Health Care Reform Center

Randall S. Stafford, M.D., Ph.D., Todd H. Wagner, Ph.D., and Philip W. Lavori, Ph.D.

New technologies, including prescription drugs and medical devices, are a major driver of increases in U.S. health care expenditures, which have grown by an estimated 71% since 2000.¹ The U.S. market for drugs and devices is regulated by the Food and Drug Administration (FDA), which scrutinizes clinical trial data for evidence of safety and efficacy. Although the FDA has been criticized for missteps and inefficiencies in its approval process, these are not the causes of increasing health care expenditures. More relevant is FDA oversight of the labeling and promotion of medical products.

Despite the potential usefulness of labeling information for controlling the unnecessary growth of expenditures, the FDA does not require the inclusion of statements regarding a product’s comparative effectiveness. As a result, drug labels may create confusion, as manufacturers strive to insulate their products from price competition through differentiation that is unrelated to health outcomes. If the FDA label were required to indicate what is and is not known about a product’s superiority to other treatments, then clinicians, patients, and payers would be less willing to pay more for a new treatment without proof that it improved health outcomes. In addition, manufacturers would have an incentive to conduct much-needed active-comparator superiority trials.

The FDA requires developers of new treatments to demonstrate that they are safe and effective in order to receive approval for market entry, but the agency demands proof of superiority to existing products only when it is patently unethical to withhold active treatment from study patients, as in the cases of therapies for AIDS and cancer.² Many new drugs are approved on the basis of demonstrated superiority to placebo. Even less is required for many new medical devices.

There are numerous reasons for this approach. Placebo-controlled trials require smaller sample sizes than active-comparator trials, are less expensive to conduct (and therefore reduce the costs of market entry), and present less risk of producing unanticipated unfavorable findings. Unlike active-comparator trials, placebo-controlled trials also prevent the possibility that two indistinguishable active comparators will be no better than placebo (see table). Yet placebo-controlled trials that are not supplemented by active-comparator trials leave clinicians, patients, and payers in the dark, providing no guidance on a new product’s advantages or disadvantages relative to existing products.

Physicians need such information to guide them in defining a product’s appropriate place in practice, and payers need it to determine whether a treatment should be elevated to first-line status, to assign it to a copayment tier, and to establish requirements for prior authorization. In the absence of evidence, other factors may dominate these coverage decisions, including fears of negative publicity. If new therapies cost no more than existing treatments, payers might simply allow physicians and patients to sort out preferred therapies over time with the help of postmarketing observational data. But new therapies are almost always more costly than previously approved treatments, particularly so when existing drugs are available in generic form.

The evidence gap, which grows with each placebo-controlled trial, enables developers to market their products as the next generation of advances, differentiating them in terms of features that bear little relation to disease outcomes. The current regulatory climate thus favors the creation of products that differ minimally from existing therapies. Although purchasers can negotiate lower prices for “me-too” drugs, it is not always clear which treatments are truly substitutes for previously approved products, and savings are not inevitable. Marketing aimed at consumers and physicians creates brand-name awareness and facilitates product selection that is not based on outcomes. As the more recently approved selective serotonin-reuptake inhibitors have shown, me-too drugs can be highly profitable even when older drugs may be better choices.³

Developers face few incentives to conduct active-comparator superiority trials and understand that they benefit from the unacknowledged deficiency of evidence. The development or marketing of me-too drugs and devices may provide a greater return on investment than research aimed at true clinical innovation. With thousands of clinical trials conducted each year and a demand for the newest generation of treatments, costs will continue to rise until prescribers and purchasers have access to more discriminating information.

We believe that it is just as important for clinicians and patients to be made aware of what is and is not known about comparative efficacy as it is for them to be given the currently required information on the risks and benefits of new products. In the absence of comparative data, drug and device labels should include a statement indicating that there is no evidence of the product’s superiority to other products. For example, the label or advertising text for a new calcium-channel blocker might state, “Although this drug has been shown to lower blood pressure more effectively than placebo, it has not been shown to be more effective than other members of the same drug class.” If payers could readily identify clinically equivalent products, they could treat them as commodities and negotiate on price. Developers would have incentives to design trials that differentiate their products with respect to clinically important attributes. This change could be made incrementally, with input from stakeholders, and it would not require capital investments or new administrative systems.

There are several potential objections to this proposal. The FDA has focused primarily on informing the medical marketplace about safety issues, and some might argue that comparative-effectiveness labeling is outside the agency’s expertise and purview. However, the FDA has considerable expertise in food labeling, and its approved labels are widely accepted by manufacturers and figure prominently in consumers’ purchasing decisions. Developing more informative labels for drugs and medical devices is also consistent with the FDA’s function as a public health agency.⁴ Physicians and consumers may pay limited attention to labeling, but requiring manufacturers to include disclaimers of superiority in their marketing materials and drug compendiums would highlight this key information and allow it to reach more people. The process of defining drug classes may be complex, but the apparently successful definitions created for Medicare Part D could provide an initial schema.⁵ Further complications may occur when a drug has multiple indications or is proved superior to only some of the drugs in a class. Such problems seem solvable, however, and only increase the value of indicating in the labeling exactly what remains to be determined.

Additional strategies could help to generate knowledge about the relative effectiveness of two drugs or devices. Large private and governmental payers could conduct clinical trials to assess the efficacy and cost-effectiveness of competing treatments, although developers should arguably bear this responsibility. In the absence of optimal evidence from clinical trials, statistical tools such as network meta-analysis and the prospective collection of observational data might indicate likely superiority. A national center for comparative-effectiveness research could not only generate assessments of new treatments but also actively facilitate changes in practice. Integrating pragmatic clinical trials into day-to-day medical practice might also provide substantial benefits. These strategies could augment the impact of revised FDA labeling.

The process of developing drugs and medical devices is lengthy, expensive, and financially risky. A few products will be breakthroughs that improve health outcomes; most will offer little, if any, advantage over existing treatments. At the time of FDA approval, it is rarely clear whether a new drug or device falls into the first category or the second. With the expanded labeling requirement we propose, stakeholders would have explicit information about proven comparative benefits or the lack thereof — and drug and device manufacturers would have an incentive to design more informative clinical trials.

Dr. Stafford reports receiving grant support from Procter & Gamble and Toyo Shinyaku; Dr. Wagner, grant support from American Medical Systems; and Dr. Lavori, consulting fees from Neuronetics, Corcept, Fibrogen, Depomed, DLA Piper, and ARCA Biopharma. No other potential conflict of interest relevant to this article was reported.

The views expressed in this article are those of the authors and do not necessarily reflect those of the Department of Veterans Affairs.

Source Information

From the Stanford Prevention Research Center, Stanford University, Palo Alto (R.S.S.); the Veterans Affairs Palo Alto Health Care System, Palo Alto (T.H.W.); and the Department of Health Research and Policy, Stanford University School of Medicine, Stanford (T.H.W., P.W.L.) — all in California.

This article (10.1056/NEJMp0906490) was published on August 12, 2009, at NEJM.org.

References

1. Sisko A, Truffer C, Smith S, et al. Health spending projections through 2018: recession effects add uncertainty to the outlook. Health Aff (Millwood) 2009;28:w346-w357.
2. Food and Drug Administration. Guidance for industry: E 10 choice of control group and related issues in clinical trials. Rockville, MD: Department of Health and Human Services, 2001.
3. Cipriani A, Furukawa TA, Salanti G, et al. Comparative efficacy and acceptability of 12 new-generation antidepressants: a multiple-treatments meta-analysis. Lancet 2009;373:746-758.
4. Hamburg MA, Sharfstein JM. The FDA as a public health agency. N Engl J Med 2009;360:2493-2495.
5. USP Model Guidelines Expert Committee. Narrative review: the U.S. Pharmacopeia and model guidelines for Medicare Part D formularies. Ann Intern Med 2006;145:448-453.