Factors Associated with ICU Admission and In-Hospital Mortality in Patients with Acute Pancreatitis: A Single-Center Retrospective Study

Abstract

Background: Early identification of patients with acute pancreatitis (AP) who may require intensive care unit (ICU) admission is critical for optimizing resource allocation and improving clinical outcomes. This study aimed to identify factors associated with ICU admission and to explore markers associated with in-hospital mortality in patients with AP using routinely available clinical and laboratory data. Methods: This retrospective cohort study included 931 patients with AP admitted to a tertiary care center between June 2020 and June 2024. Demographic, clinical, and laboratory parameters obtained within 24 hours of admission were analyzed. Multivariable logistic regression with forward stepwise selection and least absolute shrinkage and selection operator (LASSO) regression were employed to identify factors associated with ICU admission. Among ICU-admitted patients, LASSO-Cox regression was performed as an exploratory analysis to identify markers associated with in-hospital mortality. Model performance was assessed using receiver operating characteristic (ROC) curve analysis and internal validation with bootstrap resampling. Results: Of the 931 patients, 133 (14.3%) required ICU admission. Hypertriglyceridemic pancreatitis (OR 2.88, 95% CI 1.59-5.21), elevated D‑dimer (OR 1.13, 95% CI 1.07-1.18), blood urea nitrogen (BUN) (OR 1.36, 95% CI 1.25-1.48), and red blood cell distribution width (RDW) (OR 1.14, 95% CI 1.09-1.19) were independently associated with ICU admission. A nomogram incorporating these factors demonstrated good discrimination (AUC 0.89, 95% CI 0.86-0.93) and calibration (Hosmer-Lemeshow P = 0.517). Among ICU patients (n = 133, 20 deaths), LASSO-Cox regression identified elevated C‑reactive protein‑to‑albumin ratio (CAR) as a possible predictor of in‑hospital mortality (HR 2.12, 95% CI 1.76-3.27), with an AUC of 0.78 (95% CI 0.69-0.88). Kaplan‑Meier analysis revealed significantly lower survival in ICU‑admitted patients (log‑rank P < 0.001). Conclusion: Routinely available inflammatory and biochemical markers might be associated with ICU admission and in-hospital mortality in patients with acute pancreatitis. These findings may assist early clinical risk stratification and complement existing clinical assessments. However, prospective multicenter studies with larger cohorts are warranted to validate these findings.