Development of a risk prediction model for early peritoneal effusion in acute pancreatitis

Feng Xiaotian; Huang Xiaodong; Liu Siyao; Liu Xiong; Lin Jiyan

doi:10.3969/j.issn.1002-1949.2026.01.004

Abstract：Objective To develop a nomogram for predicting the risk of early peritoneal effusion in acute pancreatitis(AP)patients.Methods We retrospectively analyzed clinical data of AP patients admitted to the first affilated hospital of Xiamen university from January 2018 to November 2024,documenting the occurrence of ascites within 48 hours of admission.Patients were divided into groups based on early ascites occurrence.Candidate predictors including demographics,clinical indices and laboratory markers were collected.Multivariate Logistic regression was used to identify independent risk factors of early ascites,which were incorporated into a nomogram,and the model's discrimination was evaluated by concordance index and area under the receiver operating characteristic curve(AUC),and calibration was assessed with calibration plots.Results Ascites occurred in 59 of 415 patients(14.2％).Multivariate analysis demonstrated that elevated C-reactive protein(CRP)(OR=1.007 6,95％CI 1.003 5-1.011 6)and reduced serum calcium(Ca2+)(OR=0.017 6,95％CI 0.002 8-0.114 5)were independent risk factors for early ascites.The nomogram achieved an AUC of 0.832(95％CI 0.782-0.884)with good calibration(Hosmer-Lemeshow P=0.709).Decision curve analysis(DCA)and clinical impact curve(CIC)indicated practical clinical value of the model.Conclusions The CRP-and Ca2+-based nomogram reliably predicts the risk of early peritoneal effusion in AP patients,which may facilitate early identification of high-risk patients and prompt intervention to improve outcomes.