高甘油三酯血症性胰腺炎临床特征及重症化预测模型的建立

Clinical features of hypertriglyceridemic acute pancreatitis and establish of a clinical prediction model for severity

目的:探讨高甘油三酯血症性胰腺炎(HTG-AP)的临床特征,分析与HTG-AP重症化相关的危险因素,并建立临床预测模型。方法:分析2019年1月至2022年12月天津市中西医结合医院肝胆胰外科收治的292例HTG-AP患者的临床资料,根据亚特兰大严重程度分级将其分为轻症胰腺炎(MAP)96例、中度重症胰腺炎(MSAP)53例、重症胰腺炎(SAP)47例,比较3组患者的一般特征、临床资料、临床转归,采用多因素Logistic回归分析HTG-AP重症化的危险因素,并对模型进行检验。结果:MAP、MSAP、SAP组在APACHEⅡ评分、血糖、血淀粉酶、C反应蛋白、白蛋白、肌酐、血小板、血红蛋白、D-二聚体差异有显著性(均P<0.05);在急性胰周液体聚集、胰周感染坏死、包裹性坏死、消化道出血、胸腔积液、腹腔积液、盆腔积液、MODS、SIRS、ARDS、休克、住院时间与住院花费也具有显著差异(均P<0.05)。多因素Logistic回归分析发现,血糖(OR=1.102)、C反应蛋白(OR=1.004)、D-二聚体(OR=1.411)为HTG-AP重症化的独立危险因素,并由此得到模型,对模型进行检验,绘制受试者工作特征曲线,计算曲线下面积为0.814,该模型具有较高的鉴别力。Hosmer-Lemeshow检验的P值为0.697,说明列线图在校准方面具有良好的性能。最后在决策曲线分析(DCA)中展现了模型的临床适用性。结论:血糖、C反应蛋白、D-二聚体是HTG-AP重症化的独立危险因素,研发的列线图表现出良好的区分度、校准度和临床适用性。

Objective:To explore the clinical characteristics of hypertriglyceridemic acute pancreatitis(HTG-AP)and analyze the risk factors associated with the severity of HTG-AP.To establish a clinical prediction model.Methods:The clinical data of 292 HTG-AP patients admitted to the Department of Hepatobiliary and Pancreatic Surgery,Tianjin Hospital of Integrated Traditional Chinese and Western Medicine,from January 2019 to December 2022 were analyzed,and they were classified into 96 cases of mild pancreatitis(MAP),53 cases of moderate severe pancreatitis(MSAP),and 47 cases of severe pancreatitis(SAP)according to the Atlanta severity grading.The three groups of patients with their general characteristics,clinical data,and clinical regression were compared.The risk factors of HTG-AP severity were analyed using multifactorial Logistic regression and the model was tested.Results:The MAP,MSAP,and SAP groups had significant differences in APACHEⅡscores,blood glucose,blood amylase,C-reactive protein,albumin,creatinine,platelets,hemoglobin,and D-dimer(all P<0.05),and in acute peripancreatic fluid aggregation,peripancreatic infected necrosis,encapsulated necrosis,gastrointestinal hemorrhage,pleural effusion,peritoneal fluid,pelvic effusion,MODS,SIRS,ARDS,shock,hospitalization time,and hospitalization cost were also significantly different(all P<0.05).Backward stepwise multifactorial Logistic regression analysis revealed that blood glucose(OR=1.102),C-reactive protein(OR=1.004),and D-dimer(OR=1.411)were the independent risk factors for HTG-AP recidivism,and a model was obtained from this.The model was tested by plotting ROC curves and the area under the curve was 0.814,and the model had a high discriminatory power.The P-value of the Hosmer-Lemeshow test was 0.697,which indicated that the column line graph had good performance in calibration.Finally,the clinical applicability of the model was demonstrated in decision curve analysis(DCA).Conclusion:Blood glucose,C-reactive protein,and D-dimer are independent risk factors for HTG-AP severity.The developed nomogram shows good discrimination,calibration and clinical applicability.