Abstract
INTRODUCTION: Several randomized controlled trials have compared the effectiveness of favipiravir with that of placebo. However, evidence regarding its effect on nonsevere, early-stage coronavirus disease 2019 (COVID-19) remains insufficient. METHODS: We used the COVID-19 Registry Japan, a nationwide registry of inpatients with COVID-19, for evaluating the effectiveness of favipiravir on patients with nonsevere, early-stage COVID-19. Eligible patients, who did not need supplementary oxygen therapy at admission, were classified according to two regimens (starting favipiravir therapy within 4 days from admission vs. no favipiravir during hospitalization) and were then compared using a three-step method (cloning, censoring, and weighting). The primary outcome was supplementary oxygen requirement during hospitalization, and the secondary outcomes were the need for invasive mechanical ventilation or extracorporeal membrane oxygenation (IMV/ECMO) and overall mortality at 30 days. RESULTS: A total of 7654 cases were analyzed. The "start favipiravir" regimen did not show substantial differences in the primary outcome [hazard ratio 0.825, 95% confidence interval (CI) 0.657-1.04, p = 0.098] and both of the secondary outcomes [need for IMV/ECMO and overall 30-day mortality, hazard ratio 1.02 (95% CI 0.649-1.60) and 0.869 (95% CI 0.519-1.46), p = 0.929 and 0.594, respectively]. CONCLUSIONS: In this large cohort from a COVID-19 registry, favipiravir was not associated with a positive effect on the clinical outcome on patients with nonsevere, early-stage COVID-19, suggesting that it is not an essential drug for COVID-19 treatment.