Abstract
BACKGROUND: Corticosteroids confer a survival benefit in individuals hospitalized with coronavirus disease 2019 (COVID-19) who require oxygen. This meta-analysis seeks to determine the duration of corticosteroids needed to optimize this mortality benefit. METHODS: Electronic databases were searched to 9 March 2022, for studies reporting corticosteroid versus no corticosteroid treatment in hospitalized COVID-19 patients. We estimated the effect of corticosteroids on mortality by random-effects meta-analyses. Subgroup analyses and meta-analyses were conducted to assess the optimal duration of corticosteroid treatment while adjusting for the severity of disease, age, duration of symptoms, and proportion of control group given steroids. RESULTS: We identified 27 eligible studies consisting of 13 404 hospitalized COVID-19 patients. Seven randomized controlled trials and 20 observational studies were included in the meta-analysis of mortality, which suggested a protective association with corticosteroid therapy (risk ratio [RR], 0.71 [95% confidence interval {CI}, .58-.87]). Pooled analysis of 18 studies showed the greatest survival benefit for a treatment duration up to 6 days (RR, 0.54 [95% CI, .39-.74]). Survival benefit was 0.65 (95% CI, .51-.83) up to 7 days, and no additional survival benefit was observed beyond 7 days of treatment (RR, 0.64 [95% CI, .44-.93]). The survival benefit was not confounded by severity of disease, age, duration of symptoms, or proportion of control group given steroids. CONCLUSIONS: In this meta-analysis, optimal duration of corticosteroid treatment for hospitalized COVID-19 patients was up to 6 days, with no additional survival benefit with >7 days of treatment.