Implications of the lack of a unified research project framework: an investigation into the registration of clinical trials of COVID-19

Abstract

Objective: The aim of this study was to provide recommendations for improving the design of subsequent studies through analysis of the registered coronavirus disease 2019 (COVID-19) clinical trials. Methods: A retrospective analysis of 189 trial retrievals achieved on 20 February 2020. Results: A total of 189 trials are included in the study. There were 69.3% interventional studies, 21.7% observational studies, 5.3% diagnostic tests and 3.7% other studies. The following statistics are provided only for the interventional studies. Severity of disease: 5.3% light and common type, 17.6% severe and critically ill and 59.6% with no restricted classification. Medication use: 51.1% Western medicine, 32.1% Chinese medicine, 10.7% blood related product and 6.1% non-drug therapy. The median and inner quantile range of the sample sizes included in these studies: 104 (IQR: 60, 200). Primary outcome type most used: 45.8% with clinical characteristics and 21.4% with virological. Study design characteristics: 71% of all studies were randomized, 5% of all studies were blinded, 18% of all studies were multicenter and 76% of all studies were single center. Conclusion: Although many COVID-19 studies include randomization in their design, the lack of additional double-blind and placebo-controlled elements in their designs result in a less robust evaluation of intervention safety and efficacy. Furthermore, similar or repeated research and small sample studies that have less promise in gains of new information have possibly led to a shortage of recruitable patients and become a barrier to the completion of large multicenter clinical trial studies.