Editor's Choice

Trial results: the next battle

BMJ 2005; 330 doi: https://doi.org/10.1136/bmj.330.7483.0-g (Published 13 January 2005) Cite this as: BMJ 2005;330:0-g
  1. Kamran Abbasi (kabbasi{at}bmj.com), acting editor

    Drug companies last week announced a plan to make results from clinical trials of new drugs publicly available (p 109). Summary results of completed, industry sponsored trials will be disclosed regardless of outcome, say trade associations of the world's drug companies. On the face of it, this move is one to be cautiously welcomed.

    Cynically, you might argue that recent alarms about the safety of highly prescribed drugs and subsequent plunges in share price for AstraZeneca, Pfizer, and Merck—described by the Financial Times as “a painful six months for the pharmaceuticals industry”—may have inspired this display of virtue. Political and public pressure, and the threat of future legislation, for full disclosure of clinical trial results, particularly in the United States, is another driver of change. An initiative by leading medical journals, including the BMJ, to consider for publication only trials recorded in public registries is a small but important nudge towards transparency (BMJ 2004;329: 637-8).

    A new drug industry initiative focusing on trial results may leave you with the impression that the issue of trial registration has been resolved. But it hasn't. For several years, some companies have registered ongoing trials but this system has failed: it is voluntary and few trials are registered. Medical journals will begin implementing their registration policy in July, although it is unclear how successful it will be. There are thousands of other journals for drug companies to choose.

    The next battleground was always going to be full disclosure of trial results, but it is a battle begun before the first one—around trial registration—has run its course. None the less, the drug company initiative is a step in a direction that should meet with approval from journals, politicians, and the public. But how might the proposal be improved?

    The sting in the tail is, once again, the voluntary nature of the proposal—unless it is mandatory it will not work or be trusted. Another issue is how effectively the system will be enforced and who will monitor it? For example, if a drug company says it is participating, how can we be sure it is playing by the rules? Finally, the deal is that companies can present data on their own websites. How will this information be harmonised or how effective will be the mechanism to link these websites? Presenting information on all sorts of different websites in all sorts of different ways will be problematic.

    As companies dither over whether or not to join this venture and whether or not to cloak their commercial interests, they should remember that it is participants in those trials, who presumably entered them for personal gain and the public good, who are as much owners of that “proprietary” information as the companies. It may take several years to judge this initiative's success, by which time legislation may have made it mandatory.

    To receive Editor's choice by email each week subscribe via our website: bmj.com/cgi/customalert

    View Abstract