Early Enteral Nutrition and Sepsis-Associated Acute Kidney Injury: A Propensity Score Matched Cohort Study Based on the MIMIC-III Database

PURPOSE: We aimed to analyze the optimal timing of enteral nutrition (EN) in the treatment of sepsis and its effect on sepsis-associated acute kidney injury (SA-AKI.) MATERIALS AND METHODS: The MIMIC-III database was employed to identify patients with sepsis who had received EN. With AKI as the primary outcome variable, receiver operating characteristic (ROC) curves were utilized to calculate the optimal cut-off time of early EN (EEN). Propensity score matching (PSM) was employed to control confounding effects. Logistic regressions and propensity score-based inverse probability of treatment weighting were utilized to assess the robustness of our findings. Comparisons within the EEN group were performed. RESULTS: 2364 patients were included in our study. With 53 hours after intensive care units (ICU) admission as the cut-off time of EEN according to the ROC curve, 1212 patients were assigned to the EEN group and the other 1152 to the delayed EN group. The risk of SA-AKI was reduced in the EEN group (odds ratio 0.319, 95% confidence interval 0.245–0.413, p<0.001). The EEN patients received fewer volumes (mL) of intravenous fluid (IVF) during their ICU stay (3750 mL vs. 5513.23 mL, p<0.001). The mediating effect of IVF was significant (p<0.001 for the average causal mediation effect). No significant differences were found within the EEN group (0–48 hours vs. 48–53 hours), except that patients initiating EN within 48 hours spent fewer days in ICU and hospital. CONCLUSION: EEN is associated with decreased risk of SA-AKI, and this beneficial effect may be proportionally mediated by IVF volume.