Erick H. Turner,1,2 An-Wen Chan,3 Dan A. Oren,4 Steven Bedrick1
Many clinical trials funded by the National Institutes of Health (NIH) are not published in a timely fashion. We sought to determine how many NIH-funded R01 clinical trials go unregistered.
We used the NIH RePORT database to identify R01 grants awarded from 2009 to 2013 for clinical trials by searching the (1) abstract field for “random%;” (2) the project terms field for “Randomized controlled trial(s),” “Randomized clinical trial(s) (RCTs),” “Randomized controlled clinical trial(s),” or “Randomized placebo controlled trial(s);” and (3) the NIH Spending Category field for “Clinical Trials.” We manually examined the context around term (1) and included only grants that clearly proposed new clinical trials; by the time of presentation, each full abstract will be rated by a second individual and rating discrepancies will be resolved by reaching a consensus. Using RePORT’s Clinical Studies tab, which links grants to ClinicalTrials.gov registrations, we identified trial grants that were linked to 1 or more registered trials. For grants lacking such links, we searched ClinicalTrials.gov by principal investigator last name and manually compared any matches with the index grant; we analyzed data for a random sample of 12 such unregistered grants but will report data for all “unregistered” clinical trial grants at the time of presentation.
Among 601 grants identified by search results, we excluded 116 after manual verification leaving 485 R01-funded clinical trials. The 485 grants were associated with 357 trial registrations. Excluding 24 redundant trials (those associated with the same grant) left 333 grants linked to 1 or more registered trials, suggesting a registration rate of 68.7%. However, among the random sample of 12 “unregistered” clinical trials, we were able to manually identify ClinicalTrials.gov registrations for 7.
The proportion of NIH-R01-funded clinical trials registered at ClinicalTrials.gov is at least 68.7% and is likely higher. Reasons for nonregistration could include failure to conduct the trial after funding, which we could not measure. The RePORT database is a resource that can be used by systematic reviewers seeking comprehensive inception cohorts of NIH-funded clinical trials.
1Oregon Health & Science University, Portland, OR, USA, firstname.lastname@example.org; 2Scientific Resource Center, Agency for Health Research Quality, Rockville, MD, USA; 3Women”s College Research Institute, Toronto, Ontario, Canada; 4Yale University, New Haven, CT, USA
Conflict of Interest Disclosures:
Dr Turner receives protected time for research from the Scientific Resource Center, which is supported by the Agency for Health Research Quality. No other disclosures were reported.
The authors thank Stuart Buck and Michael Stebbins, of the Arnold Foundation for discussions of the general idea of the study and Kevin Fain, Thiyagu Rajakannan, and Nicholas Ide, of ClinicalTrials.gov for discussions regarding link between ClinicalTrials.gov and RePORT.