Fifth Meeting,
Wednesday, November 15, 2000
Attendance:
Members of the Technical Panel:
Dale Yamamoto, Chair of the Panel
Principal
Benefits Practice
Hewitt Associates
Lincolnshire, IL
James Robinson
Senior Scientist
Center for Health Systems Research and Analysis
University of Wisconsin-Madison
Madison, WI
Michael Chernew
Associate Professor
Department of Health Management and Policy
University of Michigan
Ann Arbor, MI
David Cutler
Professor of Economics
Department of Economics
Harvard University
Cambridge, MA
Len Nichols
Principal Research Associate
Health Policy Center
The Urban Institute
Washington, DC
Consultant to the Panel
Sam Gutterman
Director and Consulting Actuary
Price Waterhouse Coopers
Chicago, IL
Staff of the Health Care Financing Administration, Office of the Actuary
Richard Foster, Chief Actuary
Sol Mussey, Director of the Medicare and Medicaid Cost Estimates Group
John Shatto, Actuary
John Wandishin, Actuary
W. Kent Clemens, Actuary
Mark Freeland, Deputy Director, National Health Statistics Group
Stephen Heffler, Deputy Director, National Health Statistics Group
Jackie Carroll, Executive Officer
Donna Holt, Secretary
Staff of the Office of the Assistant Secretary for Planning and Evaluation:
Christy Schmidt, Deputy to the Deputy Assistant Secretary for Health Policy, ASPE, DHHS
Ariel Winter, Executive Director of the Panel, ASPE, DHHS
Gene Moyer, Consultant to ASPE and to the Panel
Others Present:
Mike Andrews, Social Security Administration
Phil Ellis, Economist, Treasury Department
Linda Baker, General Accounting Office
Robert Nguyen, Congressional Budget Office
Ralph Monaco, University of Maryland
Proceedings of the Meeting:
The meeting was called to order by Mr. Yamamoto, who welcomed the group and recognized Ms. Schmidt.
She said that she had talked to the Secretary and the White House and that she had been directed to extend her appreciation to the Panel for their hard work, their good thinking and documentation, and their great suggestions for the future Trustees Reports.
Mr. Winter then told the Panel that Mr. Yamamoto would be asking them to approve the minutes for the September and October meetings. Most of the rest of the day would be devoted to discussions of the draft final report and the recommendations. During the remainder of the day, Sally Burner and Greg Savord of the Office of the Actuary were to give a presentation on the way the Medicare drug proposal cost estimates were developed.
He also announced that the Trustees Meeting would be held on November 30 and that this would be the first meeting for the new Trustees, John Palmer and Thomas Savings, recently confirmed by the Senate.
At that meeting, the Secretary would be presenting the Panel's recommendations. Further, on the day before the meeting, the Secretary was to be briefed on the recommendations by the staff and by Messrs. Yamamoto and Nichols as well as any other members of the Panel who are able to come.
Mr. Winter also thanked the Panel on behalf of the Secretary. He further thanked the staffs of the Office of the Actuary and ASPE for their analyses and administrative work on behalf of the Panel.
Then he turned the meeting over to Ms. Schmidt and then to Mr. Foster for additional remarks.
Ms. Schmidt then thanked Mr. Winter for his hard work as he headed off on leave to take care of his daughter while his wife returns to work.
Mr. Foster then thanked the Panel for their open minds, their analytical abilities, and their thoughtful processes in helping to make the important Trustees Reports even better than before.
He indicated that he felt that the report of the Panel was in relatively good shape, in spite of the fact that there were some things to talk about, and announced that he had hired an editor, Cathy Curtis, some time ago to help the Office of the Actuary with their written materials and that he had "volunteered" her to edit the report of the Panel.
Mr. Yamamoto added one more "thank you" which he said must be the third or fourth offered that day. Further, he said that he had not really felt two months before that the report would be as far along as it was and that he thought they would have a very good product at the end.
Then he asked for a motion to approve the minutes from the last two meetings. Mr. Cutler made the motion. Mr. Chernew seconded it. It was passed unanimously.
Then he suggested that the Panel go through each of the recommendations not approved at the last meeting. Afterward, the Panel could go back and discuss any remaining finer points of previously discussed recommendations.
Mr. Cutler suggested that they just go through each recommendation in order. If there was nothing about a specific recommendation to discuss, they could just say so and go on.
Mr. Winter pointed out that the list was in the folder previously given to each Panel member.
List of recommendations and findings discussed by Panel:
Current Methodology
Finding I-1: The actuaries should continue to use a tripartite division of the 75-year interval into a short-term, an intermediate-term, and a long-term, with the intermediate-term being a transition between the short-term and the long-term.
Recommendation I-2: Cost growth rate assumptions in the intermediate-term and long-term should be established for HI as a whole and SMI as a whole, and then partitioned into specific services as needed and appropriate.
Finding I-3. The Panel believes the differentiation of the forecasts into a price component and a quantity component is reasonable and appropriate.
Recommendation I-4: The Panel recommends that for the short run forecast, the quantity and price components of spending for each service category should be modeled explicitly, with total spending by category being computed as the product of a price component and the quantity component.
Recommendation I-5: The Panel recommends viewing short-term forecasts in a two-part framework, with one part being an underlying growth trend and a second part being a set of service-specific shifts in spending levels away from the overall trend.
Recommendation I-6: The Panel recommends that assumptions regarding shifts in site of care should be explicit and consistent across models. This is especially important for shifts between HI and SMI, such as movements from inpatient to outpatient care settings.
Recommendation I-7: The Panel recommends that models of utilization be based on age and gender utilization patterns for all service categories.
Recommendation I-8: The Panel recommends that, within each age/gender cell, distinctions among spending for decedents and survivors should be considered.
Recommendation I-9: The Panel recommends that explicit health status models be considered for eventual incorporation into the projection methodology.
Recommendation I-10: The current method used to estimate future rates of change in managed care penetration is reasonable, although ongoing research is necessary to provide a more rigorous forecasting model.
Recommendation I-11: The Panel recommends that the allocation of aggregate managed care enrollment to age and gender cells be based upon a cohort-based model using estimated transition probabilities for managed care enrollment and disenrollment.
Recommendation I-12: The OACT forecasting models should assume that HI and SMI expenditure growth is not affected by changes in managed care enrollment.
Recommendations I-13: The Panel recommends that the model documentation should be made more explicit regarding key underlying assumptions and the reasons for various adjustments (e.g. legislation) to historical trends.
Short Term Assumptions
Finding II-1: The panel believes that the short-term forecast of price changes is reasonable.
Recommendation II-2: The panel recommends setting the technological component of the growth rate such that age-adjusted, real, per beneficiary spending growth exceeds the growth of real, per capita GDP by 1 percent per year.
Recommendation II-3: The panel recommends a trend in the next decade towards lower inpatient hospital utilization and increased utilization of ambulatory care services.
Recommendation II-4: The panel recommends that, as a general principle, forecasts of gaming [note the restriction to gaming] recognize that transitory utilization or case mix changes are partially offset over time.
Recommendation II-5: The panel recommends that the case mix increase for inpatient hospital, skilled nursing facility, home health, and hospital outpatient services paid on a PPS basis be projected to have an underlying technological component increasing at 1 percent per year.
Recommendation II-6: The panel recommends that the case mix index for inpatient hospital services be forecast to increase above the steady-state level before a new steady state is reached.
Finding II-7: The panel finds reasonable the assumptions about first year increases in case mix for home health services and hospital outpatient services. The panel further finds reasonable the assumption that case mix increase for these services and skilled nursing facilities will decline to a steady state level over a period of about 15 years.
Recommendation II-8: The panel recommends a forecast of modest reduction in use of inpatient hospital services and an increased growth rate of outpatient services as a result of continuing technological changes.
Finding II-9: The panel finds reasonable the assumption of no growth in utilization of home health and skilled nursing facilities after the transition period.
Recommendation II-10: The panel recommends an increased growth rate of physician services and the components of SMI spending not paid on a prospective basis of 0.5 percent per year, consistent with an increase in the growth of the case mix index.
Finding II-11: The panel finds reasonable the current behavioral offset assumptions.
Long Term Cost Growth
Recommendation III-1: For the period extending from 25 to 75 years into the future, real, age adjusted, per beneficiary expenditures for both SMI and HI should be assumed, in the intermediate projection, to grow at a rate 1 percentage point above real, per capita GDP growth.
Stochastic Forecasting and Uncertainty
Recommendation IV-1: OACT staff should evaluate historical experience on a regular basis to identify the sources and causes of changes in health care costs.
Recommendation IV-2: The Panel recommends that, to the extent possible, the Trustees should regularly evaluate and report on the accuracy of past HI and SMI projections over various periods, for example, over one, five, and ten year timeframes.
Finding IV-3: The current one year decomposition by source is reasonable, although it could be enhanced through further analysis of contributing causes.
Finding IV-4: The Panel's overall impression of the model and its application leads it to believe that it is a reasonable and prudent initial application of stochastic modeling.
Recommendation IV-5: The Panel recommends that the stochastic model used in the 2000 Trustees' SMI Report be further developed and enhanced and other stochastic models should be developed for SMI and HI.
Recommendation IV-6: The Panel recommends that the difference in health care costs incorporated in Alternatives I and III compared with those of Alternative II should be relatively wider in the short-term, i.e., the next five years, to recognize the greater uncertainty in the assumptions for shorter time periods and relatively wider in the very long-term (more than fifty years from the valuation date) to recognize the extreme uncertainty associated with forecasting during such a period.
Recommendation IV-7: The Panel recommends that a new set of indicators should be developed that focus on sources of uncertainty and sensitivity of health care projections with respect to underlying factors that can significantly affect future costs, including the effects of managed care, under alternative scenarios.
Finding IV-8: The current measures of actuarial balance of the HI and SMI Funds reasonably satisfy the objectives of actuarial balance. (Appendix A of the Trustees' HI Report describes the "Modified Average-Cost Method", which measures actuarial balance in a different way for the HI Trust Fund, currently resulting in the same conclusions with respect to funding of HI and SMI.)
Recommendation IV-9: The Panel recommends that a set of indicators should be developed to assess the size of estimated SMI cost growth.
Research and Presentation
Recommendation V-1: More resources should be devoted to insuring that quarterly cost and utilization data are available with a shorter time lag for analysis by the Office of the Actuary (OACT).
Recommendation V-2: More resources should be devoted to serving the analytic research priorities of the OACT and the Trustees. Without being prescriptive, this increased resource commitment should likely include both additional staff inside OACT as well as some targeted reallocation of research funds outside OACT that can address Trustee Report-related questions of paramount importance.
Recommendation V-3: There should be greater coordination of relevant research projects and agendas across the many agencies of the Department (AHRQ, HRSA, CDC, ASPE, and HCFA(now known as CMS) itself) and between the Department, the congressional support agencies (MEDPAC, GAO, CRS, CBO), private foundations (e.g., the Robert Wood Johnson Foundation, the Kaiser Family Foundation, the Commonwealth Fund, etc.) and public and private research organizations (e.g., Rand, Urban, Mathematica, universities, etc). This coordination effort may be led by a governmental entity and this will require new resources, but to be successful the effort must extend beyond the executive branch and include congressional support and private sector representatives as well. The purpose of this coordination is to produce and manage a research agenda that is focused on key behavioral responses that drive cost and quality in the Medicare program.
Recommendation V-4: The clarity of the reports' presentations would be enhanced considerably with a number of specific changes and additions. They include:
A. Emphasize cost per beneficiary, or cost per worker, rather than cost per unit of service.
B. Both the HI and SMI reports should have a section with consolidated reporting of HI dollars per beneficiary, SMI dollars per beneficiary, and total dollars per beneficiary.
C. The discussion of some key tables -- II.F.1 in each report is a salient example -- should include an explanation of how the columns relate to (and may be derived from) each other.
The Panel then discussed the Findings and Recommendations in Section I: current methodology. In general, they left the wording unchanged, although they agreed to send comments to the original authors after the meeting, and after the authors had made changes or disagreed, to have the authors send the revised wording to the professional editor, and then to look at them again after the professional editor had done her work. They agreed to delete I-11 on the grounds that it was redundant with I-5 and then decided to shorten it by leaving out the final clause. Finally, they decided to circle the recommendation and to come back to it after Mr. Robinson, Mr. Chernew, and Mr. Gutterman had done some work on it during the lunch break.
The Panel then turned to I-12. After considerable discussion, the Panel agreed to split it into two parts, one of which was "The Panel recommends the development of an explicit managed care shift based on selection of enrollees into and out of managed care." The second part, after much discussion was to read "The Panel recommends an explicit assumption of spillover effects be incorporated in to the projections." This was to become I-13. I-13 was to become I-14.
In the new I-14, there was discussion about whether the "e. g." clause belonged there, and they decided to delete the clause and to reference legislation in an appendix.
This brought up the question of an appendix of equations supplied by Mr. Chernew. The Panel decided that whether or not they included it in the report, the appendix should be made available to the public, perhaps on the web. They further decided to transmit it to Mr. Foster who would see that it was made available in some way that he could determine.
At this point the Panel took a short break.
Upon returning, the Panel began to discuss Section II on Short Term Assumptions. After some additional discussion of some new equations, the Panel went through the assumptions one by one. Finding II-1 was approved. Recommendation II-2 was turned over to Mr. Winter for rewording. Recommendation II-3 was also approved. Recommendation II-4 was approved with the proviso that the material in brackets on "gaming" be taken out. Mr. Cutler, however, objected and the Panel decided to change the wording to "The Panel recommends that, as a general principle, forecasts of transitory shifts such as gaming of payment systems are partially offset over time." There was still some question about the wording, but they finally decided to allow Mr. Cutler to determine the final wording.
There was also a question about the relationship of this recommendation to I-5 on short-term forecasts, but they decided to work on that issue in the afternoon.
Mr. Cutler agreed to work on the wording of II-5 and on whether the order should be changed. Further, he agreed to consider whether II-6 should be combined with II-3.
Then the Panel turned to Finding II-7 and approved it
The Panel then agreed to incorporate Recommendation II-8 and Finding II-9 into II-3.
They agreed to strike II-10 and to leave Finding II-11 as it was with the proviso that there might be some changes made by Mr. Robinson.
The Panel then turned to Section III and decided to leave it as it stood.
Recommendations IV-1 and IV-2 were approved without a lot of discussion. The Panel felt that recommendation IV-3 needed some rewording. Recommendation IV-4 was approved with the wording "The Panel finds that the stochastic models presented in the 2000 Trustees SMI Report are a reasonable and prudent initial application of stochastic modeling." Recommendation IV-5 was approved with the wording "The Panel recommends that the stochastic model for SMI be further developed and enhanced, and that other stochastic models should be developed for HI."
Mr. Gutterman and Mr. Yamamoto decided to try to rewrite IV-6 over lunch.
Recommendation IV-7, Finding IV-8, and Recommendation IV-9 were approved.
The Panel turned to Section V recommendations. V-1 and V-2 were interchanged at Mr. Chernew's suggestion, but approved otherwise. Ms. Schmidt asked that V-2 be changed to include the words "and research funds" so that the last phrase would say "this increased resource commitment would include both additional staff and research funds inside OACT as well as..." After discussion, the Panel decided on " Without being prescriptive, this increased research commitment should likely include both additional staff and research funds inside OACT as well as additional research funds outside OACT." Mr. Nichols said he would work on it over lunch, but the Panel approved the wording subject to whatever changes Mr. Nichols decided to make.
The Panel felt that V-3 was too long, especially that the examples needed to be moved into the text. Mr. Nichols said he would also work on this over lunch and bring new wording to the afternoon session.
The Panel approved V-4 and turned to the summary. After considerable discussion, they decided to break to allow everyone to pick up lunch and to bring it back to the meeting for a working lunch. At 12:45, the panel recessed until 1:35.
After reconvening, Mr. Yamamoto suggested that they return to the findings and recommendations individuals had worked on over the lunch break.
Mr. Robinson provided new wording for I-11 and after discussion, the Panel adopted the wording "The Panel recommends that projections of managed care enrollment be disaggregated into age, gender, and geography cells." The new II-12 and II-13 were "The Panel recommends a trend in the next decade towards lower inpatient hospitalization, increased utilization of ambulatory care services, and no change in utilization of home health and skilled nursing facilities" and "The Panel recommends setting the trend growth rate such that real age- and gender-adjusted, per-beneficiary spending growth exceeds the growth of real per capita GDP by 1 percent." II-12 was going to replace II-3, II-8, and II-9. All the Section II findings and recommendations were to be renumbered.
New wording for Section IV included IV-3 and IV-4. IV-3 had been re-written by Mr. Cutler as "The current one-year decomposition of the change in actuarial balance for HI and the program costs for SMI, by source, is reasonable, although it could be enhanced through further analysis of contributing causes."
There was more discussion of Section IV, but no more rewording. In Section V, the Panel did away with subsection A in V-4 .
The Panel then turned to Section III, but basically commented on the text rather than the recommendation since they had already approved it.
The bulk of the discussion concerned the rate of growth of trust fund spending, income, and of the health labor force, particularly the effect on overall productivity and whether historical growth rates could be sustained in the future. This was especially germane to the recommendation that real Medicare growth be set equal to real GDP growth plus 1 percentage point.
At this point, Mr. Freeland, Mr. Monaco, and Mr. Heffler were invited to join the discussion.
Mr. Freeland presented some estimates that showed that the growth rate in real health spending per health worker was not much different than that for real GDP per worker.
Mr. Cutler then brought up some measurement issues in measuring productivity of health workers. These basically concerned price measurement, quality measurement, and the resulting negative productivity growth. Mr. Freeland and Mr. Heffler agreed, but said that BLS and BEA had been doing a lot to correct these problems in the price indices with some success.
One of the major issues discussed was whether the workers remaining in the non-health sector can show sufficient productivity growth to meet the GDP goals of the society.
The Panel finally decided to leave the recommendation as it was.
Then Mr. Yamamoto declared a five minute break during which people might re-read Recommendation I-12 on dealing with managed care growth in the models, which Mr. Chernew had drafted
After the break, the Panel read through some findings and recommendations discussed earlier that the staff had typed. There were several comments, but finally the Panel decided to wait for the editor to finish her work before making major changes. There was also a question of Ms. Rosenblatt's feelings about the list. Mr. Yamamoto said he would give her a call when he returned home. Further, Mr. Yamamoto indicated that all further edits had to be transmitted no later than Monday, November 20 at 8 A. M. EST so that Mr. Winter could pass them on to the editor. There was also a discussion of what needed to be available for the Secretary's briefing on November 27 and for the Trustees meeting on November 30.
At this point, Mr. Yamamoto recognized Sally Burner and Greg Savord to discuss their recent efforts in constructing a model for estimating the cost of providing drug coverage for Medicare beneficiaries.
Ms. Burner began by discussing the model. The model was based on the 1996 Medicare Current Beneficiary Survey File. They supplemented the data with such items as a geographic cost index and the number of HMO's in the area as an index of penetration. They are working on merging a data set with all the Medigap policies in the U. S.
They "age" population on the file to a future year, currently 2002 to 2007, and alter the HMO/Fee for Service mix in the process of altering the weights. They age the expenditures using Trustees Report or budget assumptions depending on which they are working on at the time. In the process, they break up certain expenditure types so that necessary variables will be available. By a separate process, they estimate the correct number of home health and SNF users. They also adjusted the incomes on the file to match the income distribution shown by the Office of Tax Analysis for persons aged 65 or older and to get the right amounts of taxable Social Security benefits.
There was an issue on the growth rate for drugs which was lower than that used by CBO. In the next round, both CBO and OACT drug estimates will probably go up.
They attributed spending, inpatient day, and SNF day values to the institutionalized population.
Finally, they adjusted for under-reporting originally using a 30% factor, then revising it to a 15% factor.
They compared the estimates to the 1996 Medical Expenditure Panel Survey values and found they had about the same user rates, but lower expenditures per user. Thus when they did the under-reporting adjustment, they increased expenditures for those with expenditures to meet the control values. They did this according to a regression line they had developed which allowed them to conform to the Medicaid budget target as well as to a total spending target.
Mr. Savord then discussed the way the model was used to get the Medicare drug estimates. They also used the MCBS operated on the mainframe. They added up what people were currently paying for drugs less what Medicare was currently paying. They also used a discount factor of 12.75% which was developed a long time ago and which no one really understands. This factor, however, was varied by size of employer since the feeling was that large employers were already getting the discount.
Then they applied the plan deductibles and copayment rates to arrive at the amounts the beneficiary and the plan would actually pay as well as any wrap-around payments made by employer insurance, especially for the working aged for whom Medicare is a secondary payer. They apply an induction factor for the effect of lower prices on utilization and cost and re-estimate the out-of-pocket and plan payment amounts as well as a subtraction for Medicaid payments to get the appropriate Federal share of costs. There was also a calculation for an employer subsidy in which Medicare and Medicaid would subsidize employer plans under the proposal.
When Mr. Savord finished, Mr. Yamamoto thanked him and Ms. Burner.
Mr. Yamamoto adjourned the meeting at 5:30 P. M.