Abstract
BACKGROUND: Currently, some meta-analyses on COVID-19 have suggested that glucocorticoids use can reduce the mortality rate of COVID-19 patients, utilization rate of invasive ventilation, and improve the prognosis of patients. However, optimal regimen and dosages of glucocorticoid remain unclear. Therefore, the purpose of this network meta-analysis is to analyze the efficacy and safety of glucocorticoids in treating COVID-19 at regimens. METHODS: This meta-analysis retrieved randomized controlled trials from the earliest records to December 30, 2022, published in PubMed, Embase, Cochrane Library, CNKI Database and Wanfang Database, which compared glucocorticoids with placebos for their efficacy and safety in the treatment of COVID-19, Effects of different treatment regimens, types and dosages (high-dose methylprednisolone, very high-dose methylprednisolone, Pulse therapy methylprednisolone, medium-dose hydrocortisone, high-dose hydrocortisone, high-dose dexamethasone, very high-dose dexamethasone and placebo) on 28-day all-caused hospitalization mortality, hospitalization duration, mechanical ventilation requirement, ICU admission and safety outcome were compared. RESULTS: In this network meta-analysis, a total of 10,544 patients from 19 randomized controlled trials were finally included, involving a total of 9 glucocorticoid treatment regimens of different types and dosages. According to the analysis results, the 28-day all-cause mortality rate was the lowest in the treatment with pulse therapy methylprednisolone (OR 0.08, 95% CI 0.02, 0.42), but the use of high-dose methylprednisolone (OR 0.85, 95% CI 0.59, 1.22), very high-dose dexamethasone (OR 0.95, 95% CI 0.67, 1.35), high-dose hydrocortisone (OR 0.64, 95% CI 0.34, 1.22), medium-dose hydrocortisone (OR 0.80, 95% CI 0.49, 1.31) showed no benefit in prolonging the 28-day survival of patient. Compared with placebo, the treatment with very high-dose methylprednisolone (MD = -3.09;95%CI: -4.10, -2.08) had the shortest length of hospital stay, while high-dose dexamethasone (MD = -1.55;95%CI: -3.13,0.03) and very high-dose dexamethasone (MD = -1.06;95%CI: -2.78,0.67) did not benefit patients in terms of length of stay. CONCLUSIONS: Considering the available evidence, this network meta‑analysis suggests that the prognostic impact of glucocorticoids in patients with COVID-19 may depend on the regimens of glucocorticoids. It is suggested that pulse therapy methylprednisolone is associated with lower 28-day all-cause mortality, very high-dose methylprednisolone had the shortest length of hospital stay in patients with COVID-19. TRIAL REGISTRATION: PROSPERO CRD42022350407 (22/08/2022).