Editorial Note:  the King's Fund Centre breakfast debate, 3 March 1998 Proposer: Professor Michael Rawlins, Chair, Committee on Safety of Medicines Motion opposed by Charles Medawar, Director, Social Audit Ltd




1. Obsolete, misconceived…?

The drug licensing system should not be considered as part of a national drug policy, because the UK doesn't have one. We have instead a jungle of UK and European laws, introduced to address specific problems, none primarily to do with health. The 1968 Medicines Act was established to provide some basic regulation of the pharmaceutical industry, where previously there was none.

The Medicines Act is also the product of an age when patients had few rights. In the UK, at that time, the law prohibited consumers even from knowing the names of drugs prescribed for them, and harmful experimentation on unwitting humans wasn't proscribed (under the Declaration of Helsinki) until 1966. So it is not surprising that the 1968 Act makes no provision for consumer/patient involvement or participation. If the world has moved on since then, the drug licensing system has not.

The 1968 Act is commonly thought of as a response to the Thalido-mide disaster but, by the time it came into effect, that drug had been off the market for nearly ten years. The Medicines Act was in fact the product of an enquiry into The Relationship between the National Health Service and the Pharmaceutical Industry".1 The point was that the crisis had revealed that the industry was out of control, that the market was flooded with "undesirable" drugs, and that the NHS was paying for them.

The Act was a very much a first step in regulation, and the system that has grown up around it is mainly the outcome of successive, modest enough attempts at damage limitation and to secure 'fair trade'. Naturally, all this happened in the face of intensive lobbying by the main affected interests, especially pharmaceutical companies. Their success in this is underlined by the official sponsorship of the industry, not by the Department of Trade, but by the Department of Health.

Let it be said that the pharmaceutical industry detests regulation and puts about a lot of money to resist it. Drug company money reaches into every part of the system, and the industry has an almost legendary reputation for letting others get its own way. Nowadays, it even pays for the Medicines Control Agency and the Committee on Safety of Medicines. Through license fees, companies pay for all core activities in the UK regulatory system. This clearly has an important effect on the MCA/CSM budgets and on regulatory policies and practices too.

Commercial considerations also affect regulatory style. Just a few years ago, the Medicines Control Agency used to charge a 150 entry fee to its Annual General Meeting. This indicates not only whom the MCA sees as its main constituents, but also its limitations as a public agency working towards the solution of public problems. 2

Increasingly, however, the scene is now shifting to Europe, and drug licensing procedures will be defined by the EU. Does this make the current system more acceptable for patients and consumers? On the contrary: the law was created by and is still administered through DG3, the European Commission's Directorate for Competition. There has never been an EC Directorate for Health.

In short, the drug licensing system we have today does not express any clear policy about how we want use medicines and to what effect, and the public would quite mistaken if it believed that, thanks to government, patients can trust that medicines are effective and safe.


2. Secretive, unresponsive, unaccountable …?

Professor Rawlins' predecessor, when in office, said he thought that agencies like the MCA and CSM should "be immune from political and public pressure and above all free from the pressures of action groups". This word "immune" seems significant for its connotation of extreme resistance and lack of response. I offer two glimpses of evidence - both relating to the stranglehold of office - to suggest nothing much has changed.

The first involves Professor Sir William Asscher, the previous Chairman of the CSM, shortly after his retirement, by which time his Knighthood was secure. At a meeting organised in 1993, in support of a private member's Medicines Information Bill, Asscher declared:

"many's the time that I have had frustrations about not being able to communicate because of section 118 (of the Medicines Act), confining me to prison to two years if I were to communicate in the way that I would normally be able to do in my own medical school." 3

The second illustration concerns the present Chairman, Professor Michael Rawlins. Writing to me in his personal capacity, ten years ago, he volunteered that he thought the medicines control system excessively and unnecessarily secretive. Very recently I asked him if he still thought so, and if I might attribute the following opinion to him. He agreed on both counts, which says more about the calibre of the man, than for the system he helps to run:

"Much of what you (and I) seek would require repeal of section 118 of the Medicines Act which is the section concerned with "secrecy". A small part of individual applications might be regarded as commercially confidential but the major proportion could, with little loss to anyone, be made publicly available …"

Recently, the MCA has claimed it has become more open - now even disclosing the counts of Yellow Cards, adverse drug reaction reports sent in for individual drugs. This was indeed a significant step forward - for reasons Asscher had explained at the aforementioned meeting in the House of Commons in 1993. He supported the Medicines Information Bill, he said, but definitely wouldn't want to see disclosure of adverse drug reaction reports. Why not? It was because this information "is exceedingly difficult to interpret and it requires, like a good wine, a certain degree of maturation and thinking about in order to ensure the right message gets across …":

"If we released this, the journalists would have a feast of drug scandals, a real feast they would have - and a lot of good drugs would come off the market, drugs where the benefit-risk ratio is in fact quite good."

When, last year, the MCA/CSM did finally release ADR data of course no mayhem ensued. However, if there has been a slightly increased trickle of useful information in recent years, it is not primarily the result of any internal reform. It is explained mainly by the introduction of the 1994 Code of Practice on Access to Government Information and, more recently, by this government's welcome proposal for Freedom of Information law. The Yellow Card data were released only because Social Audit complained under the Code; it took nine months of serious badgering, with the MCA ducking and weaving throughout, and the CSM nowhere to be seen.

The present medicines control system remains miserably secretive. Over the years, there have been numerous, sometimes serious drug problems, but no public enquiry into drug safety has ever been held. This matters not just because it hides data and diminishes science, but also because it perpetuates two great vices in medicine - the belief that meaning well means doing good, and the consistent inability to admit mistakes and learn from them. Both should be counted as serious obstacles to patients' safety and public health.

Nor does parliamentary scrutiny of the system make accountability much more of a reality. MPs are overwhelmed with other issues, lack resources and are typically intimidated by drug issues. They are also very much subject to party policies and treasury-inspired disciplines, and therefore pretty much out of touch.

All of these factors make the present system of medicines control unusually unaccountable, and this is bad for patients and bad for the spirit of medicine itself. Secrecy and lack of accountability also have much to do with the major operating deficiencies in the present control system - of the kind elaborated in "The Antidepressant Web" 4


3. Conflicts of interest, cartels …?

From a consumer perspective, the present system operates pretty much as a cartel - a trinity of medical professionals, government and commerce. Between them, they confidently and confidentially take far-reaching decisions "in the public interest". Over the past 30 years, the leadership in the cartel has unmistakably passed from the medical profession to the industry, all the more reason to be concerned.

However, the authorities do not seem concerned. Most members of the CSM now have direct and/or indirect financial interests in pharmaceutical companies. This practice is defended on the grounds that members have to declare their interests (but not those of their families5), and members with conflicts don't vote. It is also emphasised that contacts between regulators are industry are necessary if CSM members are to stay well informed - a point that seems unpersuasive, since the CSM Chair is required to hold no personal interests at all.

In The Antidepressant Web, I give what I hope is an extreme example of the kind of thing that happens in practice. Around 1990, the CSM's leading expert on the so-called SSRI antidepressants (eg Prozac) had personal financial interests with four leading manufacturers of SSRIs and indirect interests in another one. He seems to be what is called in America a "Marquee Professor", someone who really gets around and whom companies can rely on to say nice things about their products. This man's work undoubtedly had a profound influence on licensing decisions made by the CSM and, in my view, has led to fundamental misunderstandings about the long-term effectiveness of these drugs.

Very briefly, I fear that what the regulators now think of as the long-term effectiveness of antidepressants is in fact a manifestation of drug dependence - just as it proved to be with tranquillisers, such as Valium and Ativan, only a few years ago.

While I am also very much at odds on this issue with the Royal College of Psychiatrists, their President did acknowledge in recent correspondence6 that the close links between this particular CSM expert and the pharmaceutical industry were "well known, and his views are generally treated with caution for that reason". I wonder if the CSM now thinks the same.

So how close are the regulators to the pharmaceutical industry? I particularly like the description of the present state of play by Dr Phillip Brown, the publisher of SCRIP World Pharmaceutical News, a mine of good intelligence for the industry (and a pity few others can afford it). Brown has concluded that drug regulation in the UK is "dull and uninspiring" and "self-serving", and has "little relevance to the safety of the patient in the real life situation". Apart from the curse of secrecy, he sees the main problem as the closeness between regulators and regulated, and he illustrates this by analogy with relationships in the top security of Parkhurst prison:

"Prison inspectors have found that in goals like Parkhurst, prisoners and gaolers develop close interpersonal relationships, to the extent that it is very difficult to distinguish between them, except at the close of day when the gaolers lock up the prisoners and return to the comfort of their homes" 7

Brown doesn't precisely say whether he sees the regulators as gaolers or prisoners, but it may well have occurred to him that it could be the regulators who are imprisoned - by niggling law, lack of resources, and mountains of data, etc - while the industry enjoys (and often bestows) great comforts. Dr Duilio Poggiolini is a case in point.

Until 1993, Poggiolini was Chair of the key drug agency in the European medicines control system, the Committee on Proprietary Medicinal Products. The CPMP is, in the EU control system, the equivalent of the CSM, over here. Anyway, Poggiolini was also the top man in the Italian drug control system, until brought down in the Italian anti-corruption drive. He was charged with having salted away, over about 30 years, $180m dollars in bribes from pharmaceutical companies. Following his arrest, he and his wife were found to have seventeen bank accounts between them; their safe contained a variety of gifts, ranging from valuable paintings to gold coins.


4. Inadequate scrutiny, negligent of patient interests …?

Regulation entails checking that marketed drugs are of sufficient [a] quality, [b] efficacy and [c] safety. This basically means selective scrutiny of vast quantities of data generated by manufacturers, the results of tests and trials on relatively small and highly selected groups of patients, in some (qualitative and quantitative) respects quite unrepresentative of the patients prescribed those drugs in the real world. The scrutiny is mainly handled by MCA staff or, under the European system, by an assessor nominated by the company making the licence application.

[a] Since inspection reports and recall notices are kept secret by law, I cannot confirm my impression that, as a rule, lack of drug quality (ie purity etc.) has not proved a serious problem. Though there have been exceptions, involving the manufacture/distribution of certain vaccines, I would guess the companies and the MCA between them do a good job and keep things on a fairly even keel.

[b] The system is geared to establishing whether medicines are "efficacious", when it should be trying to find out what medicines do and do not contribute to health. The word "efficacious" does not mean "effective"; it means that the drug does what it claims to do, and can be used and promoted for specified "indications". However, the claims made in the licensed indications may be so bland, yet so sweeping in their terms as to make "efficacy" mean almost everything and yet nothing. Prozac, for example, is licensed "for the symptoms of depression", which is to stop just short of saying that it is tough on stains but kind to your hands.

What patients and prescribers want to know is that a drug is effective in procuring health - but this is not something guaranteed, or even much addressed by the present system. Increasingly, for example, you will find licences granted when surrogate markers of effectiveness are used, which may prove to have little relevance to health. Some cholesterol-lowering and anti-arrhythmic drugs would be examples of this. Indeed, a product license in itself is a surrogate marker of health and, whatever is implied, gives no indication of therapeutic value for money at all.

[c] The fatal flaw in the part of the control system concerned with safety, is the preoccupation with "the safety of medicines" rather than the safety of the patients who take them. Equally, the CSM is fond of pointing to its track record "by examining the fate of all new licensed drugs", rather than the fate of patients. The evidence suggests that many do not fare well.

I know of no evidence to suggest that the situation has improved 8 since Rawlins (1981) 9 estimated that of the order of 3-5% of hospital admissions are wholly or partly the result of harm done by drugs taken with therapeutic intent. Successive surveys have confirmed this though, with elderly patients, the proportion probably rises to about 10% of hospital admissions. In addition, some proportion10 of in-patients have their stay in hospital prolonged by an adverse drug reaction (probably half of which might have been avoided). 11

I have asked many times for an official estimate of the extent of accidental drug injury in the UK, but have so far always failed. The only response has been to rubbish the best estimates I have been able to produce - these partly based on data from the US data, where prescribed drugs are believed responsible for 140,000 fatalities/year and where the cost of all drug injury is now estimated at around $150bn/year.11 On this basis, I suspect that, in the UK, each year, prescribed drugs cause a few tens of thousands of deaths, and a few hundreds of thousands of hospital admissions, plus a few millions of sometimes quite troublesome side effects, usually never counted at all. Examples of these would include impotence with antidepressants or beta-blockers, dependence and cognitive impairments with benzodiazepine tranquillisers, or movement disorders with a wide range of antipsychotic drugs.

But who decides whether the benefits are worth such risks? Initially, it is the MCA and CSM but, in their preliminary assessments, "safety" is regarded as an essentially molecular property. The reality is (or should be) that patients' safety is fundamentally to do with how well medicines are used, and with the acceptability of drug risk. The latter should be a matter for patients themselves, in consultation with doctors - but, in the absence of reliable data, these are clearly difficult decisions to make. Nor is the process helped much by the activities of pharmaceutical companies whose promotions naturally tend to celebrate the triumph of benefit over risk. The regulators have little control over drug promotion and prescribing activities, increasingly less as the industry makes deeper and deeper inroads into the system.


5. Bad science, lack of intellectual hygiene …?

When new drugs first come to market, not a great deal is known about their safety profile; it usually takes years, even decades, for the full facts to emerge. There is a simple test of this: put a ruler alongside the "warnings" or "side effects" sections in the drug data sheets issued, first, when the drug was launched, and then at later stages of the drug's life. The small-print warnings thicken and extend - though they still often appear to be designed mainly to protect, not patients, but backs.

The MCA and CSM are basically responsible for checking safety before drugs are marketed, not after. They do make re-assessments of safety but, in general, this is a passive and unsystematic activity and "fraught with a range of legal and practical impediments". The basic posture of the regulators is to wait and see - very much assuming that no news is good news, and that what they see is what there is. The strong tendency is to act, not when there are clear signs of risk, but only when there is compelling evidence of harm.

The main instrument of control over drugs on the market is the Yellow Card system - a scheme in which individual prescribers can report to the MCA/CSM what they suspect to be adverse reactions, on the basis of their patients' experience with individual drugs.

This system was started over 30 years ago, but still the vast majority of doctors never report most of the ADRs they see. The CSM itself estimates that, even with serious reactions, it rarely gets more than one Yellow Card for every ten reportable events, but I believe even this estimate to be very high.

Equally, the regulators greatly over-estimate the numbers of doctors who participate, since assessments are now based on what doctors say they do - without checking that that is what actually happens. Griffin (1984), 12 who did check, estimated that, in the first 20 years of its operation, only 16% of NHS doctors had ever sent in a Yellow Card, and that 7.4% of doctors had sent in 80% of all reports received. As the absolute numbers of Yellow Cards have since increased by only about one-third, I believe it misleading to suggest, as the MCA/CSM have recently done, 13 that "the proportion of doctors who have ever sent in a Yellow Card is encouragingly high".14

Nevertheless, the MCA/CSM (also the industry) greatly rely on the Yellow card system. They see it as the justification for quickly getting drugs to market - the reasoning being that the flow of Yellow Cards will alert the regulators to any problems there turn out to be. It is true that the Yellow Card system has flagged certain problems with particular drugs, but it has missed a great many too.

The evidence is necessarily limited, but what there is clearly suggests some generic fault in the system. Thus in 1980, the regulators concluded, on the basis of a "systematic review" that dependence on benzodiazepine tranquillisers was a very rare event. They decided this on the grounds that [a] an industry-linked evaluation15 of transparently lamentable quality had concluded the same; and [b] because the regulators had received only 28 Yellow Card reports of dependence over the previous 17 years. Five years later, the best estimates indicated that between 250,000 - 500,000 people were affected in the UK.

And the question now, with antidepressants, is whether the same thing is happening again. As described in The Antidepressant Web, the MCA and CSM in 1996 published an analysis based on Yellow Card returns which concluded that withdrawal symptoms from antidepressants were very rare - affecting something of the order of one patient in 10,000. My best estimate would be of a problem affecting at least one in ten. I am very nearly sure that the MCA/CSM is overlooking a problem with the antidepressants to compare with that for benzodiazepine tranquillisers. The Table below, based on the MCA/CSM's data, partly explains why:


Introduced Ballpark No. Yellow Card reports of:
of NHS scripts withdrawal reactions dependence- all forms
Temazepam (BDZ) 1977 50m 5 3
Diazepam (BDZ) 1963   180m   20    16
Fluoxetine (SSRI)





Paroxetine (SSRI) 1991 6m   802 9

BDZ = benzodiazepine tranquilliser; SSRI = Selective Serotonin Reuptake Inhibitor antidepressant)

On the basis of this and other case studies,16 I would be inclined to argue that the quality of scientific assessments by the MCA/CSM is often uncertain, and at times a disgrace. Factors other than secrecy have helped to obscure this. Thus, from the recent furore over third- generation oral contraceptives, the public might have inferred that just the suspicion of an excess risk of the order of two or three per thousand might be enough to get the regulators to leap into action. They would be profoundly mistaken if they thought so.


6. Conclusions …?

Because this is a debate, I have tried to give a balancing, rather than balanced account. In the available time17 it would have been hard to speak also about some of the advantages and successes of the present system, mentioning also the commitment and hard work of many individuals involved, and the obstacles that beset them. Nevertheless, I consider this motion is pretty much a travesty and that the public deserves much better than it now gets. Bear in mind too, that I have mainly focused here on what regulation does, and sets out to do - leaving aside major areas in which it makes hardly any impact at all.

My underlying fear relates to the development of a 'fast food' culture in medicine, to the provision of what I once heard a disillusioned drug rep call, "pills for losers". In a system in which drugs have become overwhelmingly the dominant treatment mode, the risk is that people lose confidence in themselves - and that their health suffers profoundly as they come to believe that the true source of their wellness comes from a pill bottle, rather than from something from within themselves.

We would be far healthier, much better off, with a system which instead operated within a framework of honest science and decent democratic and human values - which the current system does not.

That is where I would like to see the drug control system heading - and this is where I come from, in the words of WHO:

"Science and technology can contribute to the improvement of health standards only if the people themselves become full partners of the health care providers in safeguarding and promoting health … people have not only the right to participate individually and collectively in the planning and implementation of health care programs, but also a duty to do so." 18




1. Lord Sainsbury (Chairman), Report of the Committee of Enquiry into the relationship of the pharmaceutical industry with the National Health Service, 1965-1967, Cmnd 3410, London: HMSO, 1967.

2. This practice stopped, in the early 1990s. I like to think Social Audit played a part in this, mainly by drawing attention to the palpable vulgarity of the occasion: the annual general meeting is, after all, an occasion of reckoning, when organisations are called to account.

3. Asscher was speaking in Committee Room 6 at the House of Commons, on 29th April 1993, at a meeting organised by The Campaign for Freedom of Information in support of Giles Radice's Medicines Information Bill. The Bill (co-sponsored by Social Audit and the National Consumer Council) was aborted by industry lobbying with the active connivance of government and the Department of Health.

4. C. Medawar, The Antidepressant Web - Marketing depression and making medicines work, International Journal of Risk and Safety in Medicine, 1997, 10 (2), 75-126. Also at http://www.socialaudit.org.uk

5. Some wives and families hold drug shares too. Present rules do not require these to be disclosed.

6. Letter from Dr R E Kendell, dated 28 November 1997. See http://www.socialaudit.org.uk

7. P. Brown: Drug regulation -a life sentence on the industry? Scrip Magazine, Sept 1996, 23-24.

8. "We know that a significant proportion of hospital in-patients are there because of drug induced disease. In general practice there are even greater numbers of patients suffering unnecessarily as a result of adverse drug reactions" Committee on Safety of Medicines, Help Make Medicines Safer, Videotape, 1988.

9. "Adverse drug reactions (defined as unintended effects of substances used in the prevention diagnosis or treatment of disease) are common. They are responsible for 3-5% of hospital admissions, occur in 10-20% of hospital inpatients, and have recently been reported in 40% of patients receiving drugs in general practice" (M. Rawlins, Adverse reactions to drugs, Brit Med J, 21 March 1981, 282, 974-976).

10. Rawlins (supra) suggests a figure of 10 - 20 per cent, and the Boston Collaborative Drug Surveillance Project estimated that approximately 30% of hospitalised patients experienced ADRs. However, a careful three-year study conducted recently in the US reported a rate of 2.43% ADR rate associated with a significantly prolonged stay in hospital (mean 1.74 days) with an almost 2-fold increased risk of death.

11. In the US, the cost of drug-related morbidity and mortality (140,000 deaths/year) has been estimated at $136bn/year (D.C. Classen et al, Adverse Drug Events in Hospitalized Patients, JAMA, January 22/29 1997, 227, 4, 301-306 and J.A. Johnson, J.L. Bootman, Drug-related morbidity and mortality: a cost of illness model, Arch Intern Med, 1995, 155, 1949-1956.. No official estimates are available in the UK.

12. J.P. Griffin, Is better feedback a better stimulus to spontaneous adverse reaction monitoring? (letter), Lancet, 10 November 1984.

13. M.D. Rawlins et al (eds), Views on the Yellow Card Scheme, Current Problems in Pharmacovigilance, CSM/MCA, 21, July 1995, 8. This report was based on the study by K.J. Belton, S.C. Lewis, S. Payne, M.D. Rawlins and S.M. Wood, Attitudinal survey of adverse drug reaction reporting by medical practitioners in the United Kingdom, Brit. J. Clin. Pharmac., 1995, 39, 223-226.

14. A formal complaint about this was made by Social Audit. Upheld: the MCA/DoH had misled a Member of Parliament in claiming on this basis that 77% of doctors had sent in Yellow Cards.

15. J. Marks, The Benzodiazepines: use, overuse, misuse, abuse, Lancaster, MTP Press, 1978.

16. Notably: C. Medawar, Power & Dependence, Social Audit on the Safety of Medicines, London: Social Audit, 1992. This assessment is also based on work I have done on behalf of solicitors acting for legally aided plaintiffs, on a variety of drug injury cases.

17. Ten minutes was allowed for each speaker, so even this briefing was much abridged.

18. WHO Expert Committee: New approaches to health education in primary health care (TRS No 690), Geneva: WHO, 1983. (original emphasis in this quote)


Charles Medawar
3 March 1998
Contents page
Section 5.1 - Editorial notes & What's New?