Although prospective registration of clinical trials is becoming an international standard, the debate continues over public access to clinical trial results, especially investigational trials conducted by pharmaceutical companies.

Recent legislation, which has been signed into law by President Bush, reflects increasing pressure to disclose clinical trial information to the public. A solution to the disclosure problem must balance conflicting interests, including commercial interests, the public's right to be informed, unbiased publication of medical research, copyright and proprietary information, and therapeutic guidelines.

In its 2004 editorial, the International Committee of Medical Journal Editors (ICMJE) admitted that the registration of study protocols "is only part of the means to an end; that end is full transparency with respect to performance and reporting of clinical trials." This end remains elusive today. There is no open, searchable, central repository for results of most clinical studies, like ClinicalTrials.gov has done for trial registration. Simply knowing the existence of certain trials does not entirely prevent underreported negative and inconclusive study results. The editors believed that study sponsors and researchers should openly and honestly report clinical trial results—an ethical obligation they owe to trial participants' "altruism and trust that make clinical research possible."

Some researchers routinely post PubMed links to journal articles containing results of their registered trials on ClinicalTrials.gov, although it is unclear whether the National Library of Medicine, which runs PubMed, has a mechanism to ensure prompt, consistent reporting and updates of these publications. Results not published in biomedical journals would not be posted on the Web site Publication in a peer-reviewed journal can be a lengthy process, causing considerable delays in posting the citation online. As ICMJE pointed out, negative or inconclusive study results are less likely to be published than positive results. A study of published psychiatric clinical trials linked authors' conflicts of interest with a greater likelihood that study results were favorable to the intervention studied (Psychiatric News, October 21, 2005).

In a 2005 joint policy statement, the European Federation of Pharmaceutical Industries and Associations, International Federation of Pharmaceutical Manufacturers and Associations (IFPMA), Japanese Pharmaceutical Manufactures Association, and Pharmaceutical Research and Manufacturers of America (PhRMA) recommended voluntary disclosure of "the results of all clinical trials, other than exploratory trials, conducted on a drug that is approved for marketing and is commercially available in at least one country" in a free, publicly accessible clinical trial results database.

For investigational products that are not successfully developed and approved by regulatory agencies, they encourage sponsors to post study results if possible and if the results have "significant medical importance and may have an impact on...labeling." PhRMA advises members to post results of "studies conducted after a drug's approval," which may include postmarketing surveillance studies or phase 2 or phase 3 clinical trials for unapproved indications.

PhRMA has sponsored a study-results database, ClinicalStudyResults.org, since 2006. It appears to be the most comprehensive results repository available and contains clinical trial results for drugs that are already marketed in the United States. Study results are contributed by PhRMA members, which include major U.S. pharmaceutical companies but only a few biopharmaceutical companies.

Eli Lilly and Co. has adopted a broader policy of disclosing results of all trials, regardless of the phase, that the company has conducted in support of a product's initial registration "no later than when the first indication is approved and the drug is commercially available for patient use anywhere in the world." PhRMA's policy remains silent on the retrospective disclosure of preapproval clinical trial results.

It is unknown how many PhRMA member companies routinely and consistently follow the organization's guidelines and post trial results on the Web site. Reporting on ClinicalStudyResults.org remains voluntary and is not monitored by an outside group. The Web site states that "PhRMA hopes that database administration will eventually be transferred to an independent third party." By the end of April, the Web site contained clinical study results for 351 prescription drugs.

The World Health Organization Registry Platform has established the Study Group on the Reporting of Findings of Clinical Trials to provide advice to it on the reporting of clinical trial findings. A lack of consensus on international standards is cited as one hurdle to be overcome before an effective mechanism is ready to be adopted worldwide.

The ICMJE, for example, supports reporting study results "in the form of a brief (< 500 words) structured abstract or table" before they are published in a biomedical journal, due to concerns about first-publication rights. Extensive revelation of study results could be considered a form of publication and, at least theoretically, could jeopardize its eligibility to be published as primary research. Alternatively, the ICMJE suggests that study results be posted in a public registry "24 months after closure of data collection if results are not published in a peer-reviewed venue by that time."

Public Citizen, a consumer advocacy group, emphasizes the need for summaries of study results using nontechnical language in a results database." Clinical trial results should be given in a way that's palatable to the lay public. A lay person can go to PubMed and search for journal articles, but the technical language is more likely to confuse and mislead them," Peter Lurie, M.D., deputy director of the organization's Health Research Group, told Psychiatric News.

PhRMA takes a different position on the issue. Ken Johnson, senior vice president of PhRMA, stated in a September 17 press release that "it is crucial that during the evaluation of clinical trial information, researchers acknowledge that the analytical methods used—such as meta-analysis—have limitations as well as strengths. And it is important not to reach hasty conclusions until all studies and data have been evaluated."

Some researchers are concerned about misinterpretation or disproportional reaction to study conclusions. Clinical studies are often complex and yield ambiguous findings. In a commentary in the May 16 Journal of the American Medical Association, Deborah Zarin, M.D., director of ClinicalTrials.gov, and coauthors pointed out that attempting to convey complex study results using simple summaries could pose risks for misinterpretation or unintended consequences. The accuracy of non-peer-reviewed results and the reliability of data, if directly provided by sponsors with vested interests, are of particular concern. Without peer review and editorial commentaries, studies with poor design and questionable conclusions may be disseminated, and the public may not be able to distinguish these from peer-reviewed reports with higher quality.

The Food and Drug Administration (FDA) Amendments Act (HR 3580), which passed the House of Representatives and Senate in late September and has been signed into law, mandates prospective registration for clinical trials of all drugs, biologics, and medical devices (see Increased Industry Fees to Help FDA Better Monitor Drug Safety). Registration in a searchable public database, which will likely be an expanded version of ClinicalTrials.gov, will be required for all clinical trials except those in phase 1. "Basic results" from clinical trials of approved products will be required to be submitted to the database as well. The secretary of Health and Human Services will make additional rules as to whether the database will include results from trials of unapproved products and the datasets that must be submitted. The FDA will have the authority to issue civil monetary penalties in enforcing the law.

This legislation is expected to hasten the standardization and centralization of clinical trial disclosure and bring far-reaching changes to the communication among clinical research, pharmaceutical industry, biomedical journal publishing, medical practitioners, and the general public.

Following is a list of current clinical trial results databases:

PhRMA's joint policy paper is posted at<www.phrma.org/files/2005-01-06.1113.PDF>; the Federation of American Scientists' report for Congress is posted at<www.fas.org/sgp/crs/misc/RL32832.pdf>; the FDA Amendment Act is posted at<thomas.loc.gov/cgi-bin/bdquery/z?d110:h.r.03580:>.▪