Objective To investigate the risk of ICU-acquired coagulopathy (ICU-AC) in patients with severe acute pancreatitis (SAP) undergoing blood purification (BP) and develop a predictive model. Methods A retrospective observational study was conducted on 282 SAP patients treated with BP in the ICU of Shanghai Changzheng Hospital from January 2022 to December 2024. Patients were divided into ICU-AC and non-ICU-AC groups based on the occurrence of ICU-AC. Multivariate logistic regression analysis was used to identify risk factors for ICU-AC in SAP patients undergoing BP and construct a predictive model. The model’s predictive accuracy, calibration, and clinical utility were evaluated using the concordance index (C-index), calibration plots, and decision curve analysis (DCA). Results The incidence of ICU-AC among SAP patients during ICU stay was 19.50%. APACHE Ⅱ score, platelet count (PLT), D-dimer, septic shock, procalcitonin (PCT), vasoactive drug use, and filter lifespan were identified as independent risk factors for ICU-AC in SAP patients undergoing BP (P<0.05). A predictive model based on logistic regression analysis demonstrated strong discriminative ability ［area under the curve (AUC)=0.893, P<0.05］, good calibration (C-index=0.902, mean absolute error =0.003), and significant clinical benefit as shown by DCA. Conclusion APACHE Ⅱ score, PLT, D-dimer, septic shock, PCT, vasoactive drug use, and filter lifespan are risk factors for ICU-AC in SAP patients undergoing BP. The predictive model developed based on these factors exhibits high predictive accuracy and clinical applicability