Is Metformin Use Associated With a Decreased Mortality for COVID-19 Diabetic Patients? A Meta-Analysis

Introduction: Coronavirus disease 2019 (COVID-19) has been spreading globally for more than half a year. Previous studies remain controversial regarding whether metformin is associated with reduced risk for COVID-19 diabetic patients. Thus, this meta-analysis is performed. Method: A comprehensive literature search on PubMed and Web of Science was conducted to identify all relevant studies published prior to October 2020 according to the established inclusion criteria. This meta-analysis was reported in conformity to the Preferred Reporting Project declared by the Systematic Review and Meta-Analysis (PRISMA). The quality assessment was performed by the Newcastle-Ottawa Scale (NOS). The pooled odds ratio (OR) and 95% confidence intervals (CI) were calculated to estimate the association between metformin use and mortality for COVID-19 patients. A random-effect or fixed-effect model was used based on heterogeneity significance. Subgroup analysis was performed based on in-hospital-use or home-use, and different sample sizes. Sensitivity analysis and publication bias detection were also performed. All statistical analyses were performed using RevMan software (version 5.3; Cochrane library) and STATA 12.0 statistical software (Stata Corp., College Station, TX), and all P values were two-tailed, the test level was 0.05. Result: 97 articles were obtained from the database search, and 5 articles obtained from other sources. 8 articles involving 11,169 participants were included. Most studies were considered moderate quality. A statistically significant association between metformin use and decreased mortality of COVID-19 patients was found (OR 0.53, 95%CI: 0.34, 0.83, P=0.005, I(2)=77%). In the subgroup analyses, home-use of metformin was also associated with a reduced risk of mortality (OR 0.54, 95%CI: 0.35, 0.84, P=0.006, I(2)=66%), and one study reporting in-hospital use did not find reduced mortality among COVID-19 patients taking metformin (OR 1.65, 95%CI: 0.71, 3.86, P=0.247). For sample size >1,000, no statistically significant reduced risk of mortality (OR 0.84, 95%CI: 0.57, 1.26, P=0.41, I(2)=73%) was found, however, for sample ≤1,000, a statistically significant reduced risk of mortality (OR 0.29, 95%CI: 0.19, 0.44 P<0.00001, I(2)=0%) was found. Sensitivity analysis by change fixed-effect models to random-effect models and by omitting each study at a time confirmed the relative stability of the result. Begg’s test (z=0.37, P=0.711) and Egger’s test (t=-1.98, P=0.096) did not detect a significant risk of publication bias. Conclusion: The current meta-analysis demonstrates that metformin use is associated with decreased mortality for COVID-19 diabetic patients. However, only one study investigating the in-hospital use of metformin. More high-quality original studies are needed to further explore the association between metformin use and mortality risk of COVID-19.