Prescription drugs represent a large and rapidly growing share of health care expenditures – 15% in the US and 7.5% in Europe. While costly, research shows that drugs represent an enormous value to society; they both improve health1 and reduce medical costs.2
The large impact of prescription drugs on society – both on the benefits and costs side – raises many issues for policymakers. The regulation of drug safety, prices and advertising, among others, are widely debated in political circles and the popular media. Yet, despite the oftentimes heated debate over these issues, there is surprisingly little empirical evidence demonstrating how various policies will affect drug prices, patient welfare, and pharmaceutical firms’ incentives to invest in R&D. This lack of evidence in support of prescription drug policies is particularly surprising when one considers the tremendous amount of empirical evidence and research devoted to the science of these drugs.
The regulation of prescription drug safety and efficacy is one of the key policy issues. In the US, such oversight is the job of the Food and Drug Administration (FDA), which regulates drugs and medical devices, as well as some other goods. In order to market a drug, the manufacturer must demonstrate its safety and efficacy through a series of clinical trials typically lasting 6-11 years. After these trials are completed, the manufacturer then sends the results for review by the FDA, a process which can take up to two years, after which the FDA then issues its approval for the drug, as well as a drug label, whose purpose is to communicate relevant information to health care providers, such as the diseases the drug is supposed to treat, side effects, as well as any other precautions.
A similar process exists in Europe, with the European Medicines Agency (EMEA) acting as an analogue to the FDA.
Recent events have heightened interest in the regulation of drug safety and efficacy in US. The first is an outbreak of lawsuits against Merck & Co, Inc. over its drug Vioxx and its withdrawal from the market in 2004. The second is the release of a widely-read report from the US Institute of Medicine criticizing the FDA, for a perceived “drug safety crisis”, and outlining several efforts to reform the agency.3
These events suggest two critical questions in evaluating policies for regulating drug safety and efficacy.
- What is optimal point in the fundamental speed-safety tradeoff?
A longer, lengthier review process (including larger clinical trials) will reduce the probability that unsafe drugs enter the market, but at the cost of delaying patient access to potentially useful therapies. Thus, any regulatory agency, such as the FDA or the EMEA, must find the optimal speed/safety combination that efficiently balances the gains of more rapid access to drugs against the increased probability of unsafe drugs being introduced into the market. Until recently, there has been surprisingly little evidence on the performance of either agency in this regard.
- Why do drug regulatory agencies exist to begin with?
While this question may seem absurd at first, it is important to remember that other, market-based mechanisms exist to regulate safety and efficacy. First, patients and physicians are less likely to purchase drugs with poor or scant evidence of safety/efficacy, thereby leading to such drugs being eliminated from the market. In addition, legal liability gives firms incentives to provide safer drugs. Thus, it is important to understand how the pros and cons of direct government regulation compare to other, market-based systems. Moreover, given that no country relies solely on either method, it is important to understand how the two systems interact, and what the consequences of this interaction might be.
In a recent paper, we examined the first question by considering the effects of the Prescription Drug User Fee Act in the US.4 These series of acts mandated a faster FDA review process in exchange for user fees that were levied on pharmaceutical firms, and provide insight into the speed-safety balance at the FDA. If, on net, these acts improved patient welfare, this would suggest that prior to the acts, the FDA erred too heavily on the “safety” side of the speed-safety balance, and that there may be potential gains to moving towards the “speed” side of the balance. Our research suggests that this is the case. First, we found that drug approval times fell by roughly 2 percent per year in the periods before 1992. After the passage of the Prescription Drug User Fee Act of 1992, approval times fell by 9-10 percent annually, and then about 5 percent annually after the legislation was first reauthorized from 1997-2002. Thus, the Act did accomplish its stated goal of reducing review times at the FDA.
To examine the effect of these faster review times on social welfare, we began by looking at producer surplus (profits). Using sales data for a set of 663 drugs on the U.S. market between February 1998 and December 2002 and given our estimates of the effect of the Act on drug approval times, we valued greater speed by asking how much the present value of welfare would increase by allowing the observed stream of surplus to happen sooner.
Overall we estimate that the reform of 1992 increased producer surplus by roughly $11 billion (0.96 percent of sales).
Calibrating the effects of the Prescription Drug User Fee Act of 1992 on consumer and social surplus was more complicated for two reasons. First, since consumer surplus must be inferred from quantity and price data alone, some assumptions on the nature of demand had to be made. Second, consumer surplus is lower before a patent expires and higher after it expires, since generic competition occurs, thereby reducing prices. Our major finding was that the Act increased the present value of social surplus by between $18-31 billion, which amounts to about 1.6 to 2.7 percent of overall sales.
Against these estimated increases in social surplus must be weighed against the potential losses due to the possibility of less safety in the drugs being released. Using the Adverse Event Reporting System to calculate the number of lives lost from drugs that were approved under the Act but subsequently withdrawn, our analysis found that about 55,600 life-years were lost from drugs that were approved under the years of the Prescription Drug User Fee Act but were subsequently withdrawn. A reasonable range of estimates for the value of a life-year would be from $100,000 to $300,000. Thus, the losses due to reduced safety would range from $5.6 billion to $16.6 billion. Since the present value of social surplus was $18 billion to $31 billion, the gains from increased speed likely outweigh the costs.
Why should the government do it?
We examined the second question in Philipson and Sun (2007). Overall, we find that direct government regulation is useful in pharmaceutical markets because of the following reasons:
- Unsafe drugs can lead to large monetary losses, so firms can evade judgments by declaring bankruptcy.
- Because patient who use drugs often present with other co-morbidities which may lead to death or serious injury, it is often difficult to establish that an unsafe drug was the direct cause of injury.
- Since regulation is a fixed cost for each drug, while litigation can involve costs proportional to the number of those potentially harmed, regulation may be more favorable for larger economies and populations.
In addition, our analysis suggests that there can be inefficiencies that result when – as is currently the case – drugs are regulated by government regulation and product liability.
Inefficiency when control is by government regulation and product liability
These inefficiencies can occur when regulatory bodies mandate safety levels that are higher than what the firm would choose to provide under product liability alone. For example, in the US, firms seldom perform more clinical testing that what is required by the FDA, suggesting that it is the agency’s directives, as opposed to the fear of litigation, that is driving the current level of clinical testing. When this occurs, product liability can be wasteful, because while it has no additional effect on the level of safety that firms provide, it does impose legal costs on firms, which are then passed on to patients in the form of higher prices. Given these findings, we estimate that a product liability exemption for pharmaceutical firms5 could increase patient welfare by $47.8 billion (4 percent of sales) to $754.7 billion (66 percent of sales), depending on the share of firms’ marginal costs that are devoted to legal costs.
Two critical issues in drug safety regulation are the speed-safety tradeoff, and the overlap of regulation and product liability. Our results suggest that, at least in the US, regulatory agencies have historically erred on the “safety” side of the speed-safety balance; society could gains by streamlining the approval process. In addition, our results suggest potential gains from removing inefficiencies that are driven by joint regulation of drug safety through the FDA and the product liability system.
Our results are clearly a starting point, and certainly more analysis should be done, particularly for European pharmaceuticals markets. More generally, we believe that rigorous, economic based analysis can greatly inform an area whose thinking is driven primarily by physicians and policymakers. Nations routinely spend billions on research into the science of pharmaceutical drugs. Maybe it is time to spend a little more on the economics of their regulation.
Berger, J.B., Slezak, J., Stine, N., McStay, P., O’Leary, B., and Addiego Jr., J. “Economic Impact of a Diabetes Disease Management Program in a Self-Insured Health Plan: Early Results.” 2001. Disease Management 4(2):65-73.
Cowper, P.A., E.R., Whellan, D.J., LaPointe, N.M., and Califf, R.M. “Economic Effects of Beta-Blocker Therapy in Patients with Heart Failure.” 2004. Am J Med 116:104-111.
Lichtenberg, F.R. “Benefits and Costs of Newer Drugs: An Update.” 2002. National Bureau of Economic Research Working Paper no. w8996.
Lichtenberg, F.R. “The Expanding Pharmaceutical Arsenal in the War on Cancer.” 2004. National Bureau of Economic Research Working Paper no. w10328.
Matchar, D.B., and Samsa, G.P. Secondary and teriary Precention of Stroke, Patient Outcomes Research Team (PORT) Final Report-Phase 1. 2000. AHRQ Pub. No. 00-N001, Rockville, MD: Agency for Healthcare Research and Quality.
Philipson, T., and Jena, A.B. “Who Benefits from New Medical Technologies? Estimates of Consumer and Producer Surpluses for HIV/AIDS Drugs.” 2006. Forum for Health Economics and Policy 9(2), Article 3.
Philipson, T., and Sun, E. “Is the Food and Drug Administration Safe and Effective?” 2007. Journal of Economic Perspectives, forthcoming.
Philipson, T., Berndt, E., Gottschalk, E., and Sun, E. “Cost-Benefit Analysis of the FDA: The Case of the Prescription Drug User Fee Acts.” 2007. Journal of Public Economics, forthcoming.
Sun, E., Philipson, T., Lakdawalla, D., Goldman, D., and Jena, A.B. “An Economic Evaluation of the War on Cancer.” 2007. Working paper.
Wagner, E.H., Sandhu, N., Newton, K.M., McCulloch, D.K., Ramsey, S.D., and Grothaus, L.C. “Effect of Improved Glycemic Control on Health Care Costs and Utilization.” 2001. JAMA 285(2):182-9.
1 For examples, see Lichtenberg (2002, 2004), Matchar and Samsa (2000), Philipson and Jena (2005), and Sun, Philipson, Lakdawlla, Goldman, and Jena (2007)
2 For example, see Cowper, Delong, Whellan, Lapointe, and Califf (2004); Wagner, Sandhu, Newton, McCulooch, Ramsey, and Grothaus (2001), and Berger, Slezak, Stine, McStay, O’Leary, and Addiego (2001)
4 Philipson, Berndt, Gottschalk, and Sun (2007).
5 This exemption would be predicated on adhering to FDA standards, and would not include situations where a firm misled the FDA.