The coronavirus disease 2019 (COVID-19) pandemic caused an unprecedented drive to launch clinical trials.1 We assessed the recruitment and results reporting of randomized clinical trials (RCTs) to treat or prevent COVID-19 registered within 100 days of the first case reported to the World Health Organization.1
All RCTs registered in ClinicalTrials.gov and the World Health Organization International Clinical Trials Registry Platform between January 1 and April 9, 2020, were included. These trials had been previously identified in a systematic assessment.1 Between October 5 and 15, 2020, using contact information provided in the registries, we sent emails to all trial groups, requesting information on recruitment and enrollment status, reasons for trial termination or discontinuation, and any results reporting. We assessed whether trials recruited 75% or more of their target sample size, stopped before 75% recruitment, or were continuing to recruit (on schedule or delayed) but not yet 75% recruited. On October 19, we searched the Living Overview of Evidence platform2 for results posted as preprints, published in peer-reviewed journals, or reported in trial registries. The survey results were used to complement the information from our search. Associations of trial characteristics—location (ie, China vs elsewhere), planned sample size greater or less than the median (114 participants [interquartile range, 70-334 participants]), registered on or after the median registration date (March 15), and type of control group—with recruitment and results reporting were explored using logistic regressions. We used R version 3.6.2 (R Project for Statistical Computing). The P value threshold was <.05 in 2-tailed tests. Institutional review board approval and an informed consent waiver were not required for this study because the aggregated trial data collected are not subject to the health regulations of Switzerland.
There were 516 trials identified. Seventy-six trial teams (14.7%) responded to the survey, and we found results reported for 53 trials (10.3%). Overall, 56 trials (10.8%) were completed, 24 (4.6%) terminated early, 111 (21.5%) were ongoing on schedule, 126 (24.4%) were ongoing but delayed, 39 (7.6%) were ongoing without reporting an expected end date, 155 (30%) had not started or were discontinued, and 5 (1.0%) had unknown status.
Recruitment had started in 361 trials (70.0%). Of 108 started trials with detailed recruitment information obtained from investigators and/or reported results, 38 (35.1%) reached at least 75% of their target sample size, 34 (31.5%) stopped before 75%, and 36 (33.3%) had not yet recruited 75% of their target. Median (interquartile range [IQR]) recruitment was 45.2% (19.6%-100%). Trials were less likely to begin recruitment when they were less than vs greater than median sample size (152/250 [60.8%] vs 203/256 [79.3%]; adjusted odds ratio [OR], 0.60; 95% CI, 0.39-0.92; P = .02) or when they were registered in China vs elsewhere (152/267 [56.2%] vs 204/242 [84.3%]; adjusted OR, 0.32; 95% CI, 0.17-0.61; P < .001) (Table).
Of the 53 trials with reported results, 13 (24.5%) were posted as preprints, 39 (73.6%) were published in peer-reviewed journals, and 1 (1.9%) was posted on a registry. Results reporting was less likely for trials from China vs elsewhere (25/270 [9.3%] vs 28/244 [11.5%]; adjusted OR, 0.24; 95% CI, 0.10-0.62; P = .003) but more likely for trials registered before the median registration date (March 15, 2020) vs later (33/252 [13.1%] vs 20/259 [7.7%]; adjusted OR, 6.08; 95% CI, 2.5-14.99; P < .001). Type of control group was not associated with start of recruitment or results reporting.
Of the 24 terminated and 46 discontinued trials, investigators of 14 RCTs communicated the reasons for stopping. Most frequently, trials were stopped because of decreasing COVID-19 cases (8 trials [57.1%]) and emerging data raising concerns of safety (6 trials [42.9%]) or futility (2 trials [14.3%]).
Overall, 30% of COVID-19 RCTs initiated in the first 100 days of the pandemic did not begin recruitment, and only 10% had results reported by mid-October, suggesting the possibility of substantial research waste.3 Delayed completion may be explained for some countries, such as China, by slow or halted recruitment due to decreasing COVID-19 cases after the first wave.
This study is limited by a low response rate to the survey and by unknown results because of inconsistent updates of trial registries and delayed publications. Successfully conducting a trial during a pandemic is challenging. Nevertheless, collaborative efforts such as consortiums of trials prospectively planning to pool their results4 and adaptive platform trials such as the RECOVERY trial5 are promising approaches to provide reliable and timely evidence.
Accepted for Publication: January 10, 2021.
Published: March 1, 2021. doi:10.1001/jamanetworkopen.2021.0330
Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2021 Janiaud P et al. JAMA Network Open.
Concept and design: Janiaud, Ioannidis, Hemkens.
Acquisition, analysis, or interpretation of data: Janiaud, Axfors, Hemkens.
Drafting of the manuscript: Janiaud, Hemkens.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Janiaud, Ioannidis, Hemkens.
Obtained funding: Hemkens.
Administrative, technical, or material support: Axfors.
Conflict of Interest Disclosures: Dr Axfors reported receiving grants from Uppsala University, Swedish Society of Medicine, Blanceflor Foundation, and Sweden-America Foundation outside the submitted work. No other disclosures were reported.
Funding/Support: Meta-Research Innovation Center at Stanford, Stanford University, is supported by a grant from the Laura and John Arnold Foundation. COVID-evidence is supported by the Swiss National Science Foundation (grant No. 31CA30_196190).
Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Additional Contributions: We would like to thank the COVID-evidence team and supporters (www.covid-evidence.org). We would also like to thank all the investigators we reached out to who took the time to respond to our inquiries.