the technological intensity of service-that cannot be rationally explained by analytically derived considerations. Rather, they are founded on historical patterns of charges, and do not reflect demonstrable variations in the cost of producing the service or the physician's skill.

These problems have been of concern to the committee for several years, as evidenced by numerous hearings, several mandated studies, and various legislative initiatives directed toward eventual systematic reform. These problems have also been well documented by several studies and published reports, by both governmental and non-governmental agencies and parties. Among the most notable governmental studies are those from the Office of Technology Assessment (Payment for Physician Services: Strategies for Medicare Reform, February 1986) and the Congressional Budget Office (Physician Reimbursement under Medicare: Options for Change, April 1986) and the three annual reports from the Physician Payment Review Commission (see, in particular, Annual Report to Congress, March 1989).

The most recent Physician Payment Review Commission (PPRC) annual report documents the progress that has been made in developing the concepts, data and analysis needed for comprehensive reform. It is also evident that a broad consensus has developed among many of the physician and patient organizations on the broad outlines for such reform. The committee, which for some time has viewed such a fee schedule as its goal for payment reform, is now in a position to formulate the fee schedule and begin making more substantial progress in achieving that goal.

Strategy for reform. The strategy embodied in this section is to enact legislation this year, and begin implementation on April 1, 1990, for a 4 year plan of transition to a fully implemented fee schedule based on a resource-based relative value scale. The committee recognizes that we do not yet have all the answers needed for full implementation of an RBRVS fee schedule. Research is continuing on some of the important issues involved. Many procedures have yet to be surveyed, or are scheduled for re-survey, by the research project at the Harvard School of Public Health that is developing the relative value scale. Further studies and analyses will be conducted by the PPRC and the Department of Human Services. However, it is also the committee's view that we know enough to begin implementation on a phased-in basis. It is further the committee's view that we may not be able to identify all the issues and questions until the Secretary has actually begun to construct and implement the fee schedule.

We have learned a great deal about physician payment reform over the last 5 years and we know enough to set our ultimate objective and begin promptly to make substantial progress toward that objective. Since the first adoption of the fee freeze in 1984, we have mandated studies by the Secretary of Health and Human Services and we have created the Physician Payment Review Commission to evaluate alternative reforms and make recommendations to us. The Commission's recent third annual report provided us with a comprehensive and thoughtful set of specific recommendations, which are supported by a number of important physician and beneficiary groups.

Moreover, by implementing this provision on a 4 year phase-in, we have time to resolve those questions for which we do not have complete answers, and we will be better able to answer them based on real experience gained during the initial implementation stages. Undoubtedly, there will be new issues raised during implementation that we have not anticipated. Over the next 2 years, there will also be generated better data and information, and more thorough analyses, to answer both currently identified issues and new issues that arise.

The committee also recognizes, however, that implementation in this manner, and this learning process, should be undertaken in a way that minimizes any risk of a serious mistake or an over-correction of current deficiencies. Physicians, patients and policy-makers all need time to understand the changes, to adjust to them, and to monitor the consequences carefully. It will be particularly important to monitor closely the potential effects on access to care, on quality of care, and on patient out-of-pocket expenditures. Changes in the mix, intensity and volume of services will also be of great interest to the committee.

The committee's strategy for minimizing errors and disruptions during the phase-in is to begin by making marginal adjustments in the current reasonable charge methodology, by reference to the ultimate objective of a RBRVS fee schedule. During the first 2 years of the phase-in, the Secretary of HHS would construct a "reference fee schedule", based on a resource-based relative value scale. However, payments would not be made directly under this reference fee schedule. Rather, it would be used to increase or decrease the current prevailing charge screens for individual services, depending on the relationship of those prevailing charges to the reference fee schedule. The reference fee schedule would, however, be as complete as possible in incorporating all of the essential components and elements of the final fee schedule, based on the best available information.

In order to expedite the initial implementation, and avoid delays that might ensue if the Department had to resolve various policy issues, the committee bill is quite specific in giving directions to the Secretary on how to begin the initial implementation in 1990. In particular, there are the three appendices attached to this committee report. The first one would identify the specific procedures for which the payment amounts would be adjusted and specify the relative value for each. The second appendix contains instructions on calculating the conversion factor so as to preserve budget neutrality and the third sets forth specific geographic adjustments in the conversion factor.

The committee bill leaves more discretion to the Secretary to resolve important issues after 1990. The committee expects the Secretary to consult widely with interested parties in resolving such issues and the normal requirements for notice and comment rulemaking would apply. The committee also anticipates that it will review this legislation periodically and is likely to amend and refine it to reflect subsequent research and the experience gained during the interim.

Application in 1990 to selected procedures. The committee recognizes that it is not practical at the time of the initial implementa

tion to make adjustments in the payment amounts for all of the more than 7,000 procedure codes used in the Medicare program. Nor is it necessary to do so in order to make substantial progress on reform. Only those 389 procedures listed in Appendix A would be subject to adjustments under this reform in 1990. However, the procedures on this list represent many of the most significant services furnished under Medicare.

The list in Appendix A is based on the recommendations of the PPRC. It was developed with technical consultation with the Health Care Financing Administration and was subject to review by physician organizations. The list is comprised of services that were surveyed by the Harvard research team and services that are closely related to those surveyed. The specific relative values for the procedures on the list have been established by the PPRC based on the findings of the Harvard research team. Services were removed from the list whenever there was any indication of a problem or serious concern with the relative value assigned to it-such as changes in coding, insufficient or faulty data, or concern that a service was too dissimilar from those actually surveyed to warrant an extrapolation. As a result, the list represents services for which the committee has a high degree of confidence that the relative value is sufficiently accurate to begin the phase-in.

The list in Appendix A does not include services for several important medical specialties, which have not yet been surveyed by the Harvard research team or for which relative values could not confidently be determined. In particular, the schedule does not include radiology and anesthesia services. The Omnibus Budget Reconciliation Act of 1987 contained specific provisions, which originated in this committee, calling for fee schedules for those two specialties. The committee intends for the fee schedules developed under the 1987 legislation to be incorporated into the payment reform contained in this bill, beginning in 1991. This means that the work done by these specialties in developing the relative values among the services they perform should be retained to the extent consistent with overall reform. Adjustments will be made in the payment amounts for these services, as appropriate, to make sure that the payment levels for these services are consistent with their relative value compared to other services covered by the RBRVS fee schedule. The Secretary would be instructed by the bill to do this.

In the case of anesthesia services, the most practical way of doing this would appear to be to retain the relative value guide and make adjustments in the conversion factor. Since the preliminary analysis for anesthesia services indicates that they are overvalued relative to many other services, this would mean a reduction in the conversion factor.

For radiology services, it would be possible to do a similar adjustment in the conversion factors. However, this would mean that radiology services, for which the methodology of the fee schedule is otherwise identical to all services other than anesthesia, would have relative values that are not comparable and unique conversion factors. The committee concluded that the radiology fee schedule should conform as much a possible to the remainder of the fee schedule. Consequently, the bill would require the Secretary to use

the same conversion factor (and geographical adjustments) as he would for other services, and to adjust the recently constructed relative values for radiology services, as a class, to bring them into line with the RBRVS generally. In doing so, the Secretary would retain the relationships among the values assigned to radiology services under the current radiology fee schedule.

This method of treating anesthesiology and radiology places great reliance on the judgment of the physicians affected by the fee schedule for determining the relative weights to be accorded within the range of services they furnish. It also preserves the effort made in constructing the current fee schedules. The committee encourages the Secretary, the PPRC and other physician groups to use similar approaches in establishing relative values for the range of services not included in the survey conducted by the Harvard researchers. Such reliance on physician expertise will enhance the validity and acceptability of the fee schedule and is consistent with the development of the RBRVS to date.

Specialty differentials. As a general matter, the committee bill does not preserve the current specialty differentials that are frequently used by Medicare carriers under current law to pay differing amounts for apparently comparable services depending on the specialty designation of the physician furnishing the service. There has been considerable criticism about such specialty differentials, particularly since they are not applied in a consistent manner around the country, since physicians are permitted to designate their own specialty in the absence of definitive standards, and since it is difficult to prove whether the services are dissimilar. The PPRC recommended that specialty differentials be eliminated for like services and the bill does so.

It is the committee's view that those instances in which differential payments are warranted, based on the training and expertise of a specialist, can normally be accommodated with proper use of the procedure codes. Changes in the present codes may be necessary to implement this policy. There may be also particular instances, however, in which payment differentials are warranted and the committee expects the PPRC and the Secretary to review this issue carefully.

The general policy against specialty differentials does not mean, however, that specialty identifiers should be eliminated from claims forms currently submitted to Medicare. It will remain important to have claims data by specialty for several reasons. For one, treatment of malpractice expenses under the fee schedule may well be done on a specialty basis, when changes are made in 1992 as discussed below. Second, the volume and mix of services will be closely monitored along several dimensions, one of which should be by specialty. Third, specialty identifiers are important during utilization review to make sure payments for concurrent care are appropriately made. Fourth, the claims data generated by Medicare can be an important resource for valuable research, some of which is likely to require specialty designations.

Subsequent application and modification of relative values. After 1990, the reform would be applied to all of the procedures for which payment is made under Medicare. The Secretary would be responsible for establishing the relative values, based on the Har

vard research project and the recommendations of the PPRC. The Secretary would also be responsible for keeping the relative values as current as possible and modifying them, as appropriate, to reflect changes in the practice of medicine, in the delivery of services, or in technological innovation. Two avenues are provided for making such changes. The Secretary, at any time during the year, could establish a new relative value for a service that had not had a value assigned. In addition, on an annual basis, the Secretary would be expected to review some or all of the existing relative values and revise them as appropriate. However, the bill is explicit in stating that such changes can only be done to reflect corrections in relative values. Adjustments could not be made solely for the purpose of achieving reductions in expenditures.

It is important to note that, under this method, there will be a single, uniform relative value scale applicable throughout the country. This will facilitate understanding of the relative values and the derivation of the fee schedule, as well as making subsequent updates and revisions easier. However, it should also be noted that this approach means that the geographical adjustments described below, to take account of regional differences in the cost of furnishing services, will be made on the conversion factors. Consequently, while there will be a national average conversion factor, there will not be a single, standard conversion factor for each fee schedule area. Rather, each service will have an individual local conversion factor. (An "average" local conversion factor can be computed for purposes of making general comparisons, but could not be used to construct the actual payment amount for any individual service under the local fee schedule.) While the complexity of this methodology may cause some confusion, the committee believes it is the most accurate and logical method of constructing the fee schedule and, overall, the most straightforward way of understanding it.

Adjustments in prevailing charges. The reference fee schedule would be used to adjust the prevailing charges otherwise calculated for 1990 for the services on the list in Appendix A. The amount of the adjustment would be one-fifth of the difference between the prevailing charge and the reference fee schedulee. This adjustment is substantial enough to result in significant progress toward the RBRVS fee schedule, without being so large as to cause a serious disruption or to run the risk of an erroneously large adjustment that will require subsequent correction. For many services, the actual amount of the adjustment will be less than 2 or 3 percent of the current payment amount.

Most services would be individually adjusted in this manner. However, this section sets out a special rule for office visits, hospital visits, consultations, and other services identified in the bill under the caption "evaluation and management services". These services would all receive an increase in the payment level. However, under the committee bill, they would all receive the same percentage increase, rather than individually calculated increases. This uniform percentage increase would be based on the weighted average of the increase each would have received under the statutory formula. The reason for this special treatment is to take into account concerns expressed by the PPRC and others about the accuracy of the relative values for some of these services, due to am