Association between estimated glucose disposal rate and 28-day mortality in patients with sepsis: analysis of the MIMIC‑IV database

Background The aim of the study was to examine the correlation between estimated glucose disposal rate (eGDR) and 28-day mortality in septic patients. Methods Using information from the MIMIC IV 2.1 database, we conducted a retrospective cohort analysis. We gathered patient demographics and used the eGDR to characterize insulin resistance. Participants were stratified into tertiles based on eGDR values and constructed Cox regression models with 28-day mortality in sepsis patients as the primary outcome measure. In order to clearly depict the relationship between eGDR and 28-day mortality, we used Kaplan–Meier (KM) curves and curve fitting. These models were designed to actively test for a potential possible correlation between eGDR and 28-day mortality in sepsis patients. Subgroup analysis was performed to explore the stability of the primary results. Results Enrolling a total of 672 patients, 58.8% were male and the 28-day mortality was 17.4%. An association was found between higher eGDR and higher 28-day mortality using multivariate cox regression analysis. Hazard ratio (HR) for 28-day mortality per unit increase in eGDR was 1.48[95% CI 1.16–1.89]. Compared to patients in the lowest tertile of eGDR, those in the highest tertile exhibited a higher 28-day mortality rate (HR, 2.53 [95% CI 1.49–4.37], adjusting for confounding factors). Subgroup analyses demonstrated that the effect of eGDR on mortality was robust across subgroups. Conclusion The eGDR is correlated with the28-day mortality in sepsis patients. Larger prospective studies are required in the future to further validate this finding.