In the past several years, the United States has witnessed a perceived drug safety crisis. The withdrawal of Vioxx, the potentially unsafe decrease in drug approval times due to the controversial Prescription Drug User Fee Acts (PDUFA), and the undue industry influence that these Acts may have entailed have fueled these sentiments. It is in this context that the Institute of Medicine (IOM), at the request of the Food and Drug Administration's (FDA) Center for Drug Evaluation and Research (CDER), released its final recommendations for reforming the US drug safety system this past September (Institute of Medicine, 2006). The goal of the IOM panel was laudable, since despite the agency's strict adherence to evidence-based evaluation of the products it oversees, less evidence and evaluation exist on its own safety and efficacy. Put differently, no product application would pass the FDA approval process with the quality and type of evidence that currently exist for evaluating the FDA policies themselves.
Though the scope of the drug safety system is wide and diverse, the IOM report naturally focused its attention on the main regulatory body in the US responsible for both pre-and post-marketing assessments of the safety and efficacy of pharmaceutical products, the FDA. Its highly detailed recommendations can generally be divided into improvements in: (1) the organizational structure of the FDA; (2) the methodology and infrastructure to assess drug safety; (3) communication across stakeholders in the drug safety system, e.g. the FDA, patients, and physicians; (4) the regulatory power of the FDA, in particular, during the period of post-approval marketing; and (5) the resources required to fund these recommendations.
Despite the strengths of several of the Institute's recommendations, the main weakness of the report is that it is premised, with little evidence, on a perceived drug safety crisis in the US. Much of this perception centers on heavily publicized drug withdrawals, such as the withdrawal of Vioxx, as well as the PDUFA of 1992, 1997, and 2002. These Acts were designed to lower drug approval times by financing the evaluation process with user fees levied on entities applying for FDA approval. As the IOM report notes, PDUFA has been controversial for several reasons, among the most relevant being the 'too-close' relationship between the industry and the FDA, the potentially unsafe rush towards drug approval, and the pressure imposed on FDA staff due to this rush. Is there an empirical basis for these concerns and, more importantly, is there evidence that too many unsafe drugs enter the market?
With regard to PDUFA, while the IOM report nicely documents the decrease in approval times due to these Acts, there is no discussion of whether these reductions were 'rushed' in the sense that the safety and well-being of consumers were compromised. More generally, the report makes no effort to assess whether the decrease in safety due to lower approval times and the additional burden imposed on FDA staff were outweighed by the benefit to those able to consume the drug earlier, an argument suggested by existing evidence. For example, Philipson et al. (2005) found that the social benefit generated by reductions in drug approval times in the PDUFA era substantially outweighed any decreases in safety associated with more rapid approval. Focusing on 662 new molecular entities approved from 1979 to 2002 -this includes the period of Vioxx -the authors estimate that the societal gains from reduced approval times amount to $16-$32 billion dollars, an amount equivalent to 180-310 thousand life-years. In the extreme case that all drug withdrawals that took place during the PDUFA era were due to lapses from more rapid approval, the number of life-years lost would be 58 thousand. For the case of PDUFA, the trade-off between speed and safety appears to have been quite favorable in the sense of raising