Association between peak D-dimer levels within 24 hours of admission and the risk of infected pancreatic necrosis in hyperlipidemic acute pancreatitis: a retrospective analysis

Vollständiger Abstract

Worum geht es in dieser Arbeit?

Objective The goal of this study is to analyze the relationship between the peak D-dimer levels measured within 24 h of admission and the risk of infected pancreatic necrosis (IPN) in patients with hyperlipidemic acute pancreatitis (HLAP) and to create and validate an early in-hospital risk stratification model of IPN. Methods The study was conducted as a retrospective study on 286 HLAP patients from January 2023 – June 2025, divided into IPN ( n = 47) and non-IPN ( n = 239) groups. A separate time-validation group was collected ( n = 62) from July 2025 to December 2025. Multivariate logistic regression was used to identify independent risk factors for IPN and build the nomogram. The ROC curve, calibration curve, Hosmer-Lemeshow test, Delong test, and DCA were performed to evaluate model performance. Using restricted cubic spline (RCS), the dose–response relationship between D-dimer and IPN risk was evaluated. Results D-dimer, procalcitonin (PCT), MCTSI score, and persistent organ failure (POF) were independent risk factors for IPN (all p < 0.05). RCS analysis revealed a linear dose–response relationship between D-dimer and IPN risk (P for overall <0.001, P for nonlinear = 0.119). The combined model incorporating D-dimer, PCT, MCTSI, and POF achieved an AUC of 0.82 (95% CI: 0.75–0.89), with sensitivity of 0.70, specificity of 0.82, and accuracy of 0.80. Calibration was good (Hosmer-Lemeshow p = 0.806), and DCA demonstrated clinical net benefit across threshold probabilities of 0.1–0.7. In the temporal validation cohort, the model yielded an AUC of 0.73 (95% CI: 0.56–0.89), with no significant difference from the derivation cohort ( p = 0.319). Given the limited sample size of the validation cohort and its single-center origin, these findings should be considered preliminary. Conclusion Elevated peak D-dimer level within 24 h of admission is independently associated with an increased risk of IPN in HLAP patients in a linear positive manner. The early in-hospital risk stratification model comprised of D-dimer, PCT, MCTSI and POF demonstrates additionally good predictive performance, clinical utility and risk stratification during the early course of hospitalization for personalized treatment decision.