Nancy-Ann M. DeParle
Administrator
Health Care Financing Administration
Attn: HCFA-3432-NOI
Room 443-G
Hubert H. Humphrey Building
200 Independence Avenue, SW
Washington, DC 20201

Subject: RIN 0938-AJ31 Medicare Program; Criteria for Making Coverage Decisions; Notice of Intent to Publish a Proposed Rule

Dear Ms. DeParle:

On behalf of the American Medical Association (AMA), I am writing to comment on the Health Care Financing Administration's (HCFA) plans to establish criteria for Medicare coverage policy decisions. Over the past two years, HCFA has made considerable progress in addressing the AMA's concerns about the Medicare coverage policy decision process through its establishment of the Medicare Coverage Advisory Committee and its Web-based tracking system for national coverage policy proposals and decisions.

The two major issues that have been as yet unresolved and that this Notice of Intent begins to address are the need to establish criteria for Medicare coverage and the need to develop standards for use in making local Medicare coverage decisions. The Notice of Intent is a welcome indication that HCFA is beginning to move forward in establishing regulations governing the criteria to be used as the basis for Medicare coverage policy decisions. Nonetheless, the AMA has serious concerns about the proposed coverage criteria as HCFA has outlined them in this Notice.

We also welcome HCFA's stated intention, once acceptable coverage criteria are established through the rulemaking process, to apply standard criteria to local as well as national Medicare coverage policies. The AMA has long advocated that HCFA develop standard procedures and criteria for use by the Medicare carriers in making local Medicare coverage policies. Often these policies have been based solely on statistical analyses of "normal" utilization patterns instead of being grounded in good clinical information. Sound criteria for local policies are sorely needed.

On June 2, 2000, staff from the AMA and national medical specialty societies were pleased to participate in a meeting with Jeffrey Kang, MD, MPH, in which he provided a briefing on HCFA's plans for the proposed rule on criteria. The questions and comments raised at this meeting are the basis for our comments on the Notice of Intent.

A general theme throughout much of the discussion at the meeting was a concern about the desirability and feasibility of implementing the decision process that HCFA has outlined. In a perfect world, both HCFA coverage policy decisions and the decisions made by physicians and patients about treatment alternatives could be informed by high quality clinical research. Unfortunately, in the far-from-perfect real world of medical practice, we may not have adequate evidence available from clinical trials to support questions about the specific medical benefits of a service, let alone to inform decisions about comparative clinical utility of various treatment alternatives within the same clinical modality. Furthermore, even if there is evidence from clinical trials, the tools may not yet have been developed that would allow us to grade the various treatment options on the proposed "quality of life" factors, such as compliance and convenience.

The 4-step process for making Medicare coverage decisions described in the Notice inappropriately places HCFA in the position of making medical decisions about patient care. Decisions about which particular course of treatment will be selected when multiple treatments are available are a fundamental element of medical practice: such choices should be made by physicians and their patients, not by a government agency.

The AMA is also concerned that, if HCFA's coverage criteria are based on unrealistic expectations about the amount and quality of evidence from research, Medicare may fail to cover numerous diagnostic and therapeutic options that would be beneficial for patients. The AMA has not wavered in its support for evidence-based decision making. Nonetheless, the number of steps outlined in the Notice of Intent where evidence would be required is unrealistic. For example, even if evidence "demonstrates that the item or service is medically beneficial for a defined population," it is likely that evidence will not be available on the already-covered item or service that is serving as a point of comparison for the new one. Even if some evidence is available, there may not be sufficient evidence on the same patient populations or under the same circumstances to allow comparisons between the two services. If evidence is not available to support all of the decisions that must be made in Steps 1 - 4, what then provides the basis for a Medicare coverage decision? The lack of adequate evidence to address every one of these questions should not serve as a reason to issue a noncoverage decision or postpone a coverage decision.

The national medical specialty societies and other leading medical organizations have focused considerable resources and expertise on the development of clinical recommendations. In a number of instances, HCFA has looked to these clinical recommendations for guidance in shaping its own performance measurement and quality improvement activities. Medicare coverage criteria and evidentiary standards should rely upon the same body of knowledge that physicians utilize for clinical decision making. If HCFA establishes criteria that go well beyond this body of knowledge, it may frequently be the case that particular services and procedures will be recommended in widely accepted practice guidelines but that Medicare will not cover them. Conflicts between clinical recommendations and Medicare coverage policies will seriously impair efforts to improve health care quality. To avoid such conflicts, the AMA recommends that the status of particular services and items as the "standard of care" or "standard of practice" for a condition, or their inclusion in accepted clinical recommendations, be explicitly considered in Medicare coverage decisions.

Questions were also raised in our meeting about the feasibility of implementing steps in the process that do not rely on evidence. For example, how will HCFA define whether various treatment options are in the "same clinical modality?" What standards will HCFA apply when questions arise about services that are already covered? Where will HCFA obtain information on costs and how can it account for the cost decreases that often accompany diffusion of medical innovations?

Despite our concerns about how "quality of life" criteria will be applied to coverage decisions, we strongly support HCFA's inclusion of these criteria. If a particular treatment is more convenient for patients, produces fewer or more tolerable side effects, or enhances the likelihood of compliance, without increasing the risk of patient harm, then that treatment should be able to be offered.

The AMA believes that patient access should also be considered a criterion for coverage. For example, HCFA's interim final rule of June 24, 1988 on coverage of bone density measurement procedures states, "Given the differential access and convenience of various bone mass measurement techniques available to Medicare beneficiaries, the attending physician must be given the option to order the most appropriate bone mass measurement for a beneficiary in a particular set of circumstances." Such considerations of access and convenience are highly appropriate and should continue to be considered by HCFA in making Medicare coverage policy decisions.

The AMA is pleased that HCFA is making progress in establishing criteria for Medicare coverage and we look forward to reviewing the forthcoming proposed rule.

Sincerely,

E. Ratcliffe Anderson, Jr., MD