The neutrophil percentage-to-albumin ratio predicts 28-day mortality in ICU patients with acute pancreatitis: a retrospective cohort study with machine learning analysis

Abstract Background The neutrophil percentage-to-albumin ratio (NPAR) has recently been recognized as a marker of disease severity in critically ill patients. Its utility in acute pancreatitis (AP) is not well established. Given the high mortality associated with severe AP, early risk stratification is crucial. This study aimed to examine the association between NPAR and 28-day all-cause mortality in patients with acute pancreatitis (AP) in the ICU. Methods This retrospective cohort study extracted data from the MIMIC-IV database. The primary endpoint was 28-day all-cause mortality. The association between NPAR and mortality was evaluated using multivariable Cox regression, restricted cubic spline (RCS) analysis, and Kaplan-Meier survival analysis. Its predictive performance was assessed by the area under the receiver operating characteristic curve (AUC) and compared with individual components and the SOFA score. Machine learning models were further constructed to validate its importance. Results Among 532 eligible patients, 73 (13.7%) died within 28 days. Non-survivors had significantly higher NPAR levels ( P < 0.001). RCS analysis showed a positive linear association between NPAR and mortality risk (P for overall < 0.001). After adjusting for confounders (age, sex, SOFA score, comorbidities), elevated NPAR remained an independent predictor of mortality (HR 1.031, 95% CI 1.005–1.058, P = 0.021). NPAR demonstrated an AUC of 0.71 (95% CI 0.63–0.78) for mortality prediction, superior to neutrophil percentage (0.60) or albumin alone (0.67), though slightly lower than the SOFA score (0.75). Patients with NPAR > 28.81 had significantly worse survival (log-rank P < 0.001). Subgroup analyses confirmed its consistent predictive value. In machine learning models, NPAR was identified as a key predictive feature. Conclusion Elevated NPAR was independently associated with increased 28-day all-cause mortality in critically ill patients with AP. It may serve as a complementary prognostic signal that helps identify high-risk patients who warrant closer monitoring.