Prediction of early mortality in acute pancreatitis using arterial base excess: an international multicenter cohort...

Early mortality rates for severe acute pancreatitis (SAP) patients can reach 20-30%. This study aims to evaluate the predictive value of arterial base excess (BE) for early mortality in AP patients.

This study included Nanchang cohort and MIMIC-IV database. Patients were categorized into four groups-based BE values. Multivariate logistic regression analysis was used to identify independent predictors of early mortality, and a nomogram model was developed for risk prediction.

Patients in the lowest BE quartile (Q1) had significantly higher rates of SAP (48.2% vs. 15.3%), persistent multiple organ failure (21.1% vs. 2.5%), in-hospital mortality (13.3% vs. 1.3%), and 7-day mortality (5.4% vs. 0.3%) compared to those in the highest quartile (Q4) (all p < 0.001). Multivariate analysis showed that BE is an independent predictor of early mortality, with every 1-standard-deviation increase in BE was associated with a reduction in the risk of death within 7 days (OR = 0.59, 95% CI 0.49-0.71). The AUC of BE for predicting death within 7 days was 0.85 (95% CI: 0.80-0.90). A nomogram model integrating BE, age, LDH, mean arterial pressure, creatinine, and total bilirubin achieved an AUC of 0.918 in the training set and 0.885 in the test set. External validation in the MIMIC-IV cohort showed an AUC of 0.749.

Arterial BE is a powerful independent predictor of early mortality in AP patients. The nomogram model developed in this study demonstrates good discrimination and clinical utility.