| Popis: |
Background: Current evidence from randomised controlled trials (RCTs) and systematic reviews on the use of interleukin-6 receptor (IL-6R) antagonists among patients with severe coronavirus disease 2019 (COVID-19) have yielded equivocal results. We conducted a meta-analysis of reconstructed individual participant data (IPD) from RCTs, in order to test the hypothesis that IL-6R antagonists lower mortality in severe COVID-19. Methods: We searched PubMed, Embase, Cochrane, Scopus and medRXiv from 1st December 2019 to 15th April 2021 for RCTs reporting on IL-6R antagonists in patients with severe COVID-19. Survival of individual patients was reconstructed from published Kaplan-Meier curves using a plot digitiser. Weibull mixture “cure” survival models were fitted to compare the overall survival and the “cured” fractions of the IL-6R antagonist and control arms. As a sensitivity analysis, we stratified our analysis according to the type and maximum dose of the IL-6R antagonist, and pooled the aggregate odds ratio (OR) for a cure from COVID-19. Secondary outcomes were analysed using a conventional two-stage meta-analysis, and include the pooled relative risk (RR) for secondary infections, and the pooled mean difference in ventilator-free days. The risk of bias was assessed using the Cochrane Risk-of-Bias Tool for RCTs, and overall certainty of the evidence using the Grading of Recommendations, Assessment, Development and Evaluations (GRADE) approach. Findings: 9 RCTs (6831 patients) reported on survival outcomes in COVID-19, of which data from 6 RCTs (6278 patients) was used for the reconstructed IPD meta-analysis. The fractions of “cured” patients who survived were estimated to be 74·0% and 67·9% in the IL-6R antagonist and control arms respectively, which corresponds to an absolute percentage increase in the fraction of COVID-19 survivors of +6·7% (95%-CI: 4·3% to 9·4%, p |
