Methodological Rigor in COVID-19 Clinical Research: A Systematic Review and Case-Control Analysis

Jung, R. G.; Di Santo, P.; Clifford, C.; Prosperi-Porta, G.; Skanes, S.; Hung, A.; Parlow, S.; Visintini, S.; Ramirez, F. D.; Simard, T.; Hibbert, B.

Objective: To systematically evaluate the quality of reporting of currently available COVID-19 studies compared to historical controls. Design: A systematic review and case-control analysis Data sources: MEDLINE, Embase, and Cochrane Central Register of Controlled Trials until May 14, 2020 Study selection: All original clinical literature evaluating COVID-19 or SARS-CoV2 were identified and 1:1 historical control of the same study type in the same published journal was matched from the previous year Data extraction: Two independent reviewers screened titles, abstracts, and full-texts and independently assessed methodological quality using Cochrane Risk of Bias Tool, Newcastle-Ottawa Scale, QUADAS-2 Score, or case series checklist. Results: 9895 titles and abstracts were screened and 686 COVID-19 articles were included in the final analysis in which 380 (55.4%) were case series, 199 (29.0%) were cohort, 63 (9.2%) were diagnostic, 38 (5.5%) were case-control, and 6 (0.9%) were randomized controlled trials. Overall, high quality/low-bias studies represented less than half of COVID-19 articles - 49.0% of case series, 43.9% of cohort, 31.6% of case-control, and 6.4% of diagnostic studies. We matched 539 control articles to COVID-19 articles from the same journal in the previous year for a final analysis of 1078 articles. The median time to acceptance was 13.0 (IQR, 5.0-25.0) days in COVID-19 articles vs. 110.0 (IQR, 71.0-156.0) days in control articles (p<0.0001). Overall, methodological quality was lower in COVID-19 articles with 220 COVID-19 articles of high quality (41.0%) vs. 392 control articles (73.3%, p<0.0001) with similar results when stratified by study design. In both unadjusted and adjusted logistic regression, COVID-19 articles were associated with lower methodological quality (odds ratio, 0.25; 95% CI, 0.20 to 0.33, p<0.0001). Conclusion: Currently published COVID-19 studies were accepted more quickly and were found to be of lower methodological quality than comparative studies published in the same journal. Given the implications of these studies to medical decision making and government policy, greater effort to appropriately weigh the existing evidence in the context of emerging high-quality research is needed. Study registration: PROSPERO: CRD42020187318