Registration and reporting of clinical trials affiliated with California universities and with primary completion date from 2014 to 2017.

BACKGROUND: Public information on US clinical trials is shared through the ClinicalTrials.gov registry. This study's goal was to determine prospective registration, results reporting, trial registration number reporting, and publication accessibility status for trials with primary completion dates from 2014 to 2017 affiliated with seven California universities. METHODS: We identified trials with investigators, sponsors, or responsible parties affiliated with seven California universities and searched for their results publications manually. We then used semi-automatic methods to determine prospective registration, summary results reporting in a registry, publication status, and reporting of registration numbers in the abstract and the full text of manuscripts. RESULTS: We identified 1091 unique trials. Most trials were single-center (n = 752, 69%) and had a median of 50 participants (IQR 21 to 150). Overall, 64% of trials (n = 698) were prospectively registered, 46% (n = 500) had summary results reported in the registry, 69% (n = 750) had results published as articles, and an additional 3% (n = 36) as abstracts or posters. Results reporting (summary, articles, abstracts or posters) occurred for 58% of trials (n = 637) within 2 years, and for 81% (n = 888) within 5 years of the study's primary completion date. Of journal publications, 77% (n = 579) were open access publications, 37% (n = 276) had trial registration numbers listed in their abstract, and 45% (n = 336) in the manuscript. Only 92 (8%) of these trials were legally required to report results, and only 2 (2%) of those were overdue and under the primary responsibility of a California university to report. CONCLUSIONS: Almost one fifth of clinical trials with primary completion dates from 2014 to 2017 with investigators, sponsors or responsible parties affiliated with seven California biomedical research universities lacked any results reporting 5 years after their primary completion, and only 58% reported results within 2 years. Even though a large majority of these trials were completed before US legal mandates for reporting, there was an ethical requirement that the burden to research participants should be commensurate with the scientific value of the research. Research has no public scientific value if its results are not reported.