Comparison of different scoring systems for predicting 28-day mortality in critically ill patients with acute...

This study compared eight scoring systems for predicting 28-day and 1-year all-cause mortality in critically ill patients with acute pancreatitis (AP).

Data from the Medical Information Mart for Intensive Care IV were used to conduct a comparative analysis of several predictive scoring systems. Predictive performance for 28-day and 1-year mortality was assessed using receiver operating characteristic (ROC) curves (area under the curve [AUC]), restricted cubic splines (RCS) for nonlinearity testing, and multivariable logistic regression for independent predictor analysis.

A total of 694 patients were included (28-day mortality: 15.56%; 1-year mortality: 24.78%). Acute Physiology Score III (APSIII) demonstrated the highest accuracy for 28-day mortality (AUC: 0.847, 95% confidence interval (CI): 0.808-0.886), followed by Bedside Index for Severity in Acute Pancreatitis (BISAP) (AUC: 0.835, 95% CI: 0.794-0.875). Linear relationships between scores and 28-day mortality were confirmed (all p for nonlinear > 0.05). Multivariable regression identified APSIII and BISAP as independent 28-day mortality predictors. For 1-year mortality, APSIII, BISAP, and Simplified Acute Physiology Score II (SAPS II) were independent predictors.

Both APSIII and BISAP were identified as independent predictors of 28-day mortality, while APSIII, BISAP, and SAPSII were associated with 1-year mortality. Among them, APSIII showed the best overall discriminative ability for both short- and long-term outcomes. However, BISAP remains an attractive alternative for its simplicity and comparable performance.