关键词: 重症急性胰腺炎； 左侧门静脉高压症； 危险因素； 列线图预测模型

Abstract: Objective To investigate the risk factors of severe pancreatitis (SAP) and left portal hypertension (LPH),a nomogram prediction model was constructed based on the risk factors and verified.Methods The clinical data of 227 SAP patients admitted to our hospital from January 1,2019 to June 30,2022 were retrospectively collected,and the incidence of LPH was statistically analyzed.Multivariate Logistic regression was used to analyze the risk factors of LPH in SAP patients.Construct a Nomogram prediction model and evaluate its performance by C-index,calibration graph and decision curve analysis.Results During the follow-up period,66 patients were diagnosed with LPH,the incidence was about 29.07%.The proportions of men in LPH group,BMI≥28kg/m²,smoking history,hypertriglyceridemia,lesion location in pancreatic tail,splenic vein stenosis,diabetes mellitus and CT severity index（CTSI） score 9-10 points were higher than those in non-LPH group (P<0.05).Splenic vein stenosis,BMI≥28kg/m²,hypertriglyceridemia,diabetes mellitus and CTSI score 9-10 points were the risk factors for LPH in SAP patients (P<0.05).A nomogram prediction model based on Logistic regression risk factors was constructed,and it was verified that the prediction model showed good discrimination (area under the curve 0.892,P<0.05) and calibration (mean absolute error=0.029).Decision curve analysis showed that the prediction model could provide high clinical benefit.Conclusion Splenic vein stenosis,BMI≥28 kg/m²,hypertriglyceridemia,diabetes mellitus,acute fluid or necrotic accumulation are risk factors for SAP complicated with LPH.The prediction model based on risk factors has high predictive efficacy for LPH.

Key words: severe pancreatitis； left-sided portal hypertension； risk factors; nomogram prediction model