Survival of adults hospitalized with COVID-19 and remdesivir: Bayesian re-analysis and meta-analysis of two clinical trials

Abstract

INTRODUCTION: remdesivir is one of the numerous treatments evaluated against COVID-19, for which no effect on mortality has been reported. The objective of this review is to evaluate the effect of remdesivir on mortality in the first 28 days in patients hospitalized for COVID-19. METHODS: a review was made in PubMed. Only randomized trials reporting 28-day mortality were selected. The treatment effect was evaluated using a fixed effect Bayesian model, and calculating the posterior distribution of relative risk (RR) and hazard ratio (HR). RESULTS: two clinical trials were found that met the selection criteria. The probability of a combined RR of less than 1 in favor of remdesivir for mortality at 14 days was 98%, with an 95 % High Density credibility Interval (HDCrI 95%) [0.42 to 0.91], while, by re-analysis of Beigel’s ACTT-1, the probability of a HR less than 1 in favor of remdesivir was 96%, with an HDCrI 95% [0.44 to 0.99]. CONCLUSION: both the meta-analysis of and the Bayesian re-analysis support the existence of a favorable effect of treatment with remdesivir in patients with COVID-19 pneumonia.