Abstract
What is the value of fairness in medicine? Do unto others as you would have done unto you? Is it possible to say fairer than that? Fair tests of treatments presume a genuine uncertainty about which treatment is more beneficial or harmful. Yet too many clinical trials use inappropriate comparators to exaggerate the benefits of a new therapy. Fairness goes beyond transparency and disclosure of trial data. The results of many unfair trials are published, the data shared. Unfairness, in this instance, lies in selecting a comparator drug or dose of a drug that, for example, is known to be less effective or has a worse side-effect profile. In 56 out of 63 trials involving patients with highly active rheumatoid arthritis, potentially helpful treatments were withheld from 9224 out of 13,095 patients randomized to the control arms. Why? Because placebos or treatments known to be ineffective were used as controls. 1
Fairness aside, choosing appropriate comparators requires effort. Ideally a systematic review will be conducted to establish the comparator of choice and what dose to use. But such effort should be routine before submitting a grant application or seeking ethical approval for a clinical trial. Many trialists are unaware of the systematic review relevant to their studies, and so expose patients to suboptimal or potentially harmful treatments in the control arm. Admittedly, selecting the appropriate comparator is not an exact science, subjective opinions can hold sway - and the danger is that these are dominated by commerce, marketing and personal ambition.
Storm clouds are gathering for sponsors and investigators of clinical trials who withhold data and results, and who manipulate trials.2,3 Doctor and enemy of bad science, Ben Goldacre, has seen his book on how drug companies mislead doctors and harm patients, Bad Pharma, rush to the top of Amazon's bestseller list. Roche and its medical advisers have been publicly questioned over the company's unwillingness to disclose the vast majority of trial data about oseltamivir (Tamiflu), a drug that cost the NHS billions of pounds. 4 Members of parliament are beginning to listen and understand.
Publication bias and concealing unfavourable trial data are ancient undercurrents that are rising like a modern tidal wave. The dam of complicity that protects profits, bolsters careers and keeps patients from the best treatments is about to burst. 5 Why has it taken so long? Pharmaceutical company executives and clinical investigators who work with industry must ask themselves why many of them, including the most senior, have seen no evil? There is a grand deceit in medical research, and it is that clinical trials are fair by definition. It is a deceit that is camouflaged by soothing words about patient care and perpetuated by insiders who are busy buying in to the success of their latest product. There is no equipoise here, there is no doubt. There is only a drum beat for a return on investment, a desire to show a new therapy in its most seductive light. Clinicians have a moral duty to become more involved with the design of clinical trials, selection of comparators and ensuring trial data are fully published. Otherwise, seeing no evil leaves us with blood on our hands.