The ClincalTrials.gov Results Database—Update and Issues

In use for more than a decade, the Clinical Trials.gov registry was expanded in 2008 to include a database for reporting results. The Food and Drug Administration Amendments Act (FDAAA) mandates that summary results data for certain trials of drugs, biologics, and devices be submitted to ClinicalTrials.gov whether published or not. Imposing significant penalties for noncompliance, the expanded legal requirements address concerns regarding scientific and ethical issues affecting the design, conduct, and reporting of clinical trials. The database, providing publicly available summary trial data independent of the sponsor, enhances systematic public disclosure of clinical trial information. A recent article in the New England Journal of Medicine [2011;364(9):852-860] examines the contents and structure of the ClinicalTrials.gov results database and reports on updated policies affecting the state of clinical research. The researchers analyzed publicly available data collected between September 2009 and September 2010. Using a Web-based system, trial sponsors or investigators report data, usually at trial inception, including the study’s purpose, design, participant eligibility criteria, locations, and other key protocol details. A quality review of submitted information is conducted prior to public posting, and records are corrected and updated throughout the trial period. After the trial is completed, results are entered in the results database, which is designed to complement but not replace journal publication of the findings. ClinicalTrials.gov received approximately 2000 revised registrations and 330 new registrations per week as of September 27, 2010. In addition, 80 revised results and 30 new results were submitted weekly. The researchers characterized 79,413 registry and 2178 results records. Using a random sample cohort of 150 results records, 78 (52%) had associated publications within 2 years of posting. Of the 2178 clinical trials with publicly available results records, 20% had >2 reported primary outcome measures and 5% had >5. Using ClinicalTrials.gov data, the researchers examined outcome measure and analysis population, 2 data fields considered integral in interpreting study results. At registration, trial data providers are instructed to report specific measure and time frame for each primary and secondary outcome. Primary outcome measures from 100 randomly selected clinical trials were reviewed and assessed to determine whether a specific time frame was provided during the initial registration. Of these trials, 61% lacked specificity in describing the metric used; “post hoc choices of the specific metric or method of aggregation could mask the fact that multiple comparisons were conducted, potentially invalidating the reported statistical analyses and allowing for cherry-picking of results.” In examination of a sample of 700 results (representing 1749 study groups and 5160 outcome measures), the mean number of different analysis populations per study group was 2.5 (median, 1; range, 1-25). By evaluating the percentage of participants who started the study and were analyzed for the first primary outcome, researchers further explored the magnitude of these differences. In a sample of 684 eligible studies, 31% included 100% of participants in the analysis and 24% of trials reported results for ≤90% of their participants. Limitations in determining the precise state of clinical trials include those not registered with ClincalTrials.gov or any publicly available registry, records containing missing elements or imprecise entries, and a lack of coordination of new registries and policies internationally. The FDAAA is seeking to expand the basic results database through rule-making to assure complete results information. The European Medicines Agency plans to make summary protocol information and results publicly available for clinical trials of both approved and unapproved drugs. Compatibility issues of the European Union database with ClinicalTrials.gov are being addressed. In conclusion, the researchers noted that while many trial sponsors and investigators are successfully submitting summary results, others are not. The poor quality of some data entries is of concern, and the usefulness of ClincalTrials.gov ultimately “depends on the research community to submit accurate, informative data.”