Government regulation is needed to prevent biased under-reporting of clinical trials

Editor—In 1996, Schering Healthcare published details of its ongoing randomised clinical trials in the Cochrane Library. The chief executive told me that he was doing this because industry's failure to disclose the results of its phase 3 trials could not be defended ethically or scientifically. Two years later, the chief executive of GlaxoWellcome announced his company's decision to register and seek to report all its randomised clinical trials.¹ A few years after that, the Association of the British Pharmaceutical Industry commended GlaxoWellcome's policies to its other member companies.

After GlaxoWellcome had become part of GlaxoSmithKline (GSK), I wrote to the chief executive of the new company, urging him to support the efforts of those within industry who were attempting to promulgate guidelines for good publication practice (www.gpp-guidelines.org). I received no acknowledgement, and, soon after, his company sacked one of the leaders of the initiative and closed the department she headed.

In response to accusations of biased under-reporting of research, GSK has now announced that it intends to institute policies announced seven years ago by GlaxoWellcome.² It would be churlish not to welcome this. But the past record of the pharmaceutical industry, and the reactions of some other companies to GSK's announcement, prompt deep scepticism that the industry will ever voluntarily implement ethical trial registration and publication policies.

Biased under-reporting of clinical trials kills patients and wastes money, and government regulation is needed to put a stop to it.³