The association between blood urea nitrogen to albumin ratio and short- and long-term all-cause mortalities in acute pancreatitis: Insights from MIMIC-IV database

While the blood urea nitrogen/albumin ratio (BAR) is recognized for its prognostic significance in various diseases, its association in acute pancreatitis (AP) remains uncertain. This research seeks to clarify the relationship between BAR and mortality in patients with AP. Using data from Medical Information Mart for Intensive Care IV, this study examined the relationship between the BAR and mortality risk in ICU-admitted AP patients. Feature selection was performed using the Boruta algorithm, and risk stratification was conducted via X-tile analysis. Multivariable Cox regression assessed BAR's association with short- and long-term outcomes, while Kaplan-Meier analysis compared survival trends. The predictive performance of BAR was evaluated using receiver operating characteristic curve analysis. Restricted cubic splines tested for nonlinear associations, and subgroup analyses explored interactions across clinically relevant subsets. This study analyzed 492 patients stratified by BAR using an optimal cutoff of 9.62 (determined via X-tile software based on 28-day mortality). Elevated BAR (≥9.62) was independently associated with significantly higher all-cause mortality at 28, 90, and 365 days (all P < .001; all hazard ratios in Cox regression showed P < .05). Kaplan-Meier analysis analysis demonstrated significantly reduced survival in high-BAR patients (P < .001). BAR showed strong predictive capacity for mortality (receiver operating characteristic analysis), with a positive correlation with mortality risk revealed by restricted cubic splines analysis (P < .05) and no significant interaction effects across subgroups. BAR independently predicts all-cause mortality in AP.