June 23, 2020

Effectiveness and Safety of Glucocorticoids to Treat COVID-19: A Rapid Review and Meta-Analysis

Category: Article Summary

Topic: Testing and Treatment

• A meta-analysis (n=23 studies, 1 RCT and 22 cohort studies; 13,815 patients) found the use of glucocorticoid steroids was associated with a non-significant increase in mortality (RR=2.00, 95%CI 0.69, 5.75) and duration of hospitalization (increase of 2.4 days, 95%CI 1.4, 3.4) but significantly lower duration of fever (-3.23 days, 95%CI -3.56, -2.90). Heterogeneity was high across the included studies (I2 84.6% to 97.9%), likely due to differences in patient characteristics and glucocorticoid protocols, suggesting pooled effect estimates should be interpreted with caution. [EDITORIAL NOTE: While the authors used a technique to downweight the contribution of studies with higher potential for bias (e.g., due to confounding), they do not discuss the potential for confounding by indication, whereby the sickest patients may have been more likely to receive glucocorticoid treatment] 

Lu et al. (May 2020). Effectiveness and Safety of Glucocorticoids to Treat COVID-19: A Rapid Review and Meta-Analysis. Annals of Translational Medicine. https://doi.org/10.21037/atm-20-3307