New prescription drugs are developed and tested for quality, safety, and efficacy by the pharmaceutical industry, and little or no drug testing is conducted by governments in modern industrialised countries. Governments have regulatory authorities which have a legal duty to protect public health by ensuring that new drugs are not licensed unless they are of adequate quality, safety, and efficacy (box ​(boxB1).B1). The thousands of birth deformities and deaths caused by thalidomide focused public and professional concerns on how the commercial interests of pharmaceutical companies may diverge from, or conflict with, the interests of patients and public health. The reasoning behind the creation of new government regulatory authorities in the post-thalidomide era was therefore that they should be “entirely independent” of the commercial interests of the pharmaceutical industry and should act on behalf of the public interest by checking the adequacy of the test data produced by the industry.^1–3 I explain how these government regulatory authorities in the European Union, which were initially established to provide independent scrutiny of pharmaceutical firms in the interests of public health, have become increasingly responsive to the commercial interests of the industry (box ​(boxB2).B2).

European governments accepted these industry perspectives, although new drugs had in fact come to the UK or German markets in 1972-83, faster than in France, Italy, Sweden, or the United States, and more new drugs were first marketed in the United Kingdom or Germany in 1961-85 than in Austria, the Benelux countries, the eastern European countries, Italy, Scandinavia, Spain, Switzerland, or the United States (figs ​(figs11 and ​and22).⁶ Indeed, in 1988, the United Kingdom was found to have the fastest approval times for new drugs in the European Union.⁷ Moreover, in 1989, German regulatory staff argued that the speed of drug approvals was but a crude measure of efficiency, and that the drug review process was not delaying any genuine therapeutic advances because the backlog related only to products for which existing treatments were adequate.^8–11

Since January 1998, new drugs can be licensed in more than one European Union country at once in one of two ways: mutual recognition or centralised procedure (box ​(boxB3).B3). To accommodate the industry's desire for more rapid approval times, a strict timescale—210 days—has been prescribed by the European Commission for the mutual recognition procedure. Moreover, the largest amount of regulatory work (and hence the most fees from industry) comes from being the first state to which an application is submitted within the mutual recognition procedure, and being the rapporteur in the centralised procedure. As the regulatory agencies in the European Union are now largely funded by industry, and because companies look for fast approval rates as a key criterion when choosing where to submit the application for approval of a new drug, regulatory agencies are, in effect, competing with each other for “regulatory business” by attempting to approve drugs at an ever faster pace.

The competition for fees from industry means that regulatory agencies are not satisfied to meet the European Commission's requirement of approval or rejection within 210 days. For example, the British Medicines Control Agency's average net in-house assessment time for new drugs fell from 154 working days in 1989 to 44 days by 1998 (fig ​(fig33).⁴ The drug regulatory review times for Germany, Sweden, other EU countries, and the United States have also fallen dramatically.^4,20 It might be argued that this competition between agencies will enhance drug safety and efficacy, but even if competition were to encourage some regulatory agencies to try to improve their safety and efficacy regulation, this would be in a context of increasingly less time for all EU regulatory agencies to check drug safety and efficacy data.

The suggestion that such interagency competition is driven by concern to improve safety and efficacy is further undermined by the fact that different approaches have been taken towards rejections of new drug applications and rejection of “old” unproved products at present on the market. The Swedish regulatory agency has extended the response period for a company whose product licence application faced rejection from six weeks to three months. Meanwhile, in Germany, “old” products, whose efficacy has never been demonstrated against the modern standards of the 1976 German drug law, were granted extended “licences of right” until 2004. Despite the potential disadvantages of these measures for the public, they were not regarded by the government authorities as “inefficient.”⁴ Evidently, the time taken for regulatory processes and decision making may be extended rather than accelerated if this is in the interests of the industry.

Moreover, under the mutual recognition procedure, a heavy emphasis on widening the scope of new drug approvals coexists with institutional incentives not to block approval. If a member state does not wish to accord mutual recognition to the approval of a new drug by another member state on public health grounds, it must immediately inform the company, the reference member state, the other member states concerned, and the Committee for Proprietary Medicinal Products, stating the reasons for its decision and indicating how the gaps it perceives in the new drug application might be filled so as to facilitate mutual recognition. A compulsory conciliation stage then follows to facilitate the member state's recognition of the reference member state's approval.⁴ Thus, regulators are under pressure to adopt quickly a position on the reference member state's approval, and to assemble robust supporting evidence if they propose to reject an application on public health grounds. There is no such pressure if they intend to mutually recognise the reference member state's approval.

Drug regulation has gradually moved away from its public health mission. National medicines agencies and the European agency are now serving first and foremost the pharmaceutical companies, who provide their main funds. So far, independent organisations of patients and health professionals in most European Union countries have either been almost ignored or patronised by regulators.

In this context the current revision of the directive and regulation on human medicines is a source for concern. Overall, the drafts are in favour of drug companies, and it is not surprising that the pharmaceutical industry has jumped at the revision, in the hope of making the legislation even more flexible and favourable to short term competitiveness. The fact that a trade oriented directorate of the European Commission, instead of the health directorate, is in charge of EU drug policy probably explains why public health interests are overlooked.^1-3

A healthy debate is under way in the EU parliament, fuelled by amendments coming from outside the industry. Some members of the European parliament (MEPs) have realised the threats. These include shortening of drug evaluation without regular and well conducted reassessment of risks and benefits; continuing secrecy of regulatory authorities, especially regarding pharmacovigilance^1-4; priority given to industrial demands delaying generic development; and advertising direct to the consumer disguised as information.

The Committee on the Environment, Public Health, and Consumer Policy of the EU parliament has requested a delay, and voting, initially due before the summer, has been postponed to this autumn. More than 500 amendments were tabled, many of them favouring public health and patients' interests. Members of the Medicines in Europe Forum met with MEPs. Thousands of petitions, many from readers of La Revue Prescrire, were sent to MEPs, and around 2500 have recently been addressed to the president of the EU parliament. German independent drug bulletins have jointly expressed concerns over the commission's proposals. Members of the licensing committee (the Committee for Proprietary Medical Products) have voiced similar concerns to the president of the EU parliament.^1-5

The drafts have to undergo the complex co-decision process involving the European parliament and the Council of Europe. Health ministers of member states are in charge of the drafts—these are not, as was initially feared, solely in the hands of ministers of industry. The health ministers seem divided on the issue of procedures for authorising marketing, and on the respective roles of the European and national agencies, but they are aware of the detrimental effects of direct to consumer advertising. Some ministers are under public pressure to ensure the financial independence and accountability of their regulatory authorities. A number of proposals in the directive are, however, so controversial that the discussion will go on at least until mid-2003 or 2004.

During the period in which this alignment emerged and the regulatory process in the European Union became more efficient, the time required for the review of a drug's efficacy data by the United Kingdom Medicines Control Agency decreased from154 working days in 1989 to 44 in 1998.^2-1 Drug approval times in the United States fell during the same period, affecting the two most critical stages of drug development and evaluation. Between 1989 and 1998, the average time required to advance new chemical entities—which represent truly breakthrough rather than “me-too” drugs—from the investigative stage to broader clinical evaluation fell from 76.0 to 70.3 months, and the subsequent average time to full market approval fell from 34.4 to 16.8 months.

One of the clearly intended consequences of this growing “problem” is the rapid increase in the number of new chemical entities and other new drugs finding their way onto the market and into patients. During the period studied by Abraham, the mean number of new chemical entities approved by the US Food and Drug Administration, for either expanded clinical evaluation or market launch, increased from 18 to 37 per year. The mean number of all new drugs approved—new chemical entities plus “me-too” drugs—increased from 63 in 1991 to 98 in 2000 (FDA data).

Has this regulatory efficiency and its consequent gush of new drugs, sparked by the alignment of the approval process with commercial interests, put patients and the public health at risk? The data argue that it has not. During the 1990s, the total number of drugs withdrawn from the US market because of concerns about patient safety remained constant, in the range of one to three a year. Between 1984 and 1988, long before the adoption of the regulatory reforms outlined in Abraham's article, 3.5% of newly approved drugs were withdrawn from the US market because of concerns about safety; between 1994 and 1998, as the regulatory process changed in ways decried by Abraham, this number decreased to 1.2%.^2-2 This decrease is good news for the public health that Abraham's paper purports to protect.

Abraham argues that disproportionately less testing of older drugs “grandfathered” on to the market further betrays a regulatory process beholden to commercial interests. Again, an obvious response to this “regulatory lapse” should be a collective shrug. The best way to wean doctors off prescribing older, ineffective drugs is not through expensive proof—via scarce regulatory resources—of what is usually clinically obvious, but through even more rapid release of newer and better drugs.

In the United Kingdom one in seven people who are given a diagnosis of Parkinson's disease is under 40. I found it inconceivable that I should be one of them. In 1987, aged 37, I learned that to keep me mobile I would have to rely on drugs for the rest of my life. Silently and stealthily this insidious, progressive disease has taken over. Some days I blunder through my “off” periods in a fog, trying to finish whatever I had started a few moments previously. Tasmar (tolcapone), launched in Britain in 1997, relieves the “on /off” state in Parkinson's disease and subsequently improves mobility and function.^3-1,3-2 To those of us prepared to put up with monthly liver function tests, this drug was a lifesaver. Instantly, my “off” periods were dramatically reduced and the rigidity and bradykinesia, so typical of young onset Parkinson's, disappeared.

But Tasmar was withdrawn in Europe in 1998 after a few reports of severe hepatotoxicity. How does this compare, for example, with the number of strokes among women taking the contraceptive pill? Quality of life for me and many other people with Parkinson's was shattered. There were days when I would just sit there, fighting back the tears, remembering how easy it used to be to walk across the room or pick up something that I had dropped, and feeling bitter at the European drug regulators.

But a few of us who can afford to do so choose to obtain Tasmar privately, and import it on prescription from the United States, at a personal cost of approximately $700 (£450; €715) every three months. Most do not have this choice.

Healthcare professionals and policy makers need to understand that their own perceptions of health may differ from those of their patients.^3-3 Patients need to become empowered—to take control of their illness and so ensure that it does not impair their quality of life. Patients have knowledge and experience—and a desire to manage their own condition—and these three factors should be harnessed to ensure that resources are used wisely and services provided appropriately.