UK litigation over side effects of the third generation contraceptive pills reaches a crunch point this month, when the High Court judge overseeing the case must decide whether or not the risk of venous thromboembolism from the newer pills is more than twice as high as for the second generation of oral contraceptives.
Mr Justice Mackay will rule on the issue at the end of May, after submissions from both sides, based on expert evidence, finished this week. If the judge decides the relative risk is ”2.0, the compensation claim by more than 100 women against three manufacturers will be halted in its tracks. The claim will be thrown out because the women will have failed to prove that their venous thromboembolism was more likely than not to have been caused by taking a third generation pill.
The women launched their action under the Consumer Protection Act 1987, claiming that the third generation pills were defective products because there was no warning that they carried higher risks than their older counterparts.
The claims are against Schering Healthcare (manufacturer of Femodene), Wyeth (Minulet and Tri-Minulet), and Organon Laboratories (Marvelon and Mercilon). The new pills captured a large share of the UK market before the 1995 “pill scare,” when their use dropped dramatically after the Committee on Safety of Medicines, the UK regulator, issued a warning letter to doctors. The letter was in response to three studies, including a WHO study, that found that the new drugs carried a relative risk of 2.7 of venous thromboembolism, compared with second generation pills.
An analysis of seven studies since 1995 published in the BMJ last year found a relative risk of 1.7, but the analysis included industry funded studies showing a lower relative risk, which the claimants argue are unreliable.
At the heart of the case is whether the findings of a greater relative risk are accurate or, as the defendants argue, the result of bias, including prescriber and diagnostic bias, and other factors such as differences in the ages of intervention patients and controls.
Experts for the defendants argue that a small relative risk, such as 2.0, could be entirely due to bias and confounding. A striking feature of the case is the difference in the results of studies funded by the drug industry and studies not so funded, with the industry funded studies finding little difference in risk between the old and new pills.
Richard Farmer, professor of epidemiology at Surrey University, said when examined by the defendants' counsel, Nicholas Underhill QC, that early in his career he had been on the committee involved in setting up the first RCGP oral contraceptive study.
Mr Underhill said Dr Herschel Jick of the Boston collaborative drug surveillance programme had suggested that Professor Farmer did not have appropriate qualifications to conduct research in pharmacoepidemiology.
Professor Farmer said the suggestion was “a bit misleading.” There was no such discrete specialty when he did his training, and he did not think that even today there was a specialist qualification in pharmacoepidemiology. Essentially, it was applying epidemiological techniques to the investigation of the safety of medicines.
Mr Underhill said Dr Jick had condemned “in very strong language” the “pill scare” paper published by Professor Farmer in the BMJ, which looked at women who were prescribed oral contraceptives before and after the 1995 warning about third generation pills (BMJ 2000;321:477-9). It concluded there was no evidence of a twofold increase in risk from the newer pills.
Professor Farmer said the study had set out to see whether there was any evidence that the pill scare had achieved what it set out to achieve. It was reasonable to say: did the intervention work?
“If there was a major effect from the so called third generation oral contraceptives, then one would have expected it to be measurable.” He disagreed that such an analysis was controversial. “I would put it like many other pieces of epidemiology: view with caution.”
Under cross examination by Lord Brennan QC, for the claimants, Professor Farmer said he had not written on oral contraceptives and the disorder before his first paper in 1995, although he was then in his mid-50s. He had been a consultative adviser to Organon, Schering, and Wyeth on safety of medicines since the early 1990s.
Lord Brennan told Professor Farmer he was cross examining him “on the basis that your understanding of epidemiology is incomplete and your conclusions in your reports are unreliable.”
Quoting from reports to the court, he accused Professor Farmer of not understanding such a basic epidemiological term as “confounding.”
Professor Farmer admitted the words quoted were “badly drafted.” He felt the data in his 1995 study were fragile because of their sparseness.
Was it not “cavalier,” then, asked Robin Oppenheim, junior counsel for the claimants, to write a letter to the Lancet saying, “Our data do not support the view that oral contraceptives with gestodene and desogestrel as the progestogen [third generation pills] carry a significantly increased risk of thromboembolic disease”?
Professor Farmer denied it was cavalier, adding: “It does not support it. I did not say it refuted it.” He said he was asked to write the letter by Organon, Schering, and the Medicines Control Agency and that he did not disclose the funding of his study in the letter because it was not the practice of the Lancet at the time to disclose funding in letters.
His 1997 study posited that findings in previous studies of higher relative risks for third generation pills were “likely to have been the result of residual confounding by age.” He said the WHO study had used controls matched by age bands of five years, and the study by the Boston collaborative drug surveillance programme (another study not funded by the industry) used bands of two years. His study matched by year of birth.
Mr Oppenheim accused Professor Farmer of making an “extreme” case in suggesting that changing from a two year band to the year of birth would affect the relative risk.
Professor Klim McPherson, senior scientist in the department of social medicine, Bristol University, told the court: “To argue, as it is seeming to be argued here, that all relative risks of 2.0 must be a consequence of bias in case control studies seems to me to be without foundation.”
Figure.
Professor Richard Farmer, expert witness for the defendants