The Task Force Report from the Foundation for Managed Care Pharmacy (FMCP) in this issue of the journal [1] offers an overview of the welcome update  of  the  Academy  of  Managed  Care  Phar-macy (AMCP) Format for Formulary Submissions (Version 2.0), which was published in October 2002 (available from Reflecting on the experience with the first AMCP Format, the report echoes many issues that will be familiar to those developing guidance for clinical and economic evidence reimbursement submissions around the world. The challenge is how to get relevant and rigorous evidence to managed care formulary decision makers while navigating the regulatory and commercial realities of the pharmaceutical marketplace.

It is widely understood that the typical placebo-controlled trial evidence on safety and efficacy that is relevant for product licensing decisions is seldom-adequate evidence for making decisions on reimbursement of new medicines. Consider our own work on the cost-effectiveness of a new antiviral drug (oseltamivir) for treatment of influenza [2]. The effectiveness and efficiency of this treatment hinge on two key variables concerning the specificity of diagnosis and treatment: 1) what proportion of patients presenting and treated with flu-like illness will have the influenza virus; and 2) what proportion of patients will be treated beyond the efficacy window of 48 hours from symptom onset when the virus is still replicating? Close to 70% of patients in the registration trials were virus positive [3,4] and an inclusion criterion was that symptom duration was less than 48 hours. Hence a mixture of usage—that according to the “label” and that which is “off-label”—will determine the real-world effectiveness and cost-effectiveness. It is this “leakage” of a medicine's usage into nonindicated populations, where treatment effects may be diluted and cost-effectiveness eroded, which is major cause of concern [5].

The price for attempting to provide relevant evidence of real-world cost-effectiveness, complete with anticipated leakage, is that the AMCP Format needs to tread carefully around the regulatory guardians of the product label—the Food and Drug Administration (FDA). In a bizarre piece of regulatory theater, a manufacturer cannot simply provide an AMCP Format dossier to a managed-care organization in a submission for formulary listing; the initiation of such an exchange must be an “unsolicited request” from the MCO for such a dossier. What passes the solicitation test—“I may have some evidence for you but the FDA will not let me tell you what it is unless you ask me for it in a special way?” Rather than regulating the rules of solicitation it would be better to focus more energy on making sure that the content of the reimbursement dossiers is not misleading.

Cost-effectiveness modeling is an important component of drug reimbursement submissions around the  world  and  with  the  AMCP  Format. Modeling is an unavoidable part of evidence synthesis as it relates to economic questions [6]. The danger is that a model can be a “black box” if the structure, data, and assumptions are not clear and transparent. In this regard the work of the ISPOR Task Force on Good Research Practices–Modeling Studies is a useful benchmark for starting to think about quality and standards [7]. An important aspect of the AMCP format is that the manufacturer is required to submit a full electronic working copy of the model. This makes it possible for a third party to interrogate a model and further understand structure and assumptions. It also capitalizes on the modular and flexible nature of models such that local data (e.g., demographics, prices) can be substituted into the model to tailor it to local circumstances.

A potential commercial boundary on transparency arises from concerns about confidentiality of data and other evidence. The problem of how to deal with commercial-in-confidence data in cost-effectiveness submissions is a thorny problem that is not unique to US managed care but can also be found in the assessments performed by the National Institute for Clinical Excellence (NICE) in the United Kingdom [8]. Drummond has made the case for full disclosure of all evidence for public scrutiny in the context of NICE and makes convincing arguments in favor of this view, particularly with respect to the principle of decision transparency, such that the evidence used for (for example) restricting ac-cess to a medicine can be made known to patients and providers. In contrast, the AMCP view appears to be to encourage MCOs to ensure that confidentiality of dossiers will be safeguarded. There is not right or wrong here, just a difference of emphasis on the needs of transparency versus the need to protect property rights as a condition of the successful exchange of information.

The  report  also  contains  a  very  important  real-ity check concerning human resources: faced with increasingly sophisticated analyses and models, MCOs as the recipients of this information may lack the skilled labor to review and interpret the submissions. Again this problem is echoed around the world, particularly in countries such as Australia, Canada, and the United Kingdom where cost-effectiveness evidence supporting formulary listing has been required for some years. Industry can offer higher wages and attract a large share of the skilled pool of labor in outcomes research and cost-effectiveness. Part of the solution, as FMCP has been quick to understand, is offering training workshops on how to use the AMCP format. But it may also make good business sense for the pharmaceutical industry to help subsidize the continuing education of MCO dossier reviewers in contemporary methods of modeling and cost-effectiveness. Helping to create the skilled receptors in managed care for the evidence and analyses submitted is almost as important as the studies themselves.


  1. Top of page
  2. References