ABSTRACT
Objectives Although several randomised controlled trials have compared the efficacy of remdesivir with that of placebo, there is limited evidence regarding its effect in the early stage of nonsevere COVID-19 cases. Methods We evaluated the efficacy of remdesivir on the early stage of nonsevere COVID-19 using the COVID-19 Registry Japan, a nationwide registry of hospitalised COVID-19 patients in Japan. Two regimens (start remdesivir therapy within 4 days from admission vs. no remdesivir during hospitalisation) among patients without the need for supplementary oxygen therapy were compared by a three-step processing (cloning, censoring, and weighting) method. The primary outcome was supplementary oxygen requirement during hospitalisation. Secondary outcomes were 30-day fatality risk and risk of invasive mechanical ventilation or extracorporeal membrane oxygenation (IMV/ECMO). Results The data of 12,657 cases met our inclusion criteria. The ‘start remdesivir’ regimen showed a lower risk of supplementary oxygen requirement (hazard ratio: 0.861, p < 0.001). Both 30-day fatality risk and risk of IMV/ECMO introduction were not significantly different between the two regimens (hazard ratios: 1.05 and 0.886, p values: 0.070 and 0.440, respectively). Conclusions Remdesivir might reduce the risk of oxygen requirement during hospitalisation in the early stage of COVID-19; however, it had no positive effect on the clinical outcome and reduction of IMV/ECMO requirement.