Drug Approval Overregulation

Michael R. Ward

Michael R. Ward is a staff economist at the Federal Trade Commission. His views are not necessarily those of the FTC.

The Food and Drug Administration regulates drug approvals to increase public health by keeping unsafe and ineffective drugs off the market while allowing pharmaceutical firms to market safe and effective drugs. Thus, the public can think of an approved drug as one that has been demonstrated to meet a standard for safety and efficacy. There is, however, evidence to suggest that U.S. drug approval standards have become more stringent than is socially optimal. Industry studies show not only a dramatic increase in the stringency of the FDA's drug regulation but also a net effect that harms consumers. Those results are consistent with a more general theory of regulation that implies that drug approval stringency will exceed what is socially optimal because the FDA is more adversely affected by approving harmful drugs than by denying approval of beneficial drugs. Successive reforms of the drug approval process have continually attempted to relax the stringency by more fully addressing the bias in the FDA's incentives. Reforms proposed in November of 1991 that reduce the stringency of the drug approval process should be adopted.


The Drug Development Process

The very nature of the lengthy drug development process makes the pharmaceutical industry susceptible to harm from unnecessarily stringent regulations. Basic pharmaceutical research identifies promising chemical or biological properties of newly synthesized or previously known substances. Researchers usually publish their findings in academic journals and make presentations at conferences. Scientists at the major drug firms learn about the findings from those sources as well as through direct contact with the researchers. Drug firms perform their own research on the substances that show the most promise for eventual market introduction.

When a pharmaceutical firm identifies a new therapeutic use for a substance, it usually applies for a "use" patent and begins animal tests to determine safe dosage levels of the drug. To begin human clinical trials in the United States, the firm files an investigational new drug application with the Food and Drug Administration. The firm may begin testing unless the FDA intervenes within thirty days. The tests follow guidelines established by the FDA and are divided into three phases. In the first phase firms test for safety on twenty to eighty healthy volunteers. In the more extensive second-phase tests, using 100 to 300 patients, firms test for efficacy under different dosages. The third phase entails a series of wider-scale trials (usually on 1,000 to 3,000 patients), in which firms test for proof of efficacy and acceptable side-effects. If the drug still appears to be promising, the firm submits the results of the trials to the FDA in the form of a new drug application. Applications are typically 100,000 pages long. Finally, the FDA will approve the drug as a treatment for a specific indication. If it becomes known that the drug is effective in treating a different condition, the firm must file another new drug application. Safety monitoring continues after the firm markets the drug.

We can get a sense of the risks and costs involved in the development process by looking at the number of drugs at each of the milestones in the development process. At the basic research level, each year from 30,000 to 45,000 medical articles deal specifically with drug therapy. From 1979 to 1989 the Patent and Trademark Office granted between 2,000 and 4,200 drug patents annually. Each year from 1964 to 1989 pharmaceutical firms filed between 800 and 2,200 investigational new drugs with the FDA. For the investigational new drugs that represent drugs not previously tested in humans, the probability of entering the second phase of testing ranges from 70 to 75 percent, and the probability of entering the third phase is somewhat less than half of that range. Of the 80 to 250 new drug applications firms file annually, the FDA approves only 20 to 60. Many of those represent reformulations of existing products. In all, firms will market about one out of a hundred of the products for which they have developed patents.

The actual cost of synthesizing drugs is often trivial; understanding how to use them is expensive to discover. The government, largely through the National Institutes of Health, sponsors basic research, which is conducted mainly in academic medical centers but also at government and nonprofit research centers. Researchers at those institutions create knowledge, which they present at conferences and publish in academic journals. Knowledge is a public good, and elaborate citation and reward systems have evolved to maintain the "property rights" of researchers over their scholarly work. Industry continues research and development when the knowledge it creates can be embodied in a compound and intellectual property right agreements protect its investment. Elaborate, formal testing procedures create new information about the drug and serve to convince the FDA and consumers about the discovered benefits (and risks) of the new drug. Drug promotion is aimed at conveying to physicians--the purchasing agents of end consumers--the value of the discovered information. The expiration of the patent allows others to use the information, and generic drug competition ensues. Still, product differentiation between the brand-name drug and its generic competition exists in part because of the intellectual property right protection of the brand-name trademark and the reputation created around it by the firm.


The Economic Consequences the FDA Drug Approval Process

The FDA is the agency with the greatest impact on drug development in the United States. Largely as a result of the thalidomide tragedy, in which babies whose mothers took the drug during pregnancy were born severely deformed, Congress enacted amendments to the Food, Drug and Cosmetic Act in 1962 to increase FDA oversight of the drug industry. The amendments required firms to show that new drugs were safe and effective before receiving FDA marketing approval. The procedures for demonstrating safety and efficacy have become increasingly costly.

Studies of the effects of that policy change and other FDA policies have found that more stringent regulations have increased the real U.S. drug development costs by an average of 6 percent per year and have reduced by half the number of new drugs introduced in the United States. Those regulations have also increased the share of U.S. pharmaceutical R&D performed abroad. In addition, they have increased patients' suffering by denying them new drugs and have limited the usefulness of drugs by barring their use in self-medication and in proven but non-FDA-approved treatments.


Effects of a More Stringent Drug Approval Process.

Economic studies have documented the dramatic cost increases to the drug companies of developing a new drug. One study estimated the cost of bringing a new drug to the market to be $231 million, a real increase of 230 percent over fifteen years. A readily available and often used proxy for costs imposed by the FDA is the length of time required for drug development. Average drug review times rose from under two years to over three years between 1962 and 1989. Moreover, during the same period the average time firms required to test the investigational new drug to gather data necessary to submit a new drug application rose from three years to between six and seven years. Such measures are important because they highlight the loss of effective patent protection for the drug. That loss was partly remedied by the Waxman-Hatch Act in 1984. Studies have shown that cost increases stemming from the 1962 FDA amendments reduced the number of new drug introductions by half relative to other industrialized countries and created a considerable lag from the time the drug was first marketed abroad to the time it reached the U.S. markets, with no corresponding reduction in the number of inefficacious drugs. The cost in increased mortality and morbidity was valued at $330 million in 1973.

The U.S. pharmaceutical industry has a larger foreign presence today than it did three decades ago. The Pharmaceutical Manufacturers Association reports that the fraction of pharmaceutical industry employees in the United States has declined by 8 percent from 1965 to 1987. The percent of U.S. drug patents granted to foreign sponsors rose from 39 percent during the 1963 to 1975 period to 49 percent during the 1976 to 1989 period. In the early 1960s most of the drugs marketed in the United States originated here. After 1963, much of the initial research moved overseas, where human testing was less regulated.


Effects of Other FDA Policies.

Policies that limit the potential size of the market for a drug will decrease the gains to developing the drug and thus the amount of R&D spent to achieve those gains. Two such policies of the FDA are the banning of "off-label" promotion of drugs by pharmaceutical firms and the enforcement of prescription restrictions.

FDA drug approvals are for particular uses, and firms may promote their drugs only for those uses. Often, however, an approved drug is found to have uses in treatments beyond those for which it was approved. Physicians can prescribe drugs for any treatment. Crude estimates are that unapproved or off-label uses account for at least one-fourth of all prescriptions and that over half of all cancer chemotherapy patients received drugs for off-label uses. Firms may apply for a second use approval, but millions of dollars would be required and only a few years of patent protection remain. A "gray" market developed for information in which drug companies convey new information about existing drugs to practicing physicians by sponsoring symposia and seminars and by distributing journal reprints. The FDA, however, is beginning to clamp down on those activities at the same time that it is seeking to legitimize some off-label uses.

Consumers sought physicians' prescriptions for a large class of drugs before the FDA required prescriptions, and many would still seek a physician's advice without the requirement for a prescription. It is, however, likely that for some drugs the FDA's requiring a physician's prescription is a more stringent regulation than consumers would otherwise choose. For such drugs the hope is that consumers would make better decisions regarding the drugs they consume and that the adverse health effects of mistreatment would decline. Because the requirement for a prescription substantially raises the cost of consuming those drugs, some consumers seek no treatment or use either less effective nonprescription drugs or more powerful and potentially more toxic prescription drugs. Clearly, those substitutes have their own adverse health consequences. At any rate, studies have found that prescription requirements do not seem to enhance the life-saving benefits of modern drugs and do not reduce and may even increase the life-threatening risks from the improper use of those drugs. In addition, consumers stand to gain substantial savings from the FDA's changing the status of some drugs from prescription only to over-the-counter.


Why Are Drug Approval Standards Excessively High?

Why should FDA incentives favor a drug approval standard that entails adverse health and economic consequences? The answer lies in how government agencies such as the FDA respond to the risks involved in decisionmaking. The social costs in the drug approval process are approving a nonbeneficial (and possibly harmful) drug and failing to approve a truly beneficial drug. A more stringent approval standard will decrease the likelihood of the first outcome at the expense of more of the second outcome. The costs to the agency are likely to be different from the social costs because approving a nonbeneficial and possibly harmful drug results in more political pressure than failing to approve a beneficial drug. No official wants to be known as the one who approved another thalidomide. Yet those same officials have not, until recently, needed to worry about the consequences of delaying the approval of a drug. The political pressures on the agency have resulted in its maintaining a higher than optimal standard for drug approval.


The Role of Information in Assuring Safety and Efficacy.

The FDA bars consumers from buying most drugs on their own, allows physicians to prescribe only from an approved list, and prevents drug companies from disseminating information it has not approved. Thus, the agency prevents information from disciplining the market for pharmaceuticals. The ostensible reason for those stringent regulations is the FDA's belief that consumers, physicians, and producers would not act in their own best interest. Yet, the public has experienced the introduction of many new nonpharmaceutical products, such as computer programs and telecommunications systems, without the "benefit" of government guarantees of their safety and efficacy. Competitive forces discipline such markets.

Even in the health sector, with its abundance of government certification requirements, there are examples of market introductions without government assurance of safety and efficacy. Most chiropractors, acupuncturists, and even doctors and nurses do not meet FDA-type standards for safety and efficacy. Are all newly licensed health care providers more effective than the current state of the art? The market rewards the more effective, and the least effective exist only where alternatives are scarce or costly. Advertising, as in the market for optometric services, plays a key role in disseminating information about medical care providers' efficacy. Such advertising has reduced consumers' costs without impairing the quality of service they receive.

In the drug industry itself there are at least three examples of how information helps discipline the market. First, a drug introduced twenty years ago is likely to be replaced by a current product because information about competing drugs has effectively removed the product from the market. Second, prescription restrictions do not improve health because patients are able to find information about benefits and risks elsewhere. Third, the experience with off-label uses of drugs in cancer therapy suggests that the market works quite well at sorting out the benefits of different drugs without (or despite) the intervention of the FDA. Chemotherapy patients are not being victimized by unscrupulous drug companies and ignorant doctors. They are enjoying an improved chance for better health given to them by the parties most knowledgeable in the area.


Attempts to Expedite the Drug Approval Process

Measures to reform the drug approval process have focused increasingly on the decision bias that minimizes unsafe or ineffective drugs but neglects the costs of denying safe and effective drugs. The success of those measures in lowering the drug approval standard has been modest. The current package of reforms has the best chance of success because they explicitly focus on measures to alleviate the FDA's decision bias.


Past Reform Measures.

There have been a variety of measures undertaken to relieve the burden of the FDA from the drug development process. By the mid-1970s, it became apparent that the FDA's interpretation of the 1962 FDA amendments was requiring more expensive, longer tests for drug approval. As clinical tests became more complicated and required more trials, the time from the granting of the patent to market approval rose from five to seven years in 1960 to nine to twelve years in 1975. That reduced the average time that a marketed drug enjoyed patent protection by half. To partially alleviate that burden, the FDA devised a fast track for what it deemed the most "important" or potentially beneficial new drugs. Better coordination between the sponsoring firm and the FDA and end of phase reviews would quicken the approval process. A study of the fast track provision found that important drugs enjoyed no net benefit and that the remaining drugs experienced a longer approval process.

In 1983 the FDA instituted an orphan drug program to encourage drug development for diseases that affected populations too small to elicit research by the drug companies. The encouragement was in the form of R&D subsidies and, more important, a seven-year marketing exclusivity. Such intellectual property protection is much broader than a patent since it provides protection from entirely different products entering the market rather than just duplications of the existing product. If the purpose of that program was to provide a "subsidy" to undo the effect of an overly stringent drug approval "tax," the appropriate remedy is relaxing the drug approval standard, not regulating around a regulation. At any rate, it is not clear why the regulatory requirements for rare diseases differ from those for more prevalent diseases.

The Waxman-Hatch Act of 1984 extended the patent life of many brand-name drugs but eased the approval requirements for generic versions when the patent expires. The net effect of those two measures on the incentives to undertake pharmaceutical R&D is ambiguous, since the first increases intellectual property rights and the second tends to diminish them. Regardless of the net effect of the act on intellectual property rights, it is likely that, overall, consumers have benefitted, since they would gain from the lower prices of generic drugs.

The tragedy of AIDS also brought the relative costs of rejecting a beneficial drug and of accepting a nonbeneficial or harmful drug into public policy debates. For people who only have a few years to live, the last of which are bedridden and painful, the cost of an ineffective, or possibly harmful, treatment is low, while the cost of foregone treatment is high. Well-organized, vocal, and knowledgeable about the approval process, AIDS activists sought changes in the FDA's procedures to allow wider and earlier access to potential treatments. Those include the treatment use of investigational new drugs in the third phase of testing, a more liberal interpretation on restrictions governing the importation of drugs by individuals, speeded approval procedures for life-threatening diseases, and parallel testing and treatment tracks for investigational new drugs for AIDS- or HIV-related illnesses following the first phase of testing.


Current Drug Approval Reform Proposals.

Eleven new reform measures of the drug approval process, constituting the most extensive reform package since 1962, were announced in November of 1991. They are expected to reduce the drug development time from an estimated 9.75 years to 7 years or 5.5 years for drugs treating serious diseases. Those reforms are attempts to better balance the costs of rejecting a beneficial drug and of approving a nonbeneficial or harmful drug and to put checks on the authority of the FDA. The most important reforms are a new interpretation of the efficacy standard, accelerated approval of drugs for life-threatening diseases and illnesses for which there is no current treatment, the expanded use of outside experts including external review of drug applications, and the harmonization of international drug approval standards.


The Efficacy Standard.

The FDA has had considerable flexibility in interpreting the efficacy standard. Over time that standard has become more difficult to overcome, but under the proposal it will be interpreted in a manner that "maximizes rather than limits the drug's potential for approval and takes into account the risks to human life and health that result from delay of new treatments." Surrogate endpoints will be used to implement the new standard. An example of the use of a surrogate endpoint is a drug that increases the human body's production of an antibody known to lead to better health of the patients. The clinical trials need not show that the drug leads to increased health since the increased production of the antibody is a surrogate for it. The safety standard will remain unchanged.


Accelerated Drug Approval Process.

The approval process will be accelerated for drugs treating conditions that are life-threatening, serious, or severely debilitating or conditions for which there is no satisfactory alternative therapy, regardless of their severity. Specifically, market approval will be granted before third-phase testing. For some cases the FDA may require that testing to be done in postmarketing studies. Also, the FDA has indicated a willingness to broaden the definition of safety for effective but toxic drugs by tying drug distribution either to the performance of specific monitoring tests or to specially equipped facilities and specially trained physicians.


Outside Experts.

The use of outside experts on advisory committees and institutional review boards will be expanded, and external review of drug applications will be initiated. Advisory committees for each of the therapeutic classes of drugs currently become involved in the drug approval process only after the FDA has reviewed an application. Under the proposal they will monitor the progress of new drug applications and advise drug sponsors about the design and number of clinical studies necessary to test investigational new drugs. Currently, institutional review boards assist in the design and review of clinical studies. Under the reform they will be able to grant investigational new drug status without further FDA review. The FDA will use outside experts to conduct clinical reviews for eight to twelve drugs over a sixteen-month period. The FDA need not conduct its own analysis of the data but retains final authority over drug approval. Safeguards against conflict of interest and drug sponsor influence of the reviewer will be put into place.


International Accords.

Under the proposal the FDA will negotiate with other countries to develop a system for the international recognition of individual country drug approvals. Specifically, the FDA will seek reciprocity of drug approvals with other countries, consider the clinical trials conducted for drug approval in a foreign country, and develop common international standards for new drug clinical trials and marketing submissions. If instituted, those measures could lower the cost of U.S. drug approval somewhat but would considerably lower the cost of foreign drugs introduced for the U.S. market. Such a measure would tend to increase the size of the market for any single drug and thus increase the amount of R&D undertaken.



The proposed drug approval reforms will better align the FDA's incentives with social objectives. The use of surrogate endpoints and approval after the second testing phase for treatments for serious diseases will constrain the FDA from imposing unnecessary delay. Outside review of new drug applications, expanded use of institutional review boards and advisory committees, and international drug approval reciprocity place the decision to approve a drug in the hands of agents with a less pronounced incentive to minimize political risk and thus provide the best hope for lasting reform. Taken together, those reform measures will shorten the drug approval process, increase the number of drugs available while reducing their cost, and improve the quality of life for Americans.

Selected Readings

Council on Competitiveness. "Council on Competitiveness Fact Sheet: Improving the Nation's Drug Approval Process." Washington, D.C.: 1991.

DiMasi, J.A., Hansen, R.W., et al. "The Cost of Innovation in the Pharmaceutical Industry." Journal of Health Economics, Vol. 10 (1991).

Haas-Wilson, D. "The Effect of Commercial Practice Restrictions: The Case of Optometry." Journal of Law and Economics, Vol. 29 (1986).

Siegel, J. and Roberts, M. "Reforming FDA Policy: Lessons from the AIDS Experience." Regulation, Vol. 14 (1991).

Temin, P. "Realized Benefits from Switching Drugs." Journal of Law and Economics, Vol. 35 (1992).

Regulation is published four times a year by the Cato Institute. Editorial and business offices are located at 1000 Massachusetts Avenue, N.W., Washington, D.C., 20001. For subscription information, please write to Circulation Department, Cato Institute, same address, or call (202) 842-0200. Send email inquiries to subscriptions@cato.org, or subscribe online via the World Wide Web at: http://www.cato.org/pubs/regulation/reg-ordr.html

| Regulation | Home | Order Regulation | Publications | Search