs of the average beneficiary. An enrollee's costs tend to rise toward the average beneficiary's cost the longer the person remains in an HMO. We adjusted for this tendency to regress toward the mean but found that it did not eliminate the cost gap between HMO enrollees and fee-for-service beneficiaries. For example, we estimated that for enrollees who have been in an HMO for 3 years, costs are still 10 percent less than the average fee-for-service beneficiary's.(see footnote 4) Using a different method, Mathematica Policy Research also found that HMO enrollees' costs were about 11 percent below the costs of comparable fee-for-service beneficiaries.

 Page 134       PREV PAGE       TOP OF DOC
    In addition, the following indicators reflect more directly the better health status of HMO enrollees:
    •  Self-reported health status: This widely used measure of health status captures beneficiaries' perceptions of their own health.
    •  Compared with fee-for-service beneficiaries, Medicare HMO enrollees typically have a better self-reported health status.(see footnote 5) Among HMO enrollees, 22.9 percent reported being in excellent health, compared with 17.5 percent of fee-for-service beneficiaries. Similarly, 30.5 percent of HMO enrollees reported being in very good health, compared with 24.4 percent of fee-for-service beneficiaries.

    •  By contrast, HMO enrollees are significantly less likely than fee-for-service beneficiaries to report being in poor or fair health. For example, only 4.5 percent of HMO enrollees said they were in poor health, compared with 8.2 percent of fee-for-service beneficiaries.
    •  Chronic conditions: Medicare HMO enrollees are less likely to report having heart disease or several other chronic conditions.(see footnote 6) For example, 38.1 percent of fee-for-service beneficiaries said they have a heart condition, compared with 31.3 percent of HMO enrollees. Likewise, 31.0 percent of fee-for-service beneficiaries report having five or more conditions, compared with 25.3 percent of HMO enrollees.(see footnote 7) Finally, 7.7 percent of fee-for-service beneficiaries have mental disorders, compared with 3.6 percent of HMO enrollees. (Although a smaller proportion of HMO enrollees than fee-for-service beneficiaries say they have other conditions, such as emphysema, the differences are not statistically significant.)

    •  Mortality rates: In a given year, Medicare HMO enrollees are less likely to die. Compared with Medicare fee-for-service beneficiaries, Medicare HMO enrollees have lower annual mortality rates. For example, we found that in 1994, fee-for-service beneficiaries in California had a death rate of 5.2 per 100, compared with HMO enrollees' rate of 3.5 per 100.(see footnote 8)
 Page 135       PREV PAGE       TOP OF DOC

    •  People of the same sex and the same age category display the same pattern. For example, California men 85 and older who remain in the fee-for-service sector died at a rate of 17.8 per 100 in 1994, compared to a rate of 12.7 per 100 men of the same age enrolled in an HMO. It is unlikely that this pattern predominantly reflects differences in the quality of care between HMOs and fee-for-service.
    •  Activities of daily living: HMO enrollees were less likely to report difficulty with walking and other activities of daily living (ADL). For example, 21.4 percent of HMO enrollees said they had difficulty walking, compared with 27.8 percent of fee-for-service beneficiaries. The differences between the two groups of beneficiaries were statistically significant for the other five ADLs as well. Moreover, the same pattern holds for ''instrumental activities of daily living'' (IADL), such as preparing meals and paying bills.

    These data can be viewed from two perspectives. First, they reveal differences between HMO enrollees and fee-for-service beneficiaries on average. (HMO enrollees' health status is typically better.) These differences help explain why Medicare pays HMOs more than necessary. Second, these data also imply similarities between the Medicare HMO and Medicare fee-for-service populations. In both, substantial numbers of beneficiaries have chronic diseases and other signs of poorer health. For example, many HMO enrollees as well as fee-for-service beneficiaries have diabetes. Similarly, although the average new HMO enrollee had prior-year expenditures roughly 40 percent lower than the fee-for-service beneficiaries' average, some new HMO enrollees had prior-year expenditures that were considerably higher than the average HMO enrollee's.
    In summary, Medicare HMOs' distinctive features, such as a prescription drug benefit, do attract beneficiaries who are less healthy than the average. That fact is important for painting a fair and accurate portrait of Medicare HMOs. However, the relatively healthy form a higher proportion of HMO enrollees than of beneficiaries remaining in fee-for-service. That difference is critical to understanding why Medicare's method of paying HMOs overcompensates them.
 Page 136       PREV PAGE       TOP OF DOC


    Mrs. JOHNSON. Thank you.
    Chairman THOMAS. Thank you very much. Thank the panel. I look forward to using you as a resource in the future. May we have the next panel, please, which happens to be the last panel. We'll get some real world experience here. Stephen Wiggins, chairman and chief executive office, Oxford Health Plans, Inc., Norwalk, Connecticut; Jack Myers, vice president of Financial Services, Group Health Cooperative of Puget Sound, Seattle, Washington; and Craig Schub, senior vice president, PacifiCare Health Systems, Cypress, California.
    I would say to all of you that we have your written testimony. It will be made a part of the record, and you will have 5 minutes or thereabouts to inform us in any way you see fit to enlighten this panel. Might as well begin with Mr. Wiggins, the way you are structured and move across the panel.


    Mr. WIGGINS. OK. Thank you very much, Mr. Chairman. I would like to begin my testimony by making one point on an earlier reference to medical loss ratios. Our Medicare medical loss ratio is 86.7 percent, and we have a 1-percent net income in Medicare, just to set the record straight.
    First of all, we think more choice is needed in the Medicare Program. We think savings are demonstrable. It is a very simple logic. HMO costs are going up 3 percent versus the Medicare Program that is going up 8 percent. With that math, it is fairly easy to get to the President's target of $138 billion just through those savings.
 Page 137       PREV PAGE       TOP OF DOC
    One of the things we believe would happen with lower payments to HMOs is it will act as a tax on people who have joined HMOs. People joining HMOs tend to be lower income senior citizens who are least able to afford the cuts in benefits and the implicit tax of such cuts. These benefit cuts would result in fewer beneficiaries joining health plans because HMOs would become less attractive than the normal fee-for-service program.
    The reduced savings to the government would accrue as a consequence of this, and for that reason we think that 34 percent of the President's budget cuts resulting from 11 percent of the beneficiaries is too much.
    The second point is that we do not believe HMOs have healthier enrollees. I have given you some information from New York that shows HMOs actually have sicker hospital patients than the normal fee-for-service program. Studies based on self-reported health status, which were the basis of what prior testimony was referencing, do not reflect the regression to the mean that Congresswoman Johnson mentioned. The longer someone is in an HMO, the more likely they are to have the same health status as the normal fee-for-service system. In places like New York, we tend to be getting equivalent or sicker members, and we have given you information that actually shows the hospital case mix index where we have taken all Medicare fee-for-service patients and compared them to all HMO patients; and the HMOs have sicker patients. This is publicly available data that has been organized by Health Share Technologies.
    We have requested data from HCFA to compare the actual Medicare fee-for-service utilization using risk adjustors. We think there are good risk adjustors in the marketplace. We have not been able to get information from HCFA that allows us to risk adjust the Medicare fee-for-service population and compare that to HMO enrollees. We think such analysis will finally be the marker that proves that HMOs are not enrolling healthier members. As you may recall, 10 years ago, this was the same thing employers used to say about HMOs; that we were enrolling the young and the healthy. That is no longer an argument that anyone makes in the private sector as more people moved into health plans.
 Page 138       PREV PAGE       TOP OF DOC
    I would like to close with just a few points on DME and DSH—the carve-out idea. We have first-hand experience with the DME and DSH carve-out in New York where in both Medicaid and the commercial section this carve-out occurred. It is a very inflationary policy. In New York it resulted in a 3- to 12-percent increase in overall health care spending, and the reason for this is that there is double dipping that occurs. I have given you some other information that shows as a consequence of cutting out DME and DSH, you end up with a 3- to a 12-percent increase in health spending.
    We also pay community hospitals right now significantly less than teaching hospitals. We have a 20-percent premium paid to the teaching hospitals. So clearly the AAPCC dollars are being passed down to the teaching hospitals in the form of higher reimbursements.
    And then finally I just believe that this whole idea is really killing the goose that is laying the golden egg. We need to give people incentives to join HMOs, not create disincentives. There are some things that need to be addressed, but there are good people running health plans, there are good programs being put together. Beneficiaries generally like HMOs, and we need to figure out ways not to take from those beneficiaries that were the first to join, but figure out a way to get more people in HMOs.
    Thank you.
    [The prepared statement and attachments follow:]

Statement of Stephen F. Wiggins, Chairman and Chief Executive Officer, Oxford Health Plans, Inc.

    Mr. Chairman. Thank you for the opportunity to come before you today to discuss the impact of the Administration's Fiscal Year (FY) 1998 budget proposal on Medicare Risk Contract plans, such as Oxford Health Plans, and on the growing number of beneficiaries who enroll in plans such as ours. I am Stephen F. Wiggins, Chairman and CEO of Oxford Health Plans, Inc.
 Page 139       PREV PAGE       TOP OF DOC
    Oxford Health Plans currently provides healthcare to over 130,000 Medicare beneficiaries in New York, New Jersey and Connecticut. We are the largest Medicare HMO plan in New York, and one of the fastest growing Medicare HMOs nationwide. We hope to expand to new markets in the next several years.
    In addition to providing traditional healthcare services, Oxford provides such value added services as disease management, complex case management, and wellness programs. Our membership is sophisticated and demands the best healthcare available. This is especially true of our Medicare members and I am proud to note that our satisfaction rate among our seniors is 96 percent. In addition, according to a nationwide HCFA survey our voluntary disenrollment rate is 6.9 percent as compared to a national average of 11.4 percent.
    While I commend the Administration for addressing the issue of maintaining the solvency of the Medicare program, I am concerned over the impact that the Administration's Medicare proposal will have on the benefits and services provided to the 4.9 million Medicare beneficiaries who are enrolled in Medicare HMO plans nationwide. Although growing rapidly, only 11 percent of all Medicare beneficiaries are enrolled in Medicare Risk HMO plans, yet the Administration seeks to exact a full one-third of the proposed Medicare reductions from these beneficiaries- $46 billion over six years. While these reductions are technically imposed on HMOs, they will be felt directly by the beneficiaries.
    We support looking for new and innovative ways to reform this increasingly outdated program and to protect its financial viability. In fact, we are working with both the American Association of Health Plans and the Healthcare Leadership Council to meet this goal. Expanding choice and the plan options provided to Medicare beneficiaries is vital to the long-term viability of the Medicare program. Payment policies that would hinder the development of new plan options not only hurts seniors, but will result in fewer overall savings for the Medicare program.
    The Administration's proposal acknowledges the need for new plan options, though the reimbursement structure for existing Medicare alternatives will force plans to reduce and even eliminate the benefits upon which seniors have grown to rely. As a result, seniors will have access to fewer plan options, fewer benefits, and will be forced to pay higher out-of-pocket costs. The end results are reduced enrollment and reduced budget savings.
 Page 140       PREV PAGE       TOP OF DOC
    Most Medicare HMO plans compete for enrollment by offering Medicare beneficiaries benefits that are additional to those provided under traditional Medicare, such as prescription drug coverage, preventive health screening, dental care and vision care. In addition, all Oxford individual Medicare products offer zero premiums and minimal cost sharing.
    Seniors, particularly low-income seniors, find our products an attractive, low-cost alternative to traditional Medicare which has deductibles and coinsurance for all services, and to costly Medigap coverage which is necessary to fill in the gaps left by traditional Medicare. A recent survey conducted by the American Association of Health Plans (AAHP) reveals that beneficiaries in Medicare HMOs are more likely to have moderate-to low-incomes. In addition, studies indicate that the highest growth rate in Medicare HMOs is among disabled individuals (Physician Payment Review Commission, Access to Care in Medicare Managed Care, November 1996) and seniors age 85 and older (Welch study published in Health Affairs, Fall 1996).
    As you are aware, the Administration's proposal would significantly reduce our reimbursement rate by eliminating the graduate medical education (GME) and disproportionate share hospital (DSH) payment components of the AAPCC and by reducing our overall reimbursement level to 90 percent of the AAPCC from its current level of 95 percent. In addition, because the AAPCC is calculated based on Medicare's fee-for-service spending, the President's proposed reductions for traditional Medicare will result in a corresponding reduction for Medicare HMOs. Thus, not only does the Administration proposal reduce our reimbursement to 90 percent of the AAPCC over time, but it also proposes that we are reimbursed 90 percent of an overall lower AAPCC level.
    Taken together, the Administration's current proposal will force plans to either eliminate all of the additional benefits we currently offer our members or increase the senior's out-of-pocket costs for these services by charging premiums and raising co-payments. In fact, last year Oxford was forced to reduce the prescription drug benefit provided to some of our New York area seniors from $1,000 to $500 due to increased medical costs and a non-commensurate increase in reimbursement rates.
 Page 141       PREV PAGE       TOP OF DOC
    While Oxford is concerned about the total impact of the proposed reductions on Medicare beneficiaries, we are particularly concerned about the Administration's plan to remove GME and DSH directly from the AAPCC at the regional level. Administration officials have recently testified that these reductions represent an average reduction of 5.3% nationally, though in some counties health plans will experience up to a 20 percent reduction in their current reimbursement rates (ProPac 1991).
    Not only does Oxford routinely contract with teaching hospitals in the New York area, but the rates we negotiate are on average about 20 percent higher than rates with other hospitals. In addition, our Medicare members are more frequently admitted to teaching hospitals. By removing the GME component of the AAPCC, the Administration would provide a disincentive for Medicare HMOs to contract with teaching hospitals by failing to recognize and compensate Medicare HMOs for the higher costs associated with doing business with these institutions.
    Our New York experience in this area offers an important insight into the consequences of carving out GME. As you may know, New York State has participated over the years in the funding of GME. However, as part of an overall hospital deregulation plan implemented this year, teaching hospitals will receive GME funding directly from the State created GME pool. The pool is funded by payors through an assessment on the number of insured residents covered by each payor.
    The teaching hospitals are allowed to ''double dip'' by receiving funding for GME directly from the pool while continuing to receive reimbursement from payors at levels higher than what we pay non-teaching facilities. The State's response was that payors should be able to negotiate a corresponding rate reduction with teaching hospitals.
    In theory this makes sense, but in practice it does not work and in fact has been inflationary. While Oxford has been more successful than our competition with re-negotiating contracts with the teaching facilities, we nevertheless had to increase our commercial premiums this year by 3 percent to offset the difference between what we pay into the pool and what we were able to achieve in contract savings with the teaching hospitals. The consequences of a GME carveout in Medicare, as previously stated, are reduced benefits and higher out of pocket expenses for seniors.
 Page 142       PREV PAGE       TOP OF DOC
    Clearly, Oxford supports the efforts of teaching hospitals. We agree that graduate medical education is essentially a public good and, therefore, should be funded through a broad-based, general revenues mechanism. In 1995, Congress supported a general revenues trust fund for GME, funded at that time through the phasing out of a corporate tax credit. In addition, under that proposal Medicare HMOs continued to be compensated for the higher costs associated with contracting with teaching hospitals. We continue to support this type of solution to GME funding.
    Claims that Medicare HMOs are overpaid because they enroll healthier and younger seniors is either based on out-dated information or can be challenged by the results of other studies, and thus, clearly should not be the basis for making changes to the Medicare HMO payment policies. Studies by HCFA researchers Beebe and Riley show that any initial favorable selection is corrected by regression to the mean . In essence, over time a group that starts out healthier and less costly than average will move to an average level of health and cost.
    This argument is indeed the case for Oxford Medicare Advantage. When we compare the severity of illness of the Oxford Medicare Advantage population to the FFS Medicare population in our region, we find that our members have similar if not worse health status than the FFS population. On a larger scale, we find that in New York State the entire Medicare HMO population has a case mix index which is more complex than the general Medicare population. In addition, when we compare the average age of Oxford Medicare Advantage members to FFS beneficiaries in Oxford's New York service area, we find that the average ages of both populations are only months apart. This demonstrates again that our membership mix mirrors that of the general Medicare beneficiary population.
    Furthermore, studies performed by economists Laurence Baker from Stanford and Pete Welch from the Urban Institute find that increases in HMO market share are associated with a reduction in total Medicare spending levels. This is known as the spillover effect, and occurs as physician practice patterns become alligned with managed care. Baker found at estimated 1994 spending and enrollment levels, if the expenditure reduction associated with a ten percent increase in HMO market share were obtained for all Medicare beneficiaries currently receiving care in the FFS sector, this component of the Medicare program would realize savings of approximately $2.2 billion. To further support this point, a study by Welch shows that for every 10 percentage points of HMO market share, Medicare expenditures decrease per beneficiary in an MSA by 1.2 percent in the short run and 3.9 percent in the long run. In markets with large managed care penetration these numbers are even more significant.
 Page 143       PREV PAGE       TOP OF DOC
    In summary, Oxford would like to continue to grow and expand in the Medicare market. In order to successfully enroll and provide healthcare services to Medicare beneficiaries it is imperative that we are paid a fair premium to cover our costs and to prevent large out-of-pocket costs for beneficiaries. We believe that reform to the Medicare system is necessary and we will work with you and the Administration to help achieve meaningful change. The current proposal outlined by the Administration, however, will have a significant impact on the beneficiaries and will hinder the growth of Medicare managed care. Specifically, CBO's estimates of growth in Medicare managed care enrollment from 11.7 percent in FY 1997 to 22 percent in FY 2002 will be unattainable.
    Thank you Mr. Chairman for this opportunity.


    Beebe, ''Medicare Reimbursement and Regression to the Mean,'' Health Care Financing Review, Spring 1988.
    Riley, ''Biased Selection and Regression Towards the Mean in Three Medicare HMO Demonstrations: A Survival Analysis of Enrollees and Disenrollees,'' Medical Care, April 1989.
    Baker, Laurence, ''Managed Care, Health Expenditures and Physician Income,'' Princeton University Dissertation, 1994
    Welch, Pete, ''HMO Market Share and its Effect on Local Medicare Costs,'' Urban Institute, 1991
    Physician Payment Review Commission, Access to Care in Medicare Managed Care: Results from a 1996 Survey of Enrollees and Disenrollees, November, 1996
 Page 144       PREV PAGE       TOP OF DOC
    [The official Committee record contains additional material here.]


    Chairman THOMAS. Thank you, Mr. Wiggins.
    Mr. Myers.


    Mr. MYERS. Good afternoon, Mr. Chairman and Members of the Ways and Means Subcommittee on Health. On behalf of Group Health Cooperative of Puget Sound and our 55,000 Medicare beneficiaries, I appreciate being able to appear before you to share some thoughts about Medicare risk contract reimbursement. Let me first express our gratitude, Mr. Chairman, for addressing the issue of wide variation in Medicare risk contract payments in the Balanced Budget Act of 1995. We understand that the President's recently released budget may also address these issues. Group Health strongly supports the reform efforts of the Congress and administration to assure both the short-term and long-term solvency of the Medicare Program.
    Managed care can help assure that solvency. It coordinates care including prevention to meet the total range of health needs for individual, family, and community. Managed care done well can result in more affordable health care. And studies show that most seniors enrolled in health maintenance organizations find no barriers to care and are satisfied with their health plans. In fact, Group Health year after year has one of the lowest complaint rates in the country as measured by the Health Care Financing Administration.
 Page 145       PREV PAGE       TOP OF DOC
    Group Health has participated in the Medicare Program since it began in 1965. In 1976, we were the first risk contractor in the Nation. We currently are the largest risk contractor in the State of Washington. We believe in risk contracting because it is consistent with our principle of holding managed care organizations accountable for both quality and cost management. For the sake of the good health of our seniors, we would like to stay in the Medicare business.
    Here is the problem. The Seattle area is very efficient with health care costs well below the national average. Risk contract payments are based on 95 percent of the area costs. For Group Health across its various counties, this results in risk payment rates ranging from 8 to 25 percent below the national average. With these rates so significantly below the national average, we barely break even financially. We are but one example of the geographic variation that exists in the calculation of risk contract payments. Throughout the country, there is uneven reimbursement for risk contractors and uneven benefits for seniors.
    In a few areas of the country, risk contractors realize sizable financial margins. This allows them to offer their seniors additional benefits like prescription drugs at little or no charge. But in most areas of the country, health plans cannot afford to offer extra benefits even though their seniors pay the same to Medicare as those who reap extra benefits.
    In terms of dollars, Medicare monthly reimbursement to health plans ranges from a low of $221 per senior to a high of $767 per senior, a huge difference. Furthermore, low reimbursement counties tend to get lower rates of increase from year to year. Rate increases for Group Health have been 2 percent below the national average for several years. This is the paradox. Cost efficient communities receive low rate increases while higher priced inefficient communities receive reimbursements well above the average.
    Proposed reductions in payments will only exacerbate these inequities. Because Group Health supports Medicare reform efforts, we recommend the following: reduce the geographical variation in reimbursement calculation rates; encourage risk contracting options for Medicare beneficiaries in rural areas by creating a payment floor; correct the reimbursement calculation rates to include the cost of Medicare covered benefits provided in military health care facilities. Such costs are currently excluded, and this results in disproportionately low reimbursements in States and counties with significant military presence.
 Page 146       PREV PAGE       TOP OF DOC
    Consider an approach where payments are risk adjusted. That is the calculation takes into account the health status of the HMO Medicare enrollees. Under such a system, HMOs would be fairly compensated relative to the health status of their population. Preempt the ability of States to charge a premium tax on Medicare reimbursements to risk contractors. Such a tax would result in further underpayment to risk contractors. In summary, solvency of the Medicare Program can be achieved through fair and appropriate reform. Encouraging seniors to move into Medicare risk HMOs will reduce the Federal budget significantly and help ensure Medicare longevity. Mr. Chairman and Members, this concludes my remarks.
    Thank you for the opportunity to present these views to the Subcommittee. I welcome your questions.
    [The prepared statement follows:]

Statement of Jack Myers, Vice President of Financial Services, Group Health Cooperative of Puget Sound

    Good morning, Mr. Chairman and members of the Ways and Means Subcommittee on Health.
    On behalf of Group Health Cooperative of Puget Sound and our 55,000 Medicare beneficiaries, I appreciate this opportunity to appear before you today to share some of our thoughts about Medicare risk contract reimbursement. Let me first express our gratitude, Mr. Chairman, for the Balanced Budget Act of 1995. This act, for the first time, addressed the problems of low payment issues for Medicare risk contractors. We understand that the President's recently released budget may also address low payment issues.
    Group Health Cooperative of Puget Sound is a not-for-profit managed healthcare system, headquartered in Seattle, Washington. We are the Northwest's largest health maintenance organization (HMO) with an enrollment of over 650,000. We operate in 28 of the 39 counties in Washington State and in 3 counties in North Idaho. We have approximately 10,000 employees, including over 900 staff physicians. We also contract with an extensive network of over 2,400 other community physicians, group practices, and hospitals to serve our patients. Group Health is accredited by the National Committee for Quality Assurance (NCQA) and the Joint Commission for the Accreditation of Healthcare Organization (JCAHO).
 Page 147       PREV PAGE       TOP OF DOC
    The Cooperative was started by a grassroots consumer effort in 1947. It is now the nation's largest consumer-governed healthcare organization. Group Health is governed by an 11-person, volunteer, consumer Board of Trustees, elected by voting members of the Cooperative. Any enrollee can become a voting member by paying a small lifetime membership fee, currently $25. Voting members elect the Board of Trustees and vote on policy-related advisory referenda.
    Group Health strongly supports the reform efforts of the Congress and Administration directed toward assuring both the short-term and long-term solvency of the Medicare program. We also strongly support offering Medicare beneficiaries a complete range of choices for their health care coverage, including fee-for-service (FFS) and Medicare HMOs. We want to assure that these choices are available throughout the country.

Benefits of Risk Contracting for Seniors

    Group Health has participated in the Medicare program since its inception in 1965. In 1976 we were the first risk contractor in the nation. We are currently the largest Medicare risk contractor in the State of Washington, serving over 55,000 Medicare beneficiaries.
    Capitated health maintenance organizations (HMOs) have a proven track record for offering expanded benefits; simplifying the paper jungle; and improving quality, cost, and access for Medicare beneficiaries. Not surprisingly, these plans are increasingly popular with seniors. This is demonstrated by the recent, rapid increase in enrollment in Medicare risk HMOs. Because of our experience administering Medicare risk contracts, we are convinced that any payment reform that includes strong incentives for beneficiaries to choose managed care would benefit the government, the beneficiaries and the providers.

 Page 148       PREV PAGE       TOP OF DOC

    Group Health strives to provide innovative care in a variety of settings including preventive care, 24 hour consulting nurse services, home health and other care models to improve the health of our seniors. The Medicare risk contracting program, which is premised on prospective, capitated reimbursement, is consistent with this mission. Capitated payments encourage us to continually seek new and improved ways to care for our seniors.

Seattle Low Cost Healthcare Market

    The Seattle area is an efficient healthcare market where healthcare costs and utilization are below the national average. Risk contract payments are based on 95% of an areas costs and utilization. For Group Health this results in risk payment rates ranging from 8–25% below the national average. With these rates significantly below the national average we barely break even financially. We want to make certain that any reforms of the Medicare payment system assure the viability of Medicare risk contracts in all areas of the country, including low-cost, efficient healthcare markets such as Seattle's.


    Medicare HMOs attract a broad population mix and enrollees repeatedly report satisfaction with their care. Results from the Physician Payment Review Commission's (PPRC) November 1996 study Access to Care in Medicare Managed Care: Results from a 1996 Survey of Enrollees and Disenrollees show that the majority of seniors enrolled in HMOs perceive no barriers to access and nearly all are satisfied with their health plans.
    Ninety-six percent of the beneficiaries in the study rated the overall health care provided by their HMO plan as good to excellent, with 43 percent rating their care as excellent. The study found that most seniors (87 percent) would recommend their plan to family and friends. The study showed a very low disenrollment rate with only 2.8% leaving the plan and returning to fee-for-service. Of that small group, only 2% switched for reasons of satisfaction. Group Health has a similarly low disenrollment rate.
 Page 149       PREV PAGE       TOP OF DOC


    Seniors are enrolling in HMOs in increasing numbers due to the comprehensive, affordable, quality health services, and ease of use these systems offer. One important reason that Medicare HMOs are so attractive to beneficiaries is that they offer more benefits than fee-for-service Medicare at little or no additional cost. The affordability of Medicare HMOs is especially important at a time when seniors in the fee-for-service sector are experiencing double digit increases in the average premium for supplemental Medigap coverage.

Population Mix

    We are aware of several recent studies which indicate that Medicare risk HMOs are overpaid because they enroll healthier than average seniors in their programs. This is often referred to as ''favorable selection.'' This argument is being used to justify proposals for significant reductions in reimbursement to Medicare HMOs in the future. We would like to note that while this argument may be valid for some plans it is not true across the board.
    Group Health Cooperative's Center for Health Studies, a nationally recognized research center, did a study of the health status of our Medicare population and our community in 1993, applying the same methodology used in the frequently cited Mathematica study, to determine if Group Health had favorable or adverse selection. The results of the study indicated that Group Health's Medicare population was equal in health status to our general community population—not healthier, not sicker. We will be repeating this study to reassess our current Medicare population mix in 1997.
    Group Health Cooperative is a nationally accredited, high quality care provider that improves the health status of seniors and has high satisfaction indicators. The troubling news is that Group Health may not be available as a choice in our community unless payment reform occurs.
 Page 150       PREV PAGE       TOP OF DOC

Inequities of Current Methodology

    There is significant geographic variation in the adjusted average per capita cost (AAPCC) throughout the country that creates inequities both for risk contractors and for beneficiaries residing in various states. A few highly populated areas of the country benefit from high AAPCC's which produce significant margins for risk contractors enabling these contractors to pass savings from the AAPCC reimbursement on to their beneficiaries in the form of additional benefits at little or no charge. However, most areas of the country have AAPCC rates which are not high enough to permit health plans to offer their beneficiaries additional benefits or indeed to even participate profitably in the risk contract program. In fact the variation from highest to lowest 1997 county AAPCC rates is 347% (from a low of $221to $767). In comparison, the variation in Diagnosis-Related Group (DRG) payments to hospitals under the Prospective Payment System is only 14%. The wide AAPCC variation cannot be explained by local price differences. It is due in part to the variation in hospital and physician capacity and the per capita use of healthcare resources. It is also due to flaws in the current AAPCC calculation methodology. This geographic variation results in risk contracting being unattractive in low AAPCC areas and beneficiaries in low AAPCC areas receiving fewer additional benefits from the Medicare program relative to Medicare beneficiaries residing in higher AAPCC counties.
    Group Health Cooperative has been supportive of risk contracting since 1976 when we helped develop the first pre-TEFRA risk agreement. We believe in the risk contract because it is consistent with our principle of holding managed care organizations accountable for both quality and cost management. However, as we must be partners with the consumers of our healthcare services, we must also be partners with the payers for those services. We believe that our approach to managing healthcare should produce savings for the payers of healthcare. And so far we have been able to achieve those savings in one of the most efficient fee-for-service communities in the country. However, it is becoming increasingly difficult for us to participate in the Medicare risk program. For the past several years AAPCC rates in our service area have risen on average 2% below the rate of increase in the United States Per Capita Cost (USPCC). In 1997 Group Health's average payment rate increase was 3.8%, the USPCC (national average) rate increase was 5.9%. Counties in other states across the nation had increases as high as 8.9%. Currently every Washington State county AAPCC is below the USPCC national average. King County, our highest rate, is a full 8% below the national average.
 Page 151       PREV PAGE       TOP OF DOC
    This is the paradox of the AAPCC methodology; a cost efficient community's efforts are recognized with a low rate increase, while higher-priced, inefficient communities receive increases well above the average.
    Now with a combination of proposed reductions in payment to fee-for-service providers, upon which the AAPCC's are based, and absolute reduction of 5% in the AAPCC's, reimbursement will be too low to cover the cost of care in Seattle and many other markets.

Other Issues Affecting Payment Methodology

    Flaws in the AAPCC calculation exist in areas where there are Military Treatment Facilities (MTF), U.S. Treatment Facilities (USTF) or Veterans Administration Hospitals (VA). Medicare beneficiaries who are eligible for both Medicare and military medical coverage sometimes receive care at these military facilities. To simplify the AAPCC calculation, all fee-for-service (FFS) costs in a given county are divided by all Medicare beneficiaries in that county to derive the payment rate. The computation of the AAPCC includes all Medicare beneficiaries in the denominator. However, since the facilities providing care to military eligible beneficiaries do not report Medicare costs to HCFA, the numerator of the AAPCC excludes any costs Medicare beneficiaries received in these facilities. This results in an understatement of the AAPCC wherever there are military healthcare facilities. States or counties with a significant MTF/VA presence receive disproportionately low AAPCC rates due to this methodology error. While the Prospective Payment Assessment Commission (ProPAC) has estimated a national average AAPCC understatement of 3%, we have calculated Part A (inpatient) Medicare risk payment understatements of 4.3% in King County, Washington and 22.6% in Pierce County, Washington. Our estimates were based on utilization data obtained from the Department of Defense's Tricare program. The methodology needs to be revised to either include all the beneficiaries and all their Medicare allowable costs, or exclude military eligible beneficiaries and the costs of services provided by military and community providers. We believe it would be most appropriate to include both the Medicare beneficiaries and the costs for all their Medicare services including those received in FFS and at the military facilities.
 Page 152       PREV PAGE       TOP OF DOC

Premium Tax

    Some states, including Washington State, are currently considering the assessment of a premium tax on the reimbursement health plans receive from the Federal government for Medicare covered benefits. Allowing states to assess this tax would effectively reimburse risk contractors at less than 95% of the AAPCC and place risk contractors at a competitive disadvantage with indemnity insurance carriers.
    Proposals which produce Federal budget savings by reducing payment rates to all risk contractors from 95% of the AAPCC to 90% do not recognize the above stated problems inherent in the current payment methodology that result in underpayment to some healthcare plans. Furthermore, an across the board 5% reduction is based on an assumption that all risk contracting managed care plans benefit from selecting the healthiest segment of the Medicare population. In fact, not all healthplans have favorable selection.
    We are very concerned about the across the board payment reductions being proposed as a method of dealing with favorable selection. This will severely penalize plans with a balanced population mix and jeopardize their ability to participate in the risk contracting program.

Risk Adjuster

    We would prefer to see an approach where payments are risk adjusted to take into account the health status of each HMO's Medicare enrollees. Under such a system, HMOs would be fairly compensated relative to the health status of their population. We support HCFA's efforts to identify a fair and equitable risk adjustment system that can be incorporated into the risk contracting payment system soon. Group Health's Center for Health studies is collaborating with HCFA on research into this important issue. Risk adjustment needs to be part of any Medicare payment reform proposal adopted by Congress.
 Page 153       PREV PAGE       TOP OF DOC


    In your efforts to reform Medicare risk contracting, Group Health recommends:
    Reducing the variation in county AAPCC rates by blending county AAPCC rates with the national USPCC rate.
    1. Encourage risk contracting options for Medicare beneficiaries in rural counties by creating a payment floor in the AAPCC rates.
    2. Correct AAPCC rates to include the cost of Medicare covered benefits provided in military healthcare facilities so as to include all Medicare beneficiaries and their allowable costs.
    3. Address the issue of risk selection directly and specifically by developing and implementing a credible health status risk adjuster.
    4. Pre-empt the ability of states to charge a premium tax on Medicare reimbursement to risk contractors.
    Mr. Chairman and members, this concludes my remarks. Thank you for the opportunity to present our views to this committee. I welcome your questions.


    Chairman THOMAS. Thank you, Mr. Myers.
    Mr. Schub.

 Page 154       PREV PAGE       TOP OF DOC

    Mr. SCHUB. Thank you, Mr. Chairman, Members of Congress. My name is Craig Schub, senior vice president of PacifiCare Health Systems, where together with our Secure Horizons Program, we serve 3 million commercial beneficiaries and 1 million Medicare beneficiaries in 14 States covering counties with the full range of AAPCCs. We appreciate the opportunity to testify today.
    I would like to open my comments by saying that basically HMOs have demonstrated that they can provide more benefits, lower cost, higher satisfaction with equal to or greater quality than a fee-for-service Medicare Program. These HMOs are driving change, fundamental change, in health care delivery. And, in fact, reducing fee-for-service costs in markets that they operate in by fundamentally changing the cost structure in our health care delivery system.
    However, HMOs are still early in their development. They account for only 12 percent of the population, and we seriously need to question the disproportionate cuts of 30 percent of the total proposed Medicare cuts. That will clearly protect the fee-for-service system and discriminate against HMO beneficiaries. It will jeopardize benefits and costs savings to the middle- and low-income beneficiaries who tend to make up the majority of HMO members. They will destabilize health plan and provider's ability to predict future payments and threatens current investment in health plan expansion.
    Additionally, these changes are not based on any risk adjustment methodology and are, in fact, premature. Risk selection studies have identified an early growth phenomenon that may no longer exist in many markets as we clearly regress to the mean that was quoted in earlier testimony. In the 2 years since the latest study that was used for risk selection, membership in HMOs have increased by 80 percent, and there has been an unprecedented number of new plans entering the market.
 Page 155       PREV PAGE       TOP OF DOC
    Younger and healthier is the claim that is usually made. The fact is that in Secure Horizons of California, our largest and oldest plan, our average exceeds the age of the general senior population, as noted on the chart to my right. The green bar is the average age of our Secure Horizons members, clearly showing a larger proportion of older seniors, and that bar is moving to the right over time.
    The bottom line is HMOs have accepted the risk for providing care under a capitated system for our oldest and most needy of beneficiaries. Now as seniors are aging, getting sicker and staying in HMOs, we are proposing to change the rules. HMOs have demonstrated they can manage costs and are comfortable with predictable budget increases to create long-term savings that are necessary to protect Medicare. However, such things as the elimination of GME dollars allowing for contracting within HMOs to the higher cost teaching institutions, the elimination of those GME dollars threaten our stability.
    I ask Congress to caution and consider that the lack of clarity in a future payment structure together with the proposed discriminatory cuts against HMO members and together with the fact that there is currently competitive bidding demonstrations and third-party enrollment demonstrations being considered, all create a sense of instability and confusion for seniors and for future health plan investment. Currently, seniors have a choice and freedom that create an effective, competitive market.
    Rewarding cost efficiency based on consumer value, protecting beneficiaries from less quality plans by allowing them to disenroll in 30 days and encouraging competition among multiple growing integrated health care delivery systems. HMO penetration is lowering increases in fee-for-service costs by changing the community practice of medicine. However, we are far from where we need to be. Expansion of these programs need to continue if we hope to have a delivery system that can provide care to the burgeoning senior population in the future.
 Page 156       PREV PAGE       TOP OF DOC
    Introducing disproportionate cuts at this time, therefore, disrupt current market dynamics, risk to beneficiary's benefits and costs, particularly in the mid- to low-income seniors, and actually favors HCFA because we have to realize that HCFA is both a regulator and by being the operator of the sole fee-for-service system, in fact, the direct competitor to HMOs, and are not based—these changes are not based on any real understanding of actual health plan risk on a specific basis, yet are based on an aggregate in a rapidly changing market with data that is already out of date.
    So we ask you to seriously consider caution in excessive HMO cuts in an attempt to fix something that at this time is apparently not broken. Thank you.
    [The prepared statement and attachments follow:]

Statement of Craig S. Schub, Senior Vice President, PacifiCare Health Systems

    Mr. Chairman and members of the Subcommittee, I am Craig Schub, Senior Vice President of PacifiCare Health Systems. My responsibilities at PacifiCare include strategic policy for our commercial and government programs including Secure Horizons, PacifiCare's Medicare risk HMO.


    PacifiCare recently completed the acquisition of another managed care organization, FHP International Corporation, making the combined company the nation's fifth largest HMO. We will serve nearly 3 million commercial and 1 million Medicare members in 14 states and the Territory of Guam. Those states include Arizona, California, Colorado, Illinois, Indiana, Kentucky, Nevada, New Mexico, Ohio, Oregon, Oklahoma, Texas, Utah and Washington. The combination of these two companies creates an extremely strong operation with a solid leadership position. Our goal is to build a company that both members and physicians view with confidence and trust. The acquisition further solidifies PacifiCare's position as the nation's leading Medicare risk plan, with nearly 1 million members enrolled. Our organization is committed to applying its skills, experience and efficiencies to help address the complex issues confronting Medicare.
 Page 157       PREV PAGE       TOP OF DOC
    PacifiCare began operations in 1978, and by 1985, had formed ''Secure Horizons'' and landed its first Medicare-risk contract. We offer a full range of products and services, from traditional health maintenance organizations to point-of-service plans and preferred provider organizations, as well as programs to reduce the cost of prescription drugs and workers' compensation claims. PacifiCare serves a variety of market segments, including companies large and small, individuals, the military and seniors. In addition to managed care products for employer groups, and Medicare and Medicaid beneficiaries, our other specialty operations include Medicare HMO management services, coordination of managed care products for multi-region employers, health and life insurance, behavioral health, health promotion, as well as dental and vision services.
    PacifiCare also operates a non-profit organization called the PacifiCare Foundation that is devoted to charitable and educational causes that enhance the health, wellness and welfare of individuals, families, and the community.
    Today, I would like to review two subjects as the committee examines the need for change in the Medicare program:
    •  The current status of the Medicare HMO risk contracting program;
    •  The changes and issues to consider as the Administration's budget proposal is evaluated; and
    •  The issues to consider as we reform the risk payment methodology.

The Medicare Risk Contract Program

    The Medicare HMO risk contracting program was enacted in 1982 as part of the Tax Equity and Fiscal Responsibility Act (TEFRA), after a number of years of successful demonstrations of the payment models. The TEFRA program authorized HCFA to contract with HMOs and competitive medical plans (CMPs)(see footnote 9)
 Page 158       PREV PAGE       TOP OF DOC
for the provision of Medicare covered health care benefits. HMOs/CMPs which could meet or exceed health care quality, financial stability and operational standards could receive payment equivalent to 95 percent of the estimated cost of fee-for-service Medicare in the geographic area (the county)—AAPCC.
    While the specifics of the AAPCC calculation are complex, the methodology incorporates four basic steps:
    •  calculation of the national per capita cost for care under Medicare (this is the United States Per Capita Cost, or USPCC);
    •  estimating the county per capita cost of fee-for-service care by:
    —applying a county adjuster to the USPCC (the county adjuster is a moving five-year average of the ratio of county per capita costs to national per capita costs);
    —adjusting for county HMO expenditures: and
    —adjusting for the demographic mix of the county's beneficiaries.
    •  multiplying the county per capita cost by 95 percent; and
    •  applying the actuarial risk adjustments developed by HCFA to account for age, sex, Medicaid status, institutional status, and employer-based coverage of the beneficiary.
    Further, the HCFA risk plan payment process has integrated into the Adjusted Community Rate (ACR) mechanism—an actuarially-based procedure which is reviewed and approved by HCFA to determine the value of each plan's basic benefits package and any supplemental benefits. The ACR mechanism is structured to ensure that members receive the maximum benefit value from the government's payments to risk plans and, most importantly, that plans cannot profit by denying the extra benefits which could be provided with a capitated payment.
    Since the inception of the Medicare program in 1982, enrollment has increased dramatically. As of February 1997, nearly 5 million Medicare beneficiaries were enrolled in HMOs, representing 12 percent of the Medicare population. The geographic distribution of those beneficiaries in HMOs have been concentrated in limited parts of the country, only recently expanding to most metropolitan areas. Five years ago, only 6.4 percent of Medicare beneficiaries were enrolled in health plans. According to the Health Care Financing Administration (HCFA), Medicare enrollment in risk contracts is currently growing by 80,000 to 90,000 beneficiaries per month. Annual growth rates for Medicare beneficiaries enrolled in the risk program have grown steadily, from 10 percent in 1990–1991 to 32 percent in 1996–1997.
 Page 159       PREV PAGE       TOP OF DOC
    At present, only one of every eight Medicare beneficiaries is an HMO member. This contrasts notably with the private sector where more than 70 percent of Americans now receive their care through HMOs, PPOs and other types of health plans.
    Let me briefly review for the Committee some of the most successful features of the Medicare risk program: more comprehensive benefits and value than traditional Medicare, high member satisfaction, documented quality of care, and significant savings for beneficiaries and the Medicare program.

More comprehensive benefits than traditional Medicare.

    Seniors are enrolling in Medicare HMOs in increasing numbers due to the comprehensive, affordable and quality health services offered. One of the reasons that Medicare HMOs are so attractive to beneficiaries is that they offer more benefits than FFS Medicare at little or no additional cost. Under current law, Medicare HMOs must provide beneficiaries with coverage for all the benefits required in FFS Medicare. In addition, the risk plans are required to use any surplus funds between expected payment from HCFA and their projected revenue requirements for Medicare covered services to provide additional benefits, reduce premiums charged to beneficiaries, or return the savings to HCFA.
    As a result, the majority of Medicare risk plans offer seniors many benefits not covered by FFS Medicare: outpatient drugs; routine physicals, including annual mammographies and prostate cancer screenings; vision and dental care; immunizations; and preventive health screenings, such as eye and ear exams. Almost half of all Medicare risk plans offer outpatient prescription drugs. In addition, 94 percent of Medicare risk plans offer routine physicals, 86 percent offer eye exams, and 83 percent offer immunizations. For many seniors, coverage for these services would otherwise be available only through costly supplemental Medigap insurance, which is unnecessary for HMO risk enrollees.
 Page 160       PREV PAGE       TOP OF DOC
    HMOs also eliminate the costs to beneficiaries in FFS Medicare, which currently amount to $760 Part A deductibles and 20 percent Part B coinsurance. Seniors enrolled in the risk program may often have reduced or no premiums, even while receiving coverage for these additional benefits. In fact, half of all Medicare risk contractors do not charge seniors a monthly premium, and 28 percent charge a premium less than $40, while Medigap premiums range from $65 to nearly $200 per month.
    We offer a zero-premium plan in the majority of the 424 markets in which the newly-combined company, including Secure Horizons and FHP's Senior Plan operates. In each of these counties, the Average Payment Rates (APR's) and AAPCC's vary significantly from an APR of $240 per month per beneficiary in Josephine County, Oregon, to $556 per month per beneficiary in Los Angeles, County, California. In only five counties out of the 424 are the Average Payment Rates over $500, and none of PacifiCare's counties are listed in the reported twenty highest payment counties.


    PacifiCare is proud of the care we provide to older Americans, including those with multiple chronic medical conditions, and many studies have confirmed that Medicare HMOs offer high quality health care. A recent study by the Health Care Financing Administration showed that elderly HMO members with cancer are more likely to be diagnosed at an early stage than those in the fee-for-service sector.(see footnote 10) Breast, cervical and colon cancers, along with melanomas, were diagnosed significantly earlier in HMOs than in fee-for-service. The study attributed this difference to improved access to preventive care under comprehensive HMO coverage. Improved access to preventive care is also highlighted in a study by the Centers for Disease Control (CDC) and the National Center for Health Statistics that showed women in HMOs are more likely to obtain mammograms, pap smears, and clinical breast exams than those in the fee-for-service sectors.
 Page 161       PREV PAGE       TOP OF DOC
    Another study, comparing care for patients age 65 and older with acute myocardial infractions (heart attack), concluded the HMO patients received better care than that received by patients in a national fee-for-service sample.(see footnote 11)

    At PacifiCare, we recently initiated a new program entitled, ''Quality Improvement Initiatives.'' The main goal of this program is to demonstrate improvement in the quality of care our members receive. To achieve this goal, we will work closely with our providers and our market health services leadership over the next year on the following initiatives:
    •  Women's and Children's Health
    •  Diabetes Management
    •  Cardiovascular Health
    •  Depression
    In keeping with our mission to improve the health of our members, these areas address the health care needs of PacifiCare's entire member population, from newborns to the elderly, and target common medical conditions that occur frequently among our membership. Several of these initiatives are related to and building upon HEDIS (Health Plan Employer Data and Information Set) and involve collaboration with academic and professional experts in quality of care research and improvement.
    Over the next few months, each PacifiCare health plan will systematically collect data under the direction of the PHS Quality Measurement and Research Department. The data will provide information on opportunities for improvement. We will partner with our providers in each market to address these opportunities and believe this process will give our members a clear expression of quality health care.
    We also try to ensure quality care and member satisfaction by providing our seniors with ''value-added'' programs, including: member advocacy, pro-active member contacts, member education outreach, ambassador program, wellness programs, Horizons Magazine and consumer advocacy groups, community events, the PacifiCare Foundation/Senior Fund and other member satisfaction measurements. I will elaborate briefly on a few of these programs:
 Page 162       PREV PAGE       TOP OF DOC
    Member Advocacy. The Secure Horizons Member Service Department exists as a resource and helpline for our members. Members currently call our toll free number for reasons ranging from simple questions to complex problem resolution. The department is staffed with specially trained service professionals who act as plan advocates to help members obtain the most from their plan coverage.
    Secure Horizons also has in place a formal Appeals and Grievance process to ensure that members receive full and fair review of their complaints. If a member is dissatisfied with the quality of medical care received from a health plan physician, they have a place to turn for intervention. If a member desires reconsideration for a benefit or service that has been denied, Secure Horizons' Member Service departments assists the member through the Appeals and Grievance process and acts as an advocate for the member.
    Members receive immediate intervention from a medical professional when there is a medical care issue. This process allows the member to interact directly with the health plan to obtain swift recommendations and alternatives for resolution to urgent medical care needs.
    Pro-active Member Contacts. Secure Horizons attempts to contact prospective members who have submitted applications for enrollment to ensure that each individual understands the features of the Medicare risk plan and has selected a primary care physician or medical group. This program decreases the number of seniors who might not understand what type of health plan in which they are enrolled.
    Member Education Outreach. Secure Horizons conducts outreach to help members understand how to best utilize their health plan resources through new member education calls within the first 60 days of membership and monthly member education meetings.
    Wellness Programs. Secure Horizons offers a comprehensive wellness program to encourage members to lead active, healthy lives. The ongoing program includes a variety of events, such as exercise and health education, screenings, walks, wellness festivals and other social activities.
 Page 163       PREV PAGE       TOP OF DOC
    Consumer Advocacy Groups. Secure Horizons hosts open houses at our corporate offices for senior advocacy groups, where representatives meet with key individuals from our member services, health services, case management and administration departments. In addition to gaining an understanding of our role in the delivery of health care, visits provide an open forum for us to discuss common issues that consumers face with their health plans.
    Access to care is also a critical component of quality assurance. The newly-combined PacifiCare Health Systems is projecting that we will have 90,000 doctors in our network in 1997, including private practice physicians, and independent physicians associations (IPAs) comprised of private doctors located in individual offices throughout the community. Each doctor has been selected based on proven records of delivering quality care, convenient locations, and well-equipped medical facilities. Each medical provider is affiliated with one or more well-established hospital(s) in addition to a network of specialists and medical technicians located in the community nearby.
    As a part of our improvement and in an effort to speed referral time to specialists, PacifiCare/Secure Horizons has created ''Express Referral'' in our more advanced health care delivery system. A doctor may directly refer a member to a specialist without additional authorization from a medical committee. This type of innovation continues to drive change in local provider delivery systems that is improving the quality and lowering the cost of all care, including fee-for-service in those communities.
    Satisfaction: Several recent studies provide evidence of satisfaction among Medicare HMO members:
    •  In November 1996, the Physician Payment Review Commission (PPRC) released a study that found high levels of satisfaction among Medicare beneficiaries enrolled in health plans. The PPRC study found that 96 percent of those surveyed rated the overall health care provided by their plan as good to excellent, with 43 percent rating their care as excellent. The study found that most seniors (87 percent) would recommend their plan to family and friends.
 Page 164       PREV PAGE       TOP OF DOC
    •  A 1995 study by the National Research Corporation found that HMO and FFS elderly patients who say they are in fair or poor health are about equally satisfied with their coverage.
    •  A May 1994 survey conducted by the American Hospital Association found that Medicare beneficiaries enrolled in HMOs and old-style fee-for-service are equally satisfied with the quality of their care.
    •  In the most recent Secure Horizons survey of member satisfaction, 94 percent of our members say they are satisfied on an overall basis, and 95 percent surveyed would recommend Secure Horizons to family and friends.
    Recent studies have also documented low levels of disenrollment among Medicare risk enrollees which serves as further evidence of their satisfaction with their health plan:
    •  The November 1996 study released by the Physician Payment Review Commission found that only 2.8 percent of those surveyed had disenrolled from their health plans and returned to fee-for-service Medicare. Twenty-eight percent of these beneficiaries switched because they moved out of the service area or because their health plan ceased operation, leaving only 2.0 percent who chose to return to fee-for-service Medicare.
    •  A study released earlier this year by Gerald Riley of HCFA's Office of Research and Demonstrations found that Medicare HMO enrollees with cancer were no more likely to disenroll than cancer-free enrollees. In fact, patients who developed cancer while in an HMO were less likely to disenroll than cancer-free patients.(see footnote 12)


 Page 165       PREV PAGE       TOP OF DOC
    The Medicare HMO program has achieved demonstrable savings for the Medicare program as a whole. These savings are achieved in two ways: first, Medicare HMOs are paid less than the cost of caring for beneficiaries under the fee-for-service program; second, increasing levels of Medicare HMO enrollment in local areas means lower overall Medicare costs for that part of the country—commonly referred to as the ''spillover effect.''
    Numerous studies show that the presence of HMOs in health care markets help to reduce costs throughout the health care system. The phenomenon of ''spillover savings'' has been studied in relation to Medicare costs, health insurance premiums, and hospital costs. Many studies have shown that increasing HMO market share produces substantial savings in health costs whether measured by Medicare fee-for-service costs, aggregate Medicare costs, aggregate hospital costs, or indemnity insurance premiums. Studies have also shown lower hospital costs in areas with high HMO market share. This effect more than outweighs any possible short-term selection issues that may or may not exist.
    For instance, a 1995 study by Wickizer and Feldstein showed that areas with higher HMO market share experienced lower rates of increase in indemnity insurance premiums than areas with lower HMO market share.(see footnote 13) The authors found that a 10 percent increase in HMO market penetration led to a 6.6 percent decrease in the rate of growth in health insurance premiums. The authors concluded that managed care could significantly reduce health insurance premium growth rates, resulting in considerable cost savings over time.
    Another study conducted at Georgetown University in 1995 reported that hospital costs per admission increased 69 percent in markets with high HMO penetration from 1984 to 1991, compared to 96 percent for markets with low HMO penetration.(see footnote 14) In this study on spillover savings, Stanford University economist Laurence Baker found that increases in HMO market share are associated with reductions in fee-for-service Medicare spending levels.(see footnote 15) This clearly contradicts the assumption that Medicare HMO members are skimming the healthy patients and leaving the sick in FFS.
 Page 166       PREV PAGE       TOP OF DOC

    It is also important to note that risk plans are increasingly attractive to employers because of the significant savings on retiree health care costs and expanded benefit coverage they offer. In a 1995 survey, Foster Higgins reported that the percentage of large employers who offer a Medicare HMO to their retirees has increased from 7 percent in 1993 to 21 percent in 1995. In some parts of the country, this percentage is much higher. In light of the declining rate of employer-sponsored retiree coverage, it is particularly important to maintain Medicare health plans as an amiable alternative.

Administration's Budget Proposal

    The Administration's FY 98 Budget proposal contains three provisions that would impact the Medicare risk-based contracting program:
    •  Reductions in managed care payments resulting from the ''indirect'' effect that reductions in fee-for-service have on managed care payments;
    •  Removal of graduate medical education (GME) payments, indirect medical education (IME), and disproportionate share hospital (DSH) payments from the Medicare HMO reimbursement formula; and
    •  Reductions in Medicare HMO payments resulting from reducing the Adjusted Average Per Capita Cost (AAPCC) from 95 to 90 percent, beginning in the year 2000.
    We are very concerned with the proposed Medicare HMO payment cuts and the overall impact these proposals would have on beneficiaries who choose to enroll in Medicare HMOs. The Administration's proposal to take one-third of all cuts from Medicare HMO payments is unjustified and could seriously damage the growth of risk programs. Only 12 percent of Medicare beneficiaries are enrolled in risk programs, yet the risk plans are bearing one-third of all Medicare savings. Furthermore, cuts of this magnitude could eventually cause health plans to raise premiums, reduce benefits to beneficiaries or leave the Medicare risk program.
 Page 167       PREV PAGE       TOP OF DOC
    The proposed payment cuts are especially troublesome when considering the fact that larger percentages of lower and middle income beneficiaries enroll in Medicare HMOs. It is precisely these seniors who benefit from the additional benefits and reduced premiums provided by risk plans, and who would be affected by the disproportionate cuts to managed care.
    As you evaluate the Administration's Medicare payment proposal, we urge you to consider the following issues:
    •  Seniors enrolled in health plans would shoulder a disproportionate share of the proposed cuts. The Administration's proposal, if enacted, would reduce Medicare payments to risk plans by $34 billion over the next five years, which accounts for one-third of the $100 billion in Medicare savings. Although the President has made it clear that the risk program needs to continue to grow rapidly as it has over the last several years, the level of proposed cuts in the Administration's budget will make it difficult to sustain increases in enrollment and plan participation.
    •  Avoid short-sighted reductions. The Administration maintains that some studies have concluded that on average, HMO enrollees are likely to have lower costs than those beneficiaries who remain in the traditional fee-for-service program. There is good reason to believe, however, that favorable risk selection is not a significant factor in the Medicare HMO program at this time, and that any early favorable selection has dissipated as the Medicare risk program has grown.
    Between 1994—the year used in the most recent risk selection study—and 1996 alone, enrollment in the Medicare risk program has grown by 80 percent. Presently, enrollment is growing by about 1 million individuals per year. This enrollment growth makes unreliable any estimate of favorable selection based on earlier data.
    While the Administration has cited past studies on selection bias to support its proposed cuts in HMO payments, it fails to note studies in which the health status of Medicare enrollees ''regresses to the mean'' as they remain in plans. Simply put, initially healthier people will, over time, require care typical of the average beneficiary.
 Page 168       PREV PAGE       TOP OF DOC
    In PacifiCare's mature markets Secure Horizons members are disproportionately older than the general senior population. Currently, both California and Oregon are very mature markets with over 39% HMO penetration of Medicare beneficiaries. In Oregon, over 75 percent, and in California over 72 percent of members, are over age 70. In the Texas and Oklahoma emerging markets, 65 percent of members are over 70, with a Medicare risk penetration of 11 percent and 6 percent respectively. This indicates beneficiaries in Secure Horizons in both mature markets of California and Oregon are disproportionately older than in the general population of seniors in those states. It is clear that health plans in mature, highly penetrated markets have older populations of seniors and therefore, are most likely experiencing higher utilization than the general population. Mature markets may be carrying an extra financial burden not previously revealed in earlier research. (See Attachment)
    As discussed earlier, newly issued HCFA research reported that Medicare HMO members with cancer were slightly less likely to disenroll than cancer-free Medicare HMO members. Taken together, this changing picture of HMO enrollment and disenrollment strongly suggests that any favorable selection that may have occurred in the program's early years has declined steadily.
    •  Preserve the viability of the risk program. The Administration's proposed reduction in the AAPCC will make participation in the program unattractive for many HMOs and, therefore, limit the options available. Moreover, any such reduction is a double hit on HMOs—since any budgetary reductions in Medicare FFS automatically translate into a lower AAPCC. We need to secure for Medicare beneficiaries and seniors the benefits of the choices and changes that have reshaped the private marketplace, by creating a stable payment environment whose predictability supports expanded participation by HMOs and other integrated delivery systems.
    •  Maintain a comprehensive, capitated approach. The Administration proposes to split our graduate medical education payments from the AAPCC, and to make those payments separately and directly to providers on a fee-for-service or other basis.
 Page 169       PREV PAGE       TOP OF DOC
    Rather than refragmenting the payment system, the comprehensive, integrated approach offered by HMOs—and the enhanced competitive pressure that is being generated—is exactly what is needed in our health system. The reality is that HMOs do contract with academic health centers for compelling health care and competitive purposes, since beneficiaries want access to the prominent institutions in their communities.
    Currently, PacifiCare Health Systems contracts with 8 Academic Medical Centers and 41 teaching hospitals in the State of California. In Texas, some of our major providers like the University of Texas, Galveston, and Baylor University, Dallas, are our primary provider systems along with specialty services at other University of Texas facilities in San Antonio and Houston. To arbitrarily reduce the financial ability of our health plans to contract with these programs would be detrimental to the natural integration of managed care with teaching programs.

Issues for Consideration in Reforming the Payment Methodology

    •  Expand options by reducing geographic variation in risk-based payments. We should consider gradually reducing the geographic variation in risk-based Medicare payments, in order to expand the options available to beneficiaries in all regions of the country. However, any proposal to reduce this variation should not benefit lower payment areas at the expense of higher payment areas, since this could threaten the viability of the Medicare risk program in markets that are currently well-established.
    •  Establish equity in Medicare payment levels. We should ensure that any Medicare reform proposal provides fairness and equity in payment levels to both health plans and the traditional fee-for-service program. Any savings achieved through payment reform should establish parity between payments to risk contractors and funding for FFS Medicare.
 Page 170       PREV PAGE       TOP OF DOC
    •  Oppose the removal of GME from risk payments. We should consider establishing a trust fund for graduate medical education, but only if it does not undermine the foundation of the Medicare risk program. Removing funding for graduate medical education from the AAPCC calculation could disrupt coverage for beneficiaries and limit the expansion of health plans to new markets. In addition, carving out GME funding from the AAPCC formula does not address the fundamental need to identify a stable, broad-based funding mechanism for the training of our country's health care professionals.

Other Issues in Medicare Reform

    •  Maintain beneficiary choice and sovereignty. Medicare currently allows beneficiaries to ''vote with their feet''—they retain the right, at any time, to either enroll in any health plan they choose or disenroll from any plan that does not meet their expectations. Establishing an annual open enrollment and twelve month lock-in to health plans would limit beneficiary choice; and thereby eliminate competition among plans as an important indicator of quality and consumer satisfaction.
    •  Retain the role of the health plan in the enrollment process. Establishing a third party enrollment process would create an artificial barrier between the plans and beneficiaries during a critical period when plans are setting up relationships with their members. We could instead consider starting an annual, third party education process for beneficiaries to ensure that they are provided with unbiased and complete information as they select their Medicare coverage.


 Page 171       PREV PAGE       TOP OF DOC
    The Medicare risk program is working to increase needed benefits to seniors and reduce health care costs for Medicare beneficiaries, while it is re-engineering America's health delivery systems in preparation for the inevitable explosion of the senior population. PacifiCare believes that only with a sound payment mechanism can health plans continue to provide high quality, cost-effective health care to seniors through the Medicare program. However, we are concerned that our focus on keeping seniors well through preventive, coordinated care could be jeopardized by the some of the Medicare reform provisions in the Administration's budget proposal. The proposed cuts to the Medicare HMOs and the revisions to their payment methodology and operations would endanger our ability to continue to offer the value-added, quality care seniors need.
    We have been pleased to share our perspective and concerns today, and look forward to an opportunity to review with the committee a broader framework for Medicare reform. Thank you. I would be happy to answer any questions you may have.
    [The official Committee record contains additional material here.]


    Chairman THOMAS. Thank you. If I did not know otherwise, one of those last statements might buy you into the conspiracy theory about how Medicare is operating, and I know you did not intend to do that. Mr. Myers, your concerns are obviously our concerns, and we tried to partially meet them by addressing the enormous inequities in the AAPCC from region to region. Specifically, we would address these through adjustments and a partial political compromise with the rural areas—because most of these lower payments tend to be in rural areas—at $350, and the President winds up at a $350 figure through a slightly different adjustment arrangement. We probably in the process of putting a package together will go back in and change the way in which we determine it based upon testimony that we receive today and that we will get once we get more details.
 Page 172       PREV PAGE       TOP OF DOC
    But it seems as though we have locked on a dollar amount, however we get there, at $350. Apropos my colleagues who are not now here, especially Mr. Christensen from Nebraska looking at a couple of counties at $221, is $350 in today's market, and this is directed to you, but the others could respond, is $350 an OK floor or is an unrealistic floor at which to expect self-germination or advectional HMOs to come in?
    Mr. MYERS. Well, Mr. Chairman, I think in any dollar amount that we would attach to it would be fairly arbitrary at this point, but $350 seems reasonable, and I think it will attract more plans.
    Chairman THOMAS. Is that lower than any county you have in your area or higher?
    Mr. MYERS. We currently have some counties in Washington State below the $350 level.
    Chairman THOMAS. But not in the Puget Sound area?
    Mr. MYERS. Not currently in the Puget Sound.
    Chairman THOMAS. What is your lowest county?
    Mr. MYERS. Our lowest county is about $353 in our service area.
    Chairman THOMAS. Well, $3. For government work that is close enough.
    Mr. MYERS. Yeah, yeah.
    Chairman THOMAS. You are at $350 and are able to do. Do you practice what would otherwise be called balance billing between counties to try to deliver your product not withstanding the different AAPCC rates?
    Mr. MYERS. We do have a difference in rates between what we would call our central contiguous core and those outlying counties where the AAPCC rates are lower.
 Page 173       PREV PAGE       TOP OF DOC
    Chairman THOMAS. Are you shifting then, in essence, costs from the high area to the low area to provide an equal benefit package for all?
    Mr. MYERS. No, we are not shifting the costs. We are just——
    Chairman THOMAS. I understand the sensitivity of the words, but you understand what I am trying to get at? Is there a pooling of the costs which allows you then to offer an equal benefit package for all notwithstanding the differential of the AAPCC between counties?
    Mr. MYERS. HCFA allows us to pool by county. Counties that are contiguous to one another need to have the same rate and need to be rated the same way. Outlying counties we can pool separately and therefore have the same rates recognizing the differences in market rates, market competition in those areas.
    Chairman THOMAS. Any reaction from either of you about the differential county rates and your pooling to create an equitable package?
    Mr. WIGGINS. Well, first of all, I think Gail Wilensky made the very important point and that is by having a $350 floor, it will make HMOs more attractive than the fee-for-service program, and I think any time you begin to uncouple the fee-for-service program from the HMO Program, we are setting up an unsound system. In Connecticut, we have an average of $377 right now. In New Jersey in some counties we are at $399. In New York City, it goes as high as $572. We observe that there are material cost differences in our market that justify some of those differentials, and it seems to us that there is yet to be defined a surrogate that is a better measure than the historic fee-for-service costs for what the relative differences in costs are.
    I would be very interested in moving to a State that has $221 fee-for-service costs in starting an HMO where I am getting $350. I am not certain that that would be the best public policy, but it would certainly make good business sense to move into those counties. The answer to your second question——
 Page 174       PREV PAGE       TOP OF DOC
    Chairman THOMAS. Better yet would be a fee-for-service operator sitting in that county currently getting $221 and suddenly being provided with $350 with no HMO moving in.
    Mr. WIGGINS. Is that not beautiful.
    Chairman THOMAS. Which is my greater concern.
    Mr. WIGGINS. Yes. The other point that you make is true. We are losing money in some counties and making money in others. In fact, we had to change our benefits as a consequence of recent changes in the payment rates to Oxford. We instituted what we call A and B county benefits. And we have Medicare beneficiaries here today in both our A and B, Bob and Dorothy Rothman from Long Island are here, and they had a benefit cut. They came voluntarily understanding that there is a Medicare debate going on. They are in the audience. And Arlene and Alex Krasowsky are in Queens in what we call the A counties, and there are very significant benefit differentials in these various areas because there are different payment levels. And some regression to a mean is appropriate. The extreme changes that some suggest cannot be accommodated in the expensive counties like the Bronx where health care costs are just very high.
    Chairman THOMAS. So what we have in the Bronx and Queens are under the umbrella of the Oxford Plan. Folks receiving additional benefits beyond the basic package, based upon the AAPCC. You do not do that in the Puget Sound health system? If you are in one county versus another county notwithstanding the reimbursement differential, Mr. Myers, you offer the same package to both?
    Mr. MYERS. We offer the same packages, that is correct.
    Chairman THOMAS. Mr. Schub.
    Mr. SCHUB. Mr. Chairman, within PacifiCare, we do offer a range of benefits depending on the county. It is difficult for us to explain to a provider when HCFA determines that there is a payment differential and a cost differential why we should pay them different. And the other thing that I really——
 Page 175       PREV PAGE       TOP OF DOC
    Chairman THOMAS. I would be more interested rather than the provider but the beneficiaries, who if they are in the same plan receive fewer benefits because they live in Queens versus the Bronx because to them they cannot really see a differential in terms, I assume, of where they live, and the costs that they believe are associated with their health care, but receive a different package.
    Mr. SCHUB. That is correct, and they are confused by that fact. Additionally, though, the status of the delivery system within that community, you may find that within a given community with slightly less money, the health care delivery system is, in fact, more efficient and we can, in fact, provide more benefits than in another county. So there are a number of factors that affect what we can deliver, and we simply try to maximize the value to the beneficiary.
    Chairman THOMAS. Now the President has indicated in the rather scarce information we have had about his plan so far that they want to begin to standardize the add-on benefits. And you have just created a new question for me. I do not know whether they want to standardize within the county payment structure or whether you standardize it by plan which means you are going to have a significant cost shift where you do an A–B program versus those who run a single one.
    One, I would like to know what is your opinion? Would it help you from a marketing point of view? Would it concern you about rigidity and bureaucracy being established in reduction of flexibility if there was a standardization of add-on benefits?
    Mr. SCHUB. I would say absolutely. We believe——
    Chairman THOMAS. Absolutely to which one? Once again I am——
    Mr. SCHUB. Absolutely there would be——
    Chairman THOMAS. See I am not an attorney and I do not know how to ask these questions to where I just get one answer. I wind up giving you options. What is your option?
 Page 176       PREV PAGE       TOP OF DOC
    Mr. SCHUB. Having the flexibility to change benefits based on the values received by beneficiaries as opposed to some bureaucratic edict.
    Chairman THOMAS. So you would be concerned about the standardization if it did not allow more of an actuarial equivalent on standardization than specific product?
    Mr. SCHUB. Yes, sir.
    Chairman THOMAS. Is that true of the others?
    Mr. MYERS. Well, as our past practice indicates, we believe that it is better to offer the same benefit package to all of the beneficiaries within our service area.
    Chairman THOMAS. So you would be less concerned about a standardization?
    Mr. MYERS. That is right.
    Chairman THOMAS. Yes.
    Mr. WIGGINS. I think it would be very problematic if you believed that competition is not good. Some health plans are lower cost than others, and in our case we have a good cost structure in Medicare. We have an 86.7-percent medical loss ratio. We have 1 percent, just a little over 1-percent net income in that program. We are highly competitive, and if there was a standardization of benefits, we may actually end up giving away benefits that beneficiaries already receive.
    Chairman THOMAS. If any of this is proprietary, I understand why you would be reluctant to answer this question, but under the risk contract arrangement, if, in fact, the amount on the ACR is less than what the payment is, you have to offer something. It is not standardized so it is up to you folks as to what you want to offer. I notice about two-thirds of the plans, growing by about 25 percent a year, offer the money, a reduction in the premiums, or copayments. Others in varying amounts, nearing 100 percent for some items, which if you already have the structure there do not cost you really anything to do, various benefits that would be part of the program. From your perspective, and I guess I am trying to understand how you market a product as well, if you had one additional dollar to spend, where would you spend it or where do you spend it in your program? I am trying to understand the marginal value of each additional dollar available to you as an inducement or an attractor to the program that you have. What do you think is most important to you? That would help me in hierarchically arranging any standardization of benefits that might come along. So if you had one additional dollar, where would it go? Do you understand my question?
 Page 177       PREV PAGE       TOP OF DOC
    Mr. WIGGINS. Well——
    Chairman THOMAS. Notwithstanding what you do now with it, assuming everybody had the same AAPCC, because, Mr. Myers, I can understand your problem. Since you already pay for the medical stuff, and you have a very limited budget, you would not be giving dollars out. You would be giving services out, but if you have an opportunity, what do you think is the most attractive way to use that additional dollar?
    Mr. WIGGINS. Prescription drugs. That is what drives seniors into the program.
    Chairman THOMAS. Prescription drugs.
    Mr. WIGGINS. And Congresswoman Johnson was absolutely correct in stating that in those markets where prescription drugs are routinely covered for Medicare beneficiaries, you do tend to attract those that need the prescription drugs. That is a very big exposure for seniors, and if you can add the prescription drugs, they are more likely to join. You are also getting sicker people coming into the plans.
    Chairman THOMAS. But that additional factor is not available in 90, it's not available in 80, it is not available in 70. It is not even available at 60.
    Mr. WIGGINS. Under the President's proposal, in every county that we operate, if you—we have run the math actuarially, and I have given you this summary—all additional benefits are wiped out under our program, and we have to increase cost sharing.
    Chairman THOMAS. Yes, but we are not looking at the President's plan, we are looking at my hypothetical first dollar, and your first dollar would be spent on prescription drugs?
    Mr. WIGGINS. Well, I think your question was where does it deliver the most benefit, and I believe it delivers the most benefit in prescription drug area.
 Page 178       PREV PAGE       TOP OF DOC
    Chairman THOMAS. Fine. OK. Mr. Myers.
    Mr. MYERS. I agree that our studies have indicated that the prescription drug benefit is the most desired benefit.
    Mr. SCHUB. As well it is clearly the first benefit that seniors ask for.
    Chairman THOMAS. All right. We have covered the prescription, whatever the dollar amount is. The next dollar would go where?
    Mr. SCHUB. In our studies, based on consumers, they will stack up things such as dental coverage, vision coverage, assuming that you have already eliminated all deductibles.
    Chairman THOMAS. No, no, no. You can spend dollars to buy down the deductibles premium or product.
    Mr. SCHUB. If that was the question, then I misunderstood. The first place we would put dollars is eliminating the cash out-of-pocket expenses.
    Chairman THOMAS. That would be ahead of paid prescriptions?
    Mr. SCHUB. Yes.
    Mr. WIGGINS. Correct.
    Mr. SCHUB. Because that is what eliminates the need for seniors to buy supplemental insurance. And that is what allows——
    Chairman THOMAS. OK. Understanding the way the game is played, the first dollar would be to reduce any out-of-pocket dollar costs, but Mr. Myers, you do not do that in your program?
    Mr. MYERS. Yes, we do that.
    Chairman THOMAS. You do have a dollar reduction?
    Mr. MYERS. That is right.
 Page 179       PREV PAGE       TOP OF DOC
    Chairman THOMAS. Ah. OK. Then I was misinformed. Then paid prescription. Then do you not fall into a second tier of the usual add-ons you find in almost any employer benefit package of a grab bag of vision or dental, that sort of thing?
    Mr. SCHUB. Clearly vision and dental are the top in their preference.
    Chairman THOMAS. And what would be another item outside of that?
    Mr. WIGGINS. Hearing, podiatry, in rank order it would go generally dental, vision, hearing, podiatry, presuming that you are putting aside reductions in copayments.
    Chairman THOMAS. But that would be number one, I think most people would agree, a straight dollar buydown would be the most attractive additional item.
    Mr. WIGGINS. And then prescription drugs.
    Chairman THOMAS. And my assumption is that as we reduce the ability to add these items on, you would drop them in the inverse order? And the last thing you would let go would be the paid prescriptions and the buydown? Because frankly some people are arguing that what is going to occur is what I call the Prop. 13 syndrome. Those of you in California are familiar with it. When we begin putting limits on the property tax expenses, what you might think would occur from the bureaucracy was an elimination of duplication of jobs or high-paying jobs. In fact, what was eliminated was our hours on the library and inaccessibility to parks and other public services, clearly creating an artificial pressure to make change.
    What I am hearing from you gentlemen, which I assume given the universe of the availability of HMOs in front of us today, is that you would deal with the elimination of the benefits, not in terms of a most pain to beneficiary for a reaction point of view, but as I would hope you would do, from a responsible point of view, you would eliminate the secondarily desired benefits first, because I do think if you did that, I would be in the marketplace in the reverse pattern and folks would be signing up with me and not you.
 Page 180       PREV PAGE       TOP OF DOC
    Mr. SCHUB. That is correct.
    Chairman THOMAS. And that is the reason we have hopefully choice in the marketplace. Thank you.
    The gentleman from California.
    Mr. STARK. Thank you, Mr. Chairman, and I am sorry I had to step out while the panel was testifying, but I have had a chance to review the testimony, and I just wanted to comment on the articles I put in the record. I would summarize them by saying that PacifiCare looks bad, Oxford looks boring, and Puget Sound looks good. It was a straightforward search on Lexis. All the bad stuff was not selected. Or the good stuff. This is just what came out of a search asking for the three plans, and I have said this to plans before—you get a lot of anecdotal evidence, but the fact is we operate around here a lot on anecdotes. Our constituents are anecdotes, and if we get complaints, we are often required to act.
    There is a lot of information in here about some of you having to be sued in California to give up information to the State, and it looks like you are hiding stuff from time to time. And I just toss that out as saying it does not make our job any easier when voluntary compliance or voluntary openness does not seem to come very easily. I wanted to ask Mr. Wiggins, I am concerned what happens to beneficiaries if they do not have a chance to drop out and opt back into their Medigap so that they are locked in. It occurs to me that you changed—in the middle of the year—your copays for pharmaceutical benefits; is that correct?
    Mr. WIGGINS. We changed on the annual anniversary of the contract.
    Mr. STARK. Everybody knew that before they signed up again?
    Mr. WIGGINS. Yes.
    Mr. STARK. OK.
    Chairman THOMAS. Let me ask you a question because you may not have been here. They have an A and a B plan based upon what the counties pay them.
 Page 181       PREV PAGE       TOP OF DOC
    Mr. STARK. OK.
    Chairman THOMAS. If, in fact, a question refers to A or B and you kept it for A and did not keep it for B, or you made it equal for A and B, I would like to know that to understand how you would deal with the balancing required with cost factors. So in response to Mr. Stark's question where you increased the copay on the prescription, do you have paid prescription for both A and B?
    Mr. WIGGINS. Yes. And what we did——
    Chairman THOMAS. And did you increase the copayment for both A and B?
    Mr. WIGGINS. What we did is in our A counties, we have a $1,000 prescription limit. We have a $7 and a $15 copayment. And in the B counties, when reimbursement was altered, we went to a $500 limit on the drugs and a $7 and a $20 copayment so it was still $7 for generic, $20 for brand.
    Mr. STARK. This was done in advance of the people renewing? So before they renewed, they knew?
    Mr. WIGGINS. Well, to the degree that we knew. The problem is our notification happened in, I think, October.
    Mr. STARK. Who notified you?
    Mr. WIGGINS. Well, HCFA notifies us of the rate and so we notified, yes, within 30 days, which——
    Mr. STARK. HCFA did not tell you to change your copay; did they?
    Mr. WIGGINS. No. HCFA finally gave us the payment rate that they were going to reimburse us in those counties.
    Mr. STARK. And then you changed your copayment rate on the pharmaceutical plans?
 Page 182       PREV PAGE       TOP OF DOC
    Mr. WIGGINS. Right. As I mention, we make a 1-percent margin in the Medicare.
    Mr. STARK. You did not change salaries or dividends. You changed the copayment rate on the pharmaceutical?
    Mr. WIGGINS. No. We have one of the lowest profit margins in the industry.
    Mr. STARK. That is not what I asked you. I just said you did not change salaries or dividends.
    Mr. WIGGINS. No.
    Mr. STARK. You changed the copayments?
    Mr. WIGGINS. Yes.
    Mr. STARK. You just announced, I gather, that you are going to have a holding period of 4 months for provider bills, is that correct?
    Mr. WIGGINS. A holding period?
    Mr. STARK. You are going to be slow, pay by 120 days, in paying your providers. That was just announced by your plan.
    Mr. WIGGINS. I do not know quite what you are talking about.
    Mr. STARK. OK. Well, in New York, you just announced that providers may now wait 120 days to get paid.
    Mr. WIGGINS. No. Medicare right now we are within the 30-day payment requirement. We did convert to a new computer in September and we did have some delays, which we wrote letters and told everybody, and we did have unusual delays for awhile related to the conversion to a new computer system, which was, it is embarrassing to be in a public hearing talking about it.
    Mr. STARK. To the extent that it causes providers to drop out, you probably should not be allowed to do that, right? Because if I sign up with your plan and you do something that gets my doc mad and he drops out, then I am in trouble and I cannot go to the doc; right?
 Page 183       PREV PAGE       TOP OF DOC
    Mr. WIGGINS. Congressman, we are the fastest growing HMO in the Nation, and we are one of the fastest growing in the Medicare Program, and we had to upgrade our computers.
    Mr. STARK. That is the excuse America On Line uses when I cannot get to them either. So if I can get to my doc, I can get to America On Line; do not trouble me with your problems of getting rich fast.
    Mr. WIGGINS. If the truth sounds like an excuse, I apologize.
    Chairman THOMAS. Wait till Social Security and IRS try to use those arguments.
    Mr. STARK. Yeah. Leonard Schaffer tells me in the Bureau of National Affairs that ''HMOs get a guaranteed profit in high Medicare payment areas such as California.'' Schaffer goes on to say, BNA says, that he suggested the administration should move more quickly than it has proposed to reduce the amount of money Medicare pays HMOs from 95 percent of the local fee-for-service cost to 90. You could easily go to 90 percent, he said. Now, how can Schaffer do that in California and you guys cannot?
    Mr. WIGGINS. Well, what is going to happen if you cut the HMO payment rate is that the differential——
    Mr. STARK. It is going to happen to Schaffer, too; right?
    Mr. WIGGINS. Right. The differential between the fee-for-service and the HMO sector will be exacerbated and so the fee-for-service system will continue to be expensive and have runaway inflation. And fewer people will join the HMOs.
    Mr. STARK. Why can Wellpoint do this? You are missing my question. Why can Blue Cross/Blue Shield, or Wellpoint, whatever they are now, why can they do it and you cannot?
    Mr. WIGGINS. Well, I presume they would cut benefits. I presume their strategy would be the same as everyone's. They would have to cut benefits.
 Page 184       PREV PAGE       TOP OF DOC
    Mr. STARK. But they are recommending it. Why would they do that?
    Mr. WIGGINS. You will have to ask them.
    Mr. STARK. Just to get even with you——
    Mr. WIGGINS. We are not even in the same market.
    Mr. STARK. Why do you think Leonard Schaffer suggests that? I can ask him, I guess.
    Mr. SCHUB. I cannot speak for Mr. Schaffer, but clearly we are concerned that we created an opportunity for Medicare to predictably pay HMOs a budgetable amount that is reasonable going forward, but we are concerned that as we now control a substantial portion of the Medicare lives that those lives are staying with us and they are getting older and they are getting sicker, and we have accepted that responsibility, but any radical changes in the payment system really are probably going to serve to underpay HMOs. As clearly the earlier testimony indicated, we are regressing toward the mean and we may, in fact, exceed it.
    Mr. STARK. One final question, if I may, Mr. Chairman.
    Chairman THOMAS. Sure.
    Mr. STARK. In New York, where it would take a genius to not use an academic teaching hospital, it is pretty hard to find anything else in New York——
    Mr. WIGGINS. We have competitors that must be geniuses then.
    Mr. STARK. You guys have done very well.
    Now, 84 percent of HIP's referrals go to teaching or academic hospitals, compared to only 53 percent of yours. This is in the 40 most commonly used DRGs. Why is HIP doing that, and why are you not? We have come to think that these academic centers are generally better than the average hospital; are you just socking that money away that you save, or do you think that is a better way to deliver medical care?
 Page 185       PREV PAGE       TOP OF DOC
    Mr. WIGGINS. Well, teaching hospitals represent far less than 54 percent of the beds in the markets that we operate and in the geography where our membership resides. So the fact that we are 54 percent means that we are getting migration into the teaching hospitals where people are not going to the community hospitals. I think you will find that 54 percent is a very high percentage of Medicare beneficiaries and overall beneficiaries to be using teaching hospitals.
    Mr. STARK. In New York?
    Mr. WIGGINS. Yes.
    Mr. STARK. HIP has 84 percent.
    Mr. WIGGINS. Well, I am not sure about HIP. I do not know where they get that number given the hospitals that they predominantly use, and I do not know their definition of a teaching hospital versus ours. If Brookhaven and Plain View and Kings Highway are teaching hospitals, maybe that is the case. We do not count them as teaching hospitals. We count Columbia Presbyterian, Montefiore, Maimonides, those types of teaching institutions, in our calculus.
    Mr. STARK. Same list I have here. We are both reading off the same list of what is community and what is academic.
    Mr. WIGGINS. And we pay those hospitals 19 percent more than we pay the community hospitals so it is an important point here on the DME and DSH carve-out. Those moneys that are paid to us are moving through the teaching hospitals, and so if there is a carve-out of DME, somehow the reimbursement rates have to be reduced by all health plans. Otherwise, the hospitals will double dip as they have already done in the New York conversion of the private sector to this same approach of a carve-out.
    Mr. STARK. Thank you.
    Chairman THOMAS. Just a followup so that we can understand. My concern about the potential of applying apples and oranges, and I am concerned about teaching hospitals and whether or not we pay to provide a societal benefit, but having studied the current method, I think it is crazy, and we need to change it fundamentally, and I do not know that pulling it out and sending it through because people do not play the ''game'' the way it is supposed to be played is necessarily the best option, and we are going to try to come up with a different proposal.
 Page 186       PREV PAGE       TOP OF DOC
    I am more interested in product mix. Mr. Schub, you were asked a question in comparison to a quote from Wellpoint. What percentage of your business is Medicare?
    Mr. SCHUB. On a membership basis, it is about 25 percent of our members.
    Chairman THOMAS. And do you have any idea what Wellpoint's is?
    Mr. SCHUB. It is probably much smaller, and they have a very large Medicare supplement business.
    Chairman THOMAS. They do a lot of Medigap and they do some other stuff.
    Mr. SCHUB. Right.
    Chairman THOMAS. And I guess if a significant minority part of my business was what I was in, I could take a 95- to 90-percent cut since it would only affect 5 or 10 percent of my business.
    Mr. SCHUB. That is correct.
    Chairman THOMAS. But if my business was 25 or 30 percent, I would be a little bit more concerned about whether or not I could make that whack. I am always amazed at people who are more than willing, you know, the old, I should say the current statement is what does not kill me strengthens me, but I would much rather take a 5-percent hit to strengthen me than a 25- or 30-percent hit to strengthen me, and I think that is some concern about people who have committed themselves especially since the President's program also talks about lifting the 50/50 requirement. The last thing I want to do is have somebody who is at 10 percent say bring it on and somebody who is now committed to an 80, 90 percent being a little bit more cautious about the changes that are going to be made, and then wanting to know why you are machismo does not match someone else's who would receive only a 10-percent hit to the their total product line versus someone who was immersed at a very high level.
 Page 187       PREV PAGE       TOP OF DOC
    The gentlewoman from Connecticut wishes to inquire.
    Mrs. JOHNSON. Thank you. I regret that I start my own hearing in 12 minutes, and I have to go back and see some constituents between now and then. So I am not going to be able to question, but I do want to thank Mr. Wiggins for his testimony and all of you for being here today. You do represent the possibility of our developing in America a preventive system. There is no way a fee-for-service system can be anything other than an illness system. And there are ways in which managed care systems can be, and I know more about Mr. Wiggins' work in that area, but I assume the rest of you think that way or you would not still be in business.
    But I did want to ask you, Mr. Wiggins, just two things. Would you get me better information about the data? You know I want the detail behind your chart that shows that your people are actually either equally ill or more ill than fee-for-service Medicare recipients, and then I do want to read into the record, because I think this is so very important, I use that to work with others. I also want to know what information HCFA has that if you had it, you could do more too, and we cannot policy without more contemporary data, and one of the problems at the Federal level is a lot of our data is old. One year in this business is old. Two years is very old, and we use a lot of 2-year-old data.
    But I do want to read into the record because I think we are going to have to really focus on this, the fact that while it may make theoretical sense to take out DME, IME and DSH and assume that your payments to teaching hospitals will go down, your experience in New York shows that while that is a good theory, it does not work in practice, and I just want to read into the record your testimony. ''While Oxford has been more successful than our competitors with renegotiating contracts with teaching facilities, we nevertheless had to increase our commercial premiums this year by 3 percent to offset the difference between what we pay into the pool, that is to pay medical education expenses, and what we were able to achieve in contract savings with the teaching hospitals.''
 Page 188       PREV PAGE       TOP OF DOC
    And what you are really saying is that because they feel they are underreimbursed to begin with, they do not give you all that back in a lower hospital rate; is that correct?
    Mr. WIGGINS. Correct. We are the most successful in the market. We have 72 renegotiated contracts with ink dry. Of course, we are forced to do this all very quickly, and Aetna U.S. Healthcare is an example, just announced in their recent conference call to their investors that they had none. So there are many companies that are struggling with this, and having a real tough time figuring out what their 1997 costs are. I think another very important point on this idea of data, Congresswoman Johnson, is I think we could very quickly have very good information. I have offered to pay for this work by an independent organization. I have sent letters to Secretary Shalala saying that we will provide the experts that know how to use risk adjustment models. We can use Arthur Andersen or some independent body to oversee their work and compare that to Medicare fee-for-service. As a starter, just use our 130,000 Medicare enrollees in the 3 States where we have Medicare operating. I think we will be able to show finally good information using inpatient and outpatient.
    What I have here is good inpatient data on all Medicare admissions in metropolitan New York. And if we can just get from the Federal Government that information, we can deliver back to you good risk-adjusted data that shows whether we are getting good, worse or better risk selection than the fee-for-service program, and then we can set policy or you can set policy.
    Mrs. JOHNSON. Thank you very much. Appreciate it.
    [The following was subsequently received:]

New York State CMI Comparison Chart

 Page 189       PREV PAGE       TOP OF DOC

    The data were analyzed using Medicare DRGs (Diagnosis Related Groups). DRGs are disease-based groupings of clinically similar patients that form the foundation of the Health Care Financing Administration's (HCFA's) Prospective Payment System. A DRG is assigned to every inpatient hospital discharge.
    HCFA calculates a Service Intensity Weight (SIW) for each DRG which tells us the relative cost (or ''complexity'') of a particular DRG as compared with a ''typical'' DRG. Thus a 2.000 SIW is roughly twice as complex as the typical case, and a 0.500 SIW half as complex. The average of two or more SIWs is called a Case Mix Index (CMI). The average CMI for the Medicare population in New York in 1995 was 1.429, thus the services provided to this population were approximately 43% more complicated than those delivered to the average population.
    The data presented are for Medicare hospital discharges for Oxford's New York service area (New York City, Long Island and Westchester, Rockland, Orange and Putman Counties). To ensure consistency the comparison uses the same version of HCFA DRGs (version 14) across the entire sample. The dataset used was New York State Department of Health's ''SPARCS'' database for 1995, the most current available. A copy of the UB92 claim for every hospital discharge in NY must be submitted by the hospital to SPARCS. We obtained electronic copies of these through application to SPARCS' Data Protection Review Board. The data are protected and we do not have access to member specific information.


    Oxford Medicare Advantage's CMI is displayed in the left column on the attached graph. Two comparisons were calculated: 1) Medicare discharges identified by the hospitals as Medicare fee-for-service (FFS) (middle column) and 2) Medicare discharges identified by the hospitals as Medicare HMO patients (right column).
 Page 190       PREV PAGE       TOP OF DOC
    The most striking finding is that Medicare HMO patients have a higher CMI than the underlying Medicare population. This suggests that petients in Medicare HMOs are sicker than the Medicare patient population at large, and that, contrary to popular opinion, HMOs are not enrolling the healthier seniors. While Oxford's individual CMI is less complex than that for the entire Medicare HMO population, it is still more complex than the entire FFS Medicare population.
    Furthermore, when we compare the average age of Oxford Medicare Advantage members to FFS beneficiaries in Oxford's New York service area, we find that the average ages of both populations are only months apart. This demonstrates that our membership mirrors that of the general Medicare beneficiary population.
    [The official Committee record contains additional material here.]


    Mr. STARK. Could I just follow up on that one?
    Chairman THOMAS. Sure.
    Mr. STARK. Mr. Wiggins, what you are saying is if you could get the information from HCFA that I want to get, then you could look at that and tell us how you are doing; right?
    Mr. WIGGINS. We have already risk adjusted our entire population using what we think is the best risk adjustment model. It is a system called Symmetry. There are 9 or 10 risk adjustment models.
    Mr. STARK. What you were suggesting to Mrs. Johnson, I thought I heard you say, is that you need to get this information from HCFA that they will not now release it; is that it?
 Page 191       PREV PAGE       TOP OF DOC
    Mr. WIGGINS. Well, I am waiting. I cannot get a response, and I feel that given that you are the primary oversight body to that agency, I thought maybe help would be on the way.
    Mr. STARK. Well, I have got a deal for you. You are being sued in California all over the place for not disclosing your information. So how about you tell us what you are doing, and I will then get HCFA to share with you?
    Chairman THOMAS. Wiggins is in New York.
    Mr. WIGGINS. I am in New York.
    Chairman THOMAS. Yeah, but your company is also in California; right?
    Mr. WIGGINS. Negative.
    Mr. STARK. Mr. Schub.
    Mr. SCHUB. Can I buy into this deal?
    Mr. STARK. You want to get into this deal, Mr. Schub? You want to get out of your lawsuits in California and get all the HCFA information?
    Mr. SCHUB. I am not sure of the question, but the question regarding the lawsuit in California was a closed issue.
    Mr. STARK. You have a bunch of them. I am just saying you are being very cozy with the data in California. You ain't giving it up, and you are being sued to give it up or lobbying to not be regulated. And I am just saying maybe this cuts both ways. I apologize to Mr. Wiggins. I thought he was the bad guy. He is only the boring guy. You are the bad guy.
    Chairman THOMAS. I thought Myers was boring. Puget Sound was boring.
    Mr. STARK. No. Myers is the consumer. He gets the gold star award.
 Page 192       PREV PAGE       TOP OF DOC
    Chairman THOMAS. But I thought your outline was that he was boring.
    Mr. STARK. No, no.
    Chairman THOMAS. Sorry. I am confused now. I thought he was the boring one.
    Mr. STARK. Wiggins is boring. Schub is bad. Myers was great. That is according to Lexis Nexis.
    Chairman THOMAS. Oh, all right.
    Mr. STARK. And that is not a judgment that I am willing to make. I am just saying that is what the computer pulls out when you ask for information. So my question really is moot. Mr. Wiggins, who wants the information, does not have the information I want to give up and trade so we cannot make a deal.
    Chairman THOMAS. We can do a barter here if you guys could work out a deal.
    Mr. SCHUB. The information that was requested in California was not information Mr. Wiggins was referring to. It was actual confidential information on patients' medical records.
    Mr. STARK. Why is that confidential, Mr. Schub?
    Mr. SCHUB. Well, under the Knox-Keene Act, the State requires that we hold personal information about people's health care in confidence and do not just release it to any uncontrolled environment where such information could wind up coming back on people about their healthcare and personal problems. So when requested for literally thousands of files with no backup in terms of why, we asked for protection from the court, and ultimately the Department of Corporations reduced the request from across-the-board thousands of files to 188 files which we fully gave them, and we have not heard from them since.
 Page 193       PREV PAGE       TOP OF DOC
    Mr. STARK. Thank you.
    Chairman THOMAS. You obviously touched a nerve. Let me ask you a question that I do not know the answer to, but you folks in your marketing acumen may have come across. Mr. Wiggins, you offer based upon different counties an A or a B plan in terms of add-on benefits. Mr. Myers, you have chosen not to do that, and my assumption is, Mr. Schub, you do not do that either. To your knowledge, is it permissible under the law to talk about a benefits package which would have a dollar equivalency? And that people could then choose which benefits they want under that dollar amount, a cafeteria plan for benefits, so that not everybody gets the same add-on but picks what they want? Since although maybe paid prescription would be chosen by everyone and you only have $2 left and some people might want dental and others might want vision, instead of arbitrarily picking what the next benefit would be, you would allow folks to choose. Is that legal, to your knowledge?
    Mr. SCHUB. If you are referring to a defined contribution?
    Chairman THOMAS. No. I am talking about the additional benefits above the ACR and AAPCC formula that you either turn back to HCFA or provide in benefits. We played that game earlier where your first dollar would be to buy down costs, your second dollar would be for paid prescription. Do you know if it is legal to offer, say, 100 dollars' worth of benefits, but the individual gets to choose which ones they want, which may be different than someone else. Is that legal?
    Mr. SCHUB. It is and it depends on the way you file your ACR because I know our plan, the recently acquired FHP plan in Arizona has such an option where you can choose, say, the chiropractic or dental or a number of the supplementals, and they are all basically valued about the same.
    Chairman THOMAS. Yes.
    Mr. SCHUB. You can pick whichever one you choose.
 Page 194       PREV PAGE       TOP OF DOC
    Chairman THOMAS. Because if we move to standardization, you could almost pick the benefit in terms of product and then whatever dollar is left over apply to a buydown so that you could get a mix that made some sense to you. I am just trying to think about the offset on standardization.
    Mr. WIGGINS. The risk that would come along with that strategy would be adverse selection significantly into a benefit category so that everyone with a hearing problem chooses the hearing benefit.
    Chairman THOMAS. They would opt for that.
    Mr. WIGGINS. And you lose the concept of insurance which is to spread the risk, and so what you would have a very significant increase in the cost of that benefit because you are assured of 100 percent utilization of the benefit because people that have hearing problems or vision problems would choose it. In the case of PacifiCare, I am presuming that those are capitated benefits, and you cannot do that for very long before your capitated providers realize that the adverse selection is going to work against them.
    Mr. SCHUB. So actuarially——
    Chairman THOMAS. I understand the actuarial and the shared risk concept. I was just looking at benefits that folks really wanted versus benefits that were there because someone thought they wanted them, and I guess if there is a percentage that needs hearing assistance, that it is better under a managed care concept of taking care of the whole person to deal with that than to be more concerned about the pooling aspects of insurance risks. There may come a time when we have a more serious discussion about that based upon what plan moves down the pike. But I do want to thank you folks for your willingness to testify. We oftentimes do not get a real world look at the results that have been produced by virtue of the policies that have been made.
    Thank you very much. The Subcommittee stands adjourned.
 Page 195       PREV PAGE       TOP OF DOC
    [Whereupon, at 1 p.m., the hearing was adjourned.]
    [A submission for the record follows:]

Statement of American Hospital Association

    The American Hospital Association appreciates this opportunity to present our views on Medicare's health plan payment policies on behalf of AHA's 5,000 hospitals, health systems, networks, and other providers of care. The way in which Medicare pays health plans is important to the entire health care community.
    The adequacy of health plan payments undergirds the opportunity for Medicare beneficiaries to select alternatives to traditional fee-for-service Medicare. This fundamental purpose becomes even more significant with the growing role of managed care in delivering health care to seniors—up 36 percent from 1995 to 1996, to 4 million enrolled beneficiaries, according to the Health Care Financing Administration (HCFA). And, the Congressional Budget Office (CBO) estimates that one-quarter of Medicare beneficiaries will receive services from health plans by 2002. This will only be realized if plan payments are adequate to assure that more plan choices are available.
    Health plan payment policy is important to the entire health care community because these payment rates also affect the hospitals, physicians and other caregivers that provide care to Medicare beneficiaries under contract to Medicare health plans today. If health plan payments are set unrealistically low so that they cannot accommodate reasonable costs of delivering care, then hospitals and physicians will find it increasingly difficult to participate in health plans.
    Finally, the adequacy of health plan payments is vital to ensuring that a greater range of private-sector health plan options becomes available to Medicare beneficiaries. The AHA believes that provider-sponsored organizations (PSOs) should be among the new options. PSOs are community-based, integrated networks of physicians, hospitals and other caregivers that provide directly a spectrum of coordinated care in exchange for a fixed capitated payment. PSOs achieve the cost efficiency necessary to hold down health care costs by directly managing both the utilization of services and the cost of producing those services. They hold particular promise in serving rural areas, which have not been attractive to HMOs. However, the success of new health plan options, including PSOs, will depend in part on the appropriateness of payments under Medicare.
 Page 196       PREV PAGE       TOP OF DOC
    Unfortunately, the current method used by HCFA to pay Medicare health plans—the adjusted average per capita cost, or AAPCC—is seriously deficient in several ways.

Current AAPCC Shortcomings

    The AAPCC's payment calculations are based on regional fee-for-service costs generated from Medicare claim files. Given the wide variations in historic fee-for-service utilization patterns, there is a resulting wide variation in AAPCC payment—more than 300 percent among counties across the United States. In some counties where payment rates are particularly low, the AAPCC appears to pose barriers to health plans being offered to seniors at all.
    A further problem: today the AAPCC includes money which the fee-for-service system earmarks for graduate medical education and for those hospitals treating a large volume of low-income individuals—the disproportionate share hospitals. Health plans, however, are not required to spend these funds for these purposes. Plans can use them to support added services or additional profits rather than essential provider missions. There is strong experience indicating that these designated funds, currently paid to Medicare health plans, do not reach the hospitals that actually incur these costs.

Proposed Solutions

    The AHA believes that four objectives should provide the framework for adequate and equitable Medicare health plan payments:
    •  Promote provider and health plan efficiency
 Page 197       PREV PAGE       TOP OF DOC
    •  Promote personal responsibility by beneficiaries
    •  Provide administrative simplicity
    •  Maintain some beneficiary choice in plan and provider
    The AHA Board of Trustees has adopted the following principles for Medicare's health plan payments:
    1. Medicare should be viewed primarily as a program that provides access to specific benefits rather than an income transfer program like Social Security. Medicare should remain a program for the provision of medical services for seniors and the disabled and should not be transformed into a program that provides cash transfers or income supplements which can be used for non-medical services.
    2. As a health benefits program, Medicare should be based on a standard minimum benefits package. As the Medicare program evolves, it should continue to be a program that assures access for seniors and the disabled to a minimum set of benefits. The Medicare payment should allow plans to offer that minimum set of benefits. Any additional premiums from beneficiaries should reflect their choice for a preferred style of medicine or for added benefits.
    3. Payments should be equal per person but adjusted to reflect regional input price differences and utilization that result from differences in beneficiary health status—for example, disability or renal dialysis. The current system by which Medicare health plan payments are determined results in enormous variation in payment rates from county to county throughout the United States. Medicare health plan payments should be uniform across the country, but then adjusted for differences in the cost of delivering health care services that are beyond the control of the health care system. Specifically, health plan payments should be adjusted to reflect differences in the cost of labor, supplies, equipment, and of other goods needed to provide care. Health plan payments also should be adjusted to reflect differences across the country in the cost of delivering care due to the fact that some areas may care for less-healthy, more costly Medicare beneficiaries.
 Page 198       PREV PAGE       TOP OF DOC
    AHA has developed a number of suggestions for designing and implementing a new payment method to achieve these goals:
    •  The current AAPCC should be blended with a new payment rate that eliminates differences in historical patterns of use across counties. We believe the transition period to a new, more equitable, methodology should be as brief as possible, in order to encourage rapid development of health plan options. A ''blend'' approach over two to three years would minimize payment variation and provide an interim point at which the impact of the change could be evaluated—and any needed adjustments made. In addition, a payment floor should be quickly established to raise payments in the lowest-rate areas.
    •  Geographic price differences should be addressed by accommodating real price differences—but not inefficiency. Various measures should be tested and the best selected on the basis of sound policy judgment.
    •  In regard to risk adjustments: don't delay changing the AAPCC because we lack good adjustments today. HCFA, in the Medicare Choices demonstration, is in the process of testing several methods. Research must continue, but we can't wait to begin changing health plan payments until it is complete.
    •  Payments should continue to be made on a county basis. The blending of current and new payment methodologies will dramatically reduce existing differences in county payment rates.
    4. Clinical education and disproportionate share payments should be removed from Medicare health plan payments and paid directly to organizations serving those responsibilities. The Medicare program traditionally has paid hospitals directly for the special, additional costs associated with teaching and with treating large numbers of low-income individuals. As these special payments remain buried within a fixed, Medicare health plan payment, health plan organizations receiving the payment are not passing on the funding to those institutions actually incurring the added costs. Medicare payments for clinical education and for hospitals treating a disproportionately large share of low-income individuals should be removed, or ''carved out,'' from the Medicare health plan payment and paid directly to the organizations fulfilling those responsibilities.
 Page 199       PREV PAGE       TOP OF DOC
    5. Medicare's private sector alternatives to traditional fee-for-service Medicare must include integrated, at-risk provider-sponsored organizations (PSOs). As the Medicare program evolves and seeks to make more private-sector health plan options available to seniors, those options must include PSOs. PSOs are community-based, integrated networks of providers that would offer comprehensive care in exchange for a fixed payment from the Medicare program. PSOs offer an alternative to fee-for-service care for seniors—one that holds the promise of better, more coordinated care; one that dedicates more of the health care dollar to direct patient care rather than to insurer overhead and profits; and one that assures seniors can continue to get care from the very hospitals and health care professionals with whom they already have long-standing and caring relationships.
    6. Medicare's share of the cost of coverage should be income-related so that available funds will provide an adequate minimum package for those with little or no personal funds for their coverage. As health care costs continue to rise and baby boomers begin to retire, the dollars available to fund Medicare will become even more constrained than today. To better allocate Medicare dollars, Medicare's share of the cost of coverage should be income-related—Medicare would pay for the full costs of coverage (excluding required copayments and deductibles) for lower-income seniors and would pay a very limited amount for upper-income seniors. Mechanically, income-relating Medicare coverage could be achieved in a variety of ways. The effect, however, is the same—to assure that limited Medicare dollars can continue to provide adequate care for those most in need.
    7. To facilitate the best use of limited Medicare funds, federal tax policy should allow for a tax-sheltered savings vehicle by which individuals could save personal funds during their work life to use to purchase health care coverage during retirement. Following from the principle above, if Medicare's share of the cost of coverage becomes income-related, individuals should have the opportunity to save during their working years and to set aside funds that could be used to defray medical expenses during their retirement. People can be encouraged, through tax-preferred vehicles, to save their own funds during their working years for future health care. These savings vehicles should not be confused with Medical Savings Accounts, which are designed to allow already-retired seniors to purchase a high-deductible, catastrophic coverage plan and accumulate tax-preferred savings to defray any non-covered health care costs. The tax-preferred savings vehicles encompassed by this principle are for working individuals and could be combined with any type of health coverage plan.
 Page 200       PREV PAGE       TOP OF DOC

The President's Budget Proposal

    President Clinton has proposed reducing Medicare spending by $100 billion as part of his five-year balanced budget plan. Of these reductions, $34 billion comes from reductions in Medicare health plan spending, achieved in three ways:
    •  The largest amount, $18 billion, results because the fee-for-service reductions to hospitals, physicians and other providers proposed in the President's budget are automatically incorporated into the AAPCC rates. Lowering fee-for-service payments to providers automatically lowers AAPCC payments to health plans, because the latter is based on the former. Congress should seek to ''decouple'' Medicare health plan payments from Medicare fee-for-service payments.
    •  An additional $10 billion in reductions results from the ''carve out'' from Medicare health plan payment of graduate medical education and disproportionate share costs incurred by hospitals. These funds would now go directly to those institutions incurring these costs. AHA supports this change, which is consistent with the principle outlined above. Without this carve out, there are no assurances that health plan organizations receiving the payment will pass on the congressionally earmarked hospital funding to those institutions actually incurring the costs.
    •  The President's plan also removes $6 billion from Medicare health plan payments. The Administration's rationale is that today's payment, pegged at 95 percent of adjusted average per capita costs, overpays health plans because they have attracted healthier seniors whose care costs less than current payments. This adjustment is a one-size-fits all arbitrary proposal. It penalizes all plans without regard to the actual health status of each plan's enrollees. Rather than a fixed percentage reduction, the appropriate payment should be made through the design and use of a better risk adjustment mechanism to be applied according to the characteristics of each health plan's enrollees, rather than through an arbitrary ratcheting down of the rate, adversely affecting all health plans.
 Page 201       PREV PAGE       TOP OF DOC
    The American Hospital Association is deeply concerned about the impending Medicare financing crisis. Action must be taken soon to make fundamental structural changes that will allow this nation to continue to meet the health care needs of the elderly. Repairing the AAPCC's present variability and unpredictability should be a part of this effort. Broadening beneficiaries' choice of Medicare health plans, including PSOs, is another.
    Overall, the Medicare program has been an outstanding success in bringing health care security to the elderly. In a nation where eroding access to health care coverage in the working population is already contributing to a steady rise in the uninsured, we cannot afford a future in which we lack the resources to keep the Medicare promise. Addressing the current shortcomings in the AAPCC payment rate can be an important step in keeping that promise.

(Footnote 1 return)
B. Dowd and others, ''An Analysis of Selectivity Bias in the AAPCC,'' Health Care Financing Review, Vol. 17, No. 3, (1996) found no favorable selection in 1988 data for the Twin Cities area in Minnesota. A Price Waterhouse study, sponsored by the Association of American Health Plans, found little favorable selection. Price Waterhouse estimated that HMO enrollees in 1992 were 2 percent less costly than fee-for-service beneficiaries (Is There Biased Selection in Medicare HMOs? (Washington, D.C.: Price Waterhouse, Mar. 1996)). Noting several technical weaknesses in this study, however, the Congressional Budget Office (CBO) has concluded that the Price Waterhouse estimate understates favorable selection by a factor of 4. (That is, according to CBO, enrollees' costs are at least 8 percent lower than fee-for-service beneficiaries'.) For a review of recent studies including the Price Waterhouse report and the CBO critique, see ''Policy Implications of Risk Selection in Medicare HMOs: Is the Federal Payment Rate Too High?'' Issue Brief, No. 4, Center for Studying Health System Change (Washington, D.C.: Nov. 1996).

(Footnote 2 return)
For a list of related studies as well as an analysis of 1989-94 data, see Annual Report to Congress, Physician Payment Review Commission (Washington, D.C.: 1996), ch. 15.

(Footnote 3 return)
We calculated this cost ratio from data underlying MEDICARE HMOS: HCFA Can Promptly Eliminate Hundreds of Millions in Excess Payments (GAO/HEHS–97–16, Apr. 25, 1997), p. 38, figure II.2.

(Footnote 4 return)
GAO/HEHS–97–16, Apr. 25, 1997, p. 38.

(Footnote 5 return)
Estimates of self-reported health status are from an analysis of the Medicare Current Beneficiary Survey for 1994. See Gerald Riley and others, ''Health Status of Medicare Enrollees in HMOs and Fee-for-Service in 1994,'' Health Care Financing Review, Vol. 17, No. 4, (1996) pp. 70-71.

(Footnote 6 return)
Estimates of self-reported health conditions are from the article cited in footnote 5.

(Footnote 7 return)
Beneficiaries were asked whether they had any of these conditions: heart, hardening of arteries, hypertension, cancer (nonskin), stroke, arthritis, diabetes, osteoporosis, emphysema, mental disorders, vision problems, hearing problems, Parkinson's disease, amputations, skin cancer, Alzheimer's disease, mental retardation, broken hip, and partial paralysis.

(Footnote 8 return)
GAO/HEHS–97–16, Apr. 25, 1997, p. 42.

(Footnote 9 return)
Competitive medical plans (CMPs) are HMOs that have not chosen to become federally qualified but meet similar federal standards. For the remainder of the testimony, we use the term ''HMO'' to refer to both HMOs and CMPs.

(Footnote 10 return)
Gerald F. Riley, et al, ''Stages of Cancer at Diagnosis for Medicare HMO and Fee-for-Service Enrollees,'' American Journal of Public Health, 1994; 84:1598–1604.

(Footnote 11 return)
David Carlisle, et al, ''HMO vs. Fee-for-Service care for Older Persons with Acute Myocardial infarction,'' American Journal of Public Health, 1992; 82: 1626–1630.

(Footnote 12 return)
Riley, et al, ''Disenrollment of Medicare Cancer Patients from Health Maintenance Organizations,'' Medical Care, August 1996.     •  Secure Horizons of California has one of the lowest disenrollment rates among health plans, an average of 0.8%, which reflects the monthly figure of those who voluntarily disenroll. This low disenrollment rate also reflects Medicare beneficiaries' satisfaction with their HMO selection.

(Footnote 13 return)
Wickizer and Feldstein, ''Does HMO Competition Reduce Health Insurance Premiums? Recent Evidence,'' Medicare Practice Management, August 1996.

(Footnote 14 return)
Gaskin and Hadley, ''The Impact of HMO Penetration on the Rate of Hospital Cost Inflation, 1984–1993,'' Georgetown University Medical Center, 1995.

(Footnote 15 return)
Laurence Baker, ''Managed Care, Health Expenditures and Physician Income,'' Princeton University Ph.D. dissertation, 1994.