High admission blood glucose is an independent risk factor of poor prognosis in COVID-19: A systematic review and dose-response meta-analysis

Aims To investigate the prognostic value of admission blood glucose (BG) in predicting COVID-19 outcomes, including poor composite outcomes (mortality/severity), mortality, and severity. Materials and methods Eligible studies evaluating the prognostic value of fasting BG (FBG) and random BG (RBG) levels in predicting COVID-19 outcomes were included and assessed for risk of bias with the Quality in Prognosis Studies tool. Random-effects high-vs-low meta-analysis followed by dose-response analysis using generalized least squares model in a two-stage random-effects meta-analysis were conducted. Potential non-linear association was explored using restricted cubic splines and pooled using restricted maximum likelihood model in a multivariate meta-analysis. Results The search yielded 35 studies involving a total of 14,502 patients. We discovered independent association between admission FBG and poor prognosis in COVID-19 patients. Furthermore, we demonstrated non-linear relationship between admission FBG and severity (Pnon-linearity<0.001), where each 1 mmol/L increase augmented the risk of COVID-19 severity by 33% (risk ratio 1.33 [95% CI: 1.26-1.40]). Albeit exhibiting similar trends, study scarcity limited the strength of evidence on the independent prognostic value of admission RBG. GRADE assessment yielded high-quality evidence for the association between admission FBG and COVID-19 severity, and moderate-quality evidence for its association with mortality and poor outcomes, while the other assessments yielded very low-to-low quality. Conclusion High level of FBG at admission was independently associated with poor prognosis in COVID-19 patients. Further researches to confirm the observed prognostic value of admission RBG and to ascertain the estimated dose-response risk between admission FBG and on COVID-19 severity are required.