From The Guardian, 27 July 1999, 2 & 8 & 17. See also letters 30 and 31 July and 3 August 1999

Drug trials risk to patients

Audit shows flawed tests are a danger to health, scientists argue

Sarah Boseley
Health Correspondent
Trials of new medicines to establish that they are safe and effective are so badly flawed that they endanger the health of the patients who take part and those who will take the drug once it has been licensed, according to scientists who have been auditing them in confidence for 10 years.

Wendy Bohaychuk and Graham Ball are appalled at the poor standards they have consistently found. "Frankly, after 10 years of detailed auditing, I would never go into a clinical study myself and I would certainly try to discourage anyone in my family from doing so," Dr Bohaychuk writes in an editorial for the industry journal Clinical Research Focus.

Problems she and Dr Ball have found include medicines being stored at the wrong temperatures, patients not being properly informed of what will happen in the trial, and the side-effects of new drugs not being reported.

In a sample of 226 trials, they found that 43% of patients were not given clear instructions for using the medicine. In 55%, there was inadequate proof that the drugs had been properly stored: one patient complained that a gel ran down his leg after it had been kept through a hot summer in the doctor's attic. In a prostate cancer drug study, the GP was not given proper instructions from the drug company on its use on the grounds that it was already on the market. It was - for treating high blood pressure. In a third of the trials, they found "significant under-reporting" of side-effects.

On the basis of these trials medicines are given a licence, the seal of approval for general use. But if they have been wrongly conducted, says Dr Ball, we cannot be sure the drug is safe. "If the data are not adequate, there is a risk to the patient who will be exposed to the drug in the future."

The scientists' company, Good Clinical Research Practices, is called in by pharmaceutical companies to establish whether trials meet international standards. Drug companies are obliged to have an audit, but not obliged to disclose the result to anyone. The audit certificate that goes to the licensing authority - the medicines control agency (MCA) in this country - states only that the audit has been carried out.

Because of the secrecy surrounding clinical trials information, the MCA and the government are in no position to know about the failings, argue the auditors. "The government has no idea what is going on," said Dr Ball.

The doctors are concerned that patient safety is not at the heart of these trials. "I think patients probably don't know what is happening to them in the vast majority of cases. They are not adequately informed. If they are informed, it is a biased information process," he said. "They are not aware that a clinical study is putting them at risk."

The MCA admits that there is a problem. "The whole business of clinical trials and good practice is down to a voluntary code," said a spokeswoman. It has been invited by drug companies to do only about 30 voluntary inspections since 1997. But the EU is considering a draft directive, which the industry is fighting, that would make inspections mandatory. If that comes into force the situation could change radically.

The Association of the British Pharmaceutical Industry said it would expect members to adhere to guidelines on trials. It denied that there were widespread failings.



Trial and error puts patients at risk

Drug test Auditors claim that rules to minimise danger to human guinea pigs are neglected

Sarah Boseley
Health Correspondent
There are about 3,000 clinical trials in the UK at any one time, involving hundreds of thousands of patients. But every one is a human experiment, with risks attached.

There are rules to minimise the dangers to the human guinea pigs but two auditors say patient safety is being neglected.

Those who take part in the first phase of clinical trials are usually healthy volunteers, often students, who may be paid. The drug will have been tested in animals.

Phases two and three trials are on patients with the diseases the drug is supposed to treat, first a small group and then a large-scale, often multi-national trial involving thousands. Those trials are usually random - patients and their doctor are not supposed to know whether they are getting the new drug or a dummy. The trials have to be approved and overseen by ethics committees from the relevant hospital or health authority.

Drugs may have unpleasant or harmful side-effects but Wendy Bohaychuk and Graham Ball, whose company, Good Clinical Research Practices Ltd, has been auditing these trials for 10 years, say patients are not properly informed.

Andrew Herxheimer of the Cochrane Collaboration which trawls clinical trials for evidence of what treatments work, said: "The implications for the systemised review of randomised trials are horrible because it means we cannot really trust them as much as we thought. We need far more detailed accounts of exactly how trials are run."

Bohaychuk and Ball have put together a database from over 800 audits, mainly from the UK.

They use checklists to ensure that patients had been asked if they wanted to join a trial and told what to expect. They look for signatures and dates on consent forms and check that the trial documentation specifying what drugs the patient is on and at what doseage tallies with the patient's notes - often it does not.

They divide compliance into grade I non-compliance - serious breaches, which could create a dangerous situation for patients or allow an unsafe product to get to market - and grade II - less serious but still potentially unsafe.

Out of 226 sites (GP surgeries or hospital departments), they reported in the journal Applied Clinical Trials that: "88% had at least one event of grade I non-compliance, with an average of 2.3 events per site. All of the sites had at least one event of grade II non-compliance, with an average of 9.5 events per site."

At 31% of the surgeries or departments, they found "significant discrepancies" between the patient notes and the forms for the trial which should both have detailed the medication and tests the patient was receiving. A further 31% had either no reports or inadequate reports of serious side-effects of the new drug in 10% to 20% of patients. At 55% of the sites, the medication was either stored badly or there was no evidence that it had been properly stored.

In 81% of these studies, ethics committees were not told about all the serious side-effects and in 31% they say "there was significant under-reporting of safety information". In 37%, patients were not asked to sign the consent forms until after the study had begun.

Dr Ball says he believed the problem was principally one of doctors and drug companies cutting corners in their determination to prove that a new medicine worked. "The pharmaceutical company is making millions and the clinician is being paid thousands of pounds. It is not wilful disobedience - it is just that people get biased in terms of making money."

What starts as cutting corners could end up as fraud. Frank Wells and Peter Jay of MedicoLegal Investigations have successfully taken 18 cases to the General Medical Council, the doctors' disciplinary body, and have 12 more in the pipeline. Often the doctors involved were quite good researchers, they say, but they had not got enough suitable patients or enough time and so they made up data, assuming that they knew what the outcome would be.

Mr Jay said: "Most patients will do exactly what the doctor says. The vast majority of GPs are beyond reproach but every now and then we hit upon one who is an outright rogue.

"I think they possibly start by cutting corners. They are expected to recruit 20 patients and can only find 18 so they tell a patient the ECG hasn't gone properly and have to run it again so they have got two traces which they can use. It is very easy to fabricate it. Next time they recruit 10 and invent 10. Next time they realise all they have to do is fill in the forms."

The United States is the only country which sends in its own inspectors from the food and drugs administration (FDA). Drug licence applications will be thrown out if they do not come up to standard.

There is a lot of subtle pressure on patients

The fraudster
James Alfred Bochsler was a GP in Gipsy Hill, south London.

Twice in 1994 he signed deals to enter patients into trials for angina and depression. His fees totalled 22,500.

He claimed to have enrolled 36 of his patients in the two trials, but 25 of the consent forms were forgeries, signed by Dr Bochsler himself.

One of the trials, for Solvay Healthcare, was for an anti-depressant. One of the patients entered with a faked signature had twice been admitted to hospital as a psychiatric patient while she was supposed to be on the new medication under Dr Bochsler's supervision.

A second patient had been on other medication for lack of appetite and constipation. None of those details was recorded on the forms that went back to the company.

The second trial, for Bayer, involved a drug for hypertension (high blood pressure) and angina. Dr Bochsler had been asked to take ECG readings for the patients involved. For five pairs of patients, he submitted identical printouts.

Dr Bochsler was struck off last November. "Scientific dishonesty producing bogus results undermines the integrity of drug trials," said John Ball, chairman of the professional conduct committee.

The investigator
Peter Jay of MedicoLegal Investigations

"We hear the most bizarre stories. There was a husband and wife in the same depression study, which was a bit unusual in itself. The wife clearly had a problem when I saw her. She agreed that she had signed the consent form and taken the medication and her husband had too. I said to him: 'What's your history of depression?' He said: 'I have never been depressed in my life. Dr X told me that it would be in my wife's best interests if I went into the study as well.' It was a placebo controlled trial. The wife said the medication was bloody useless. The husband said that for four months he had never felt so ill."

The counsellor
Heather Goodare, counsellor to breast cancer sufferers and chairwoman of Breast-cancer Research Ethics and Advocacy Strategy

"Patients are very vulnerable when they are being treated for cancer. When you have just been told the bad news you are not in a fit state to make sensible decisions. I think it is not a good idea to try to recruit people into clinical trials at that point. There is an awful lot of subtle pressure and people don't want to offend the doctor.

"I know of a patient who has spent five years on one cancer drug and would be pretty grateful to stop it, but was asked to enter a study into the longer-term effects. She thinks it is almost her patriotic duty to enter the trial."



Daylight on drugs

Americans are better protected. Why?

The control of medicines in Britain has always been shrouded in secrecy. Even when the last government produced its patient's charter, which guaranteed every patient the right "to be given a clear explanation of any treatment proposed, including any risks and any alternatives, before they decide whether they will agree to treatment", the secrecy of the regulatory process remained uninterrupted. The 1968 Medicines Act, which makes it a criminal offence for officials involved in licensing drugs to disclose information about their decisions, remained in force. In the United States, patients have a right to know why a new medicine has been licensed, what type of clinical studies were undertaken, and what detailed information exists on side effects. Here, all information supplied to the Medicines Control Agency (MCA) remains strictly confidential. Social reformers have long pressed for the MCA to follow the open approach pursued by America's food and drug administration. Now we know something more fundamental is needed. Even the MCA is not being provided with important facts.

Our health correspondent reports today on the disillusionment of two of the senior auditors whom drug companies are obliged to engage

to monitor the trials that all new medicines are required to undergo to test their safety and effectiveness. The MCA gets an auditor's certificate. It receives up to 600 volumes of material for just one product. But what it does not get are specific details on the conduct of the trials. The auditors report that 43% of patients are not given clear instructions in using medicines. In a third of the trials they found "significant under-reporting" of side effects. The auditors want much more systematic inspection of the trial process.

A voluntary inspection system already exists, under which drug companies can invite the MCA to send in inspectors. But insiders suggest the invitations are only issued when they are least needed. In America, a cadre of FDA inspectors apply a much more systematic approach. The European Union is considering a draft directive, which would make inspections mandatory. Our powerful medicines industry is opposed, but this is one issue on which Brussels must be backed. There are two ways of improving drug safety and effectiveness: a more open process (sunshine is the best disinfectant) and more systematic inspections.


The Guardian, 30 July 1999
Letter from Dr Richard Tiner, Director of Medicine, ABPI
It is not true to suggest that people taking apart in UK clinical trials are at risk (Drug trials risk to patients, July 27). Stringent safeguards are in place throughout the different phases of clinical trials to ensure that any unwarranted side-effects of a new medicine are immediately reported. If necessary, the trial will be stopped. Your article refers to research carried out over a 10year period, but clearly only involving a very small percentage of clinical trials - probably fewer than 1%. Even then, the authors indicate that the research also covered trials done outside the UK.

Much has changed in clinical practice within that 10-year period. Independent ethics committees have been appointed to ensure that clinical trials are run to the most exacting standards.


It is suggested that pharmaceutical companies and investigators may cut corners. Quite apart from ethical considerations, no company can afford to do that. It takes 10-12 years and some 350m to research and develop a new medicine. This investment would be thrown away if the regulatory authorities had' cause to believe that guidelines had not been followed.

Your leader (July 27) suggested more regulation. There is no real need for further regulation -, indeed, the pharmaceutical industry is already one of the UK's most regulated industries. Contrary to the impression given by your article, the Association of the British Pharmaceutical Industry (ABPI) has no problem with the development of a European directive on good clinical practice.


Letter to The Guardian (unpublished), 31 July 1999, from Wendy Bohaychuk PhD
Director, GCRP Consultants; Editor-in-Chief, Quality Assurance Journal
Dr Tiner (Director of Medicine, Association of the British Pharmaceutical Industry), in response to an article describing some of our clinical trial audit findings, wrote that stringent safeguards are in place throughout clinical trials to protect study subjects. According to our data, they are not stringent enough.

The industry has always complained about too much regulation, but given that our health is at risk, surely the whole situation deserves tight control. Anyway, who is checking that these "safeguards" are effective - certainly not the ABPI. Dr Tiner criticised the "small" number of studies we have inspected (over 800, mostly in the UK, and more than any other organisation we know of, apart from the US Food and Drug Administration which makes its findings publicly available through Freedom of Information rules). It is indeed a small percentage of the total number of clinical studies being

conducted, which makes our findings even more worrisome as we are convinced that we are invited to audit the "better" studies. As far as we know, the ABPI has inspected no studies, so we wonder on what grounds Dr Tiner makes his statements assuring us that all is well.

As to Dr Tiner's statement that much has changed in the last 10 years - perhaps implying that our findings may not reflect current practice (although we have audits conducted up to last year on our database, with no suggestion that things have improved) - we look forward to seeing some hard data to support that statement. (In fact, the ABPI and the Royal College of Physicians published some good standards 10 years ago, but self-regulation did not work then as it does not work. now.) It is true that many more new rules have emerged in the last decade to try to improve the situation (as the industry and the public apparently recognised

there was a problem) but again, who is checking that these rules are working? Certainly not the overworked ethics committees in this country which review lots of paper but do not have the time and resources to visit study sites and confirm that the studies they approve are, running properly and meeting "exacting standards". And certainly not our government which is still waiting for legislation (why?) to conduct mandatory inspections.

Ten years ago, a senior ABPI spokesperson reported to a European-wide audience at a meeting in France that all clinical trials in the UK were safe and waved about an ABPI document of Good Clinical Research Practice guidelines (in fact, a very good document) to apparently establish this point. Our reaction then, as it remains today, is "how do you know"? The ABPI, the ethics committees and the government were not conducting inspections at that time either.


Letter to The Guardian, 3 August 1999
from Charles Medawar, Social Audit, Ltd
Given the strength of the evidence in the Guardian report (27 July) of risks to patients in clinical trials, it is extraordinary to see the Association of the British Pharmaceutical Industry deny (letters, 30 July) that any problem exists.

If there really weren't much of a problem, presumably the ABPI would have no objection to making audit reports publicly available - or at least available for inspection by ethics comm- ittees and/or the Medicines' Control Agency? As things stand, the authorities have no access to such documentation, and therefore no evidence of the many problems that emerged in the 800 audits reported by Drs. Bohaychuk & Ball.

The only information now available to the authorities is a certificate indicating that a trial has been audited - a one-page document that omits all qualifications and gives no indication of the standards actually achieved. If member companies of the ABPI really want to reduce the amount of external regulation, they should publish the evidence they have. In the meantime, patients and doctors should beware.


Contents page
What's New?