Predictive value of albumin-corrected anion gap for acute kidney injury risk assessment in sepsis: Insights from MIMIC-IV

Abstract

Acute kidney injury (AKI) represents a major cause of death among patients with sepsis. While recent evidence suggests that albumin-corrected anion gap (ACAG) might serve as an early indicator of AKI, its prognostic capabilities require further investigation. We conducted a retrospective analysis of critical care data from the Medical Information Mart for Intensive Care IV electronic health record repository. Restricted cubic splines (RCSs) and Cox proportional hazards models were employed to quantify ACAG’s association with AKI occurrence and in-hospital mortality. Kaplan–Meier survival analyses were performed to compare outcomes across ACAG-based groups, and subgroup analyses were carried out to evaluate the robustness of these associations and potential interactions between variables. Among 19,445 included patients, elevated ACAG emerged as a strong predictor of AKI (hazard ratio, HR (95% confidence interval, CI): 16.75 (14.50, 19.75), p < 0.001) and in-hospital mortality (HR (95% CI): 16.75 (14.50, 19.75), p < 0.001). RCS analysis showed a predominantly linear relationship between ACAG and clinical outcomes (AKI: p for nonlinear = 0.059; mortality: p for nonlinear = 0.794), with 17 emerging as a critical ACAG threshold value. Significant interactions were identified between ACAG and factors such as sex, age, and chronic kidney disease status. Our findings demonstrate ACAG’s robust ability to predict adverse outcomes in sepsis, particularly regarding kidney function deterioration and survival. These findings highlight the potential clinical utility of ACAG as a prognostic marker to guide early therapeutic interventions.