News and Events > A New Phase in Clinical Trials Reporting

A New Phase in Clinical Trials Reporting

September 10, 2004

Physicians and patients have long criticized pharmaceuticals companies' tendency to hide negative drug trial results, but until recently, the critiques lacked force. Actions for change have coalesced since June, when the New York Attorney General filed suit against GSK alleging deliberate suppression of negative results from clinical trials of anti-depressant Paxil prescribed to children and adolescents.

The AMA adopted a policy in June to support a national clinical trials registry. Addressing the public's concern about commercial influences in drug trials, the AMA warns that industry-funded studies may be more likely to use placebos or inactive controls, increasing the chances of achieving positive results, and that furthermore, positive findings are more likely to see publication than are negative or null results.

GSK settled in late August, agreeing to post online summaries of all its clinical trials completed after December 2000, the date of merger between Glaxo Wellcome and SmithKline Becham. GSK's clinical trials register purports to disclose whether the goals or elements of the study were changed mid-stream, and when studies are called off, the reason for early termination. GSK posted its first release of clinical trials data on September 1. (http://ctr.gsk.co.uk/welcome.asp). Also as a result of inquiry into clinical study disclosure practices, Forest Labs agreed in September to initiate a public registry of its marketed drugs.

The pharmaceutical industry is advancing its own initiatives, while Congress has begun hearings and medical journal publishers have joined to force fuller disclosure. Current industry initiatives include The Pharmaceutical Research and Manufacturers of America (PhRMA's) October 2004 date for inaugurating a public database containing the results of all controlled clinical trials (mainly Phase III and IV studies), both positive and negative, completed since October 2002 for PhRMA-member company drug products approved in the United States. The scope of industry's proposal, limited to summaries of completed trials of approved products, arguably responds to prescribing physician and patient information needs, but skeptics argue it does little to advance researchers' access to comprehensive, unskewed data. PhRMA's voluntary registry falls short of GSK's and Forests' registries, failing to meet the AMA's recommendation to include revised and halted trials and their reasons for termination.

On September 9, eleven top medical publishers announced that beginning in mid-2005 they will not publish results of studies unless they've been registered in advance in an independent, public database such as the NIH's www.clinicaltrials.gov. As The Economist points out (9/9/04), "The editors' initiative will help, but it will serve only to flush out now-hidden trials so that questions can be asked about what happened to them if no public report of their results ensues" and will not force answers to those questions. Congress will consider deeper reforms, including adoption of the AMA's recommendation to ensure public registry compliance by making it a condition of IRB approval for human trials. Federal law and FDA rules currently require registration of certain kinds of trials, but lacks enforcement power. According to the AMA, industry dollars made up more than 70% of clinical trials funding in 2002.

The medical journal editors' stance followed the NIH's September 3 announcement that it proposes to improve researchers' access to federally-funded medical studies. The agency proposes to establish a comprehensive, searchable electronic database for posting the results of all NIH-funded research. The proposal, which NIH has been considering for more than a year, would require federally funded researchers to submit their papers to the agency after they have been accepted for publication and edited by a journal. Six months after an NIH-supported research study's publication by a commercial publisher, the manuscript will be freely available to the public through the NIH's digital documents repository. (http://grants.nih.gov/grants/guide/notice-files/NOT-OD-04-064.html)