重度急性胰腺炎并发胰性脑病早期预测模型的建立和评价  

Establishment and Evaluation of A Early Prediction Model for Severe Acute Pancreatitis Complicated With Pancreatic Encephalopathy

作　　者：伏添 栾正刚[2] 张谞丰[3] FU Tian;LUAN Zhenggang;ZHANG Xufeng(Department of Intensive Care Unit, Hanzhong 3201 Hospital, Hanzhong, Shaanxi Province 723000;Department of Intensive Care Unit, the First Hospital of China Medical University, Shenyang;Department of Hepatobiliary Surgery, the First Affiliated Hospital of Xi’an Jiaotong University, Xi’an)

机构地区：[1]陕西省汉中市三二O一医院重症医学科,723000 [2]中国医科大学附属第一医院重症医学科 [3]西安交通大学第一附属医院肝胆外科

出　　处：《胃肠病学》2020年第12期740-744,共5页Chinese Journal of Gastroenterology

基　　金：辽宁省自然基金资助计划(2019-MS-09);西安交通大学第一附属医院临床研究重点项目(XJTU1AF-CRF-2017-004)。

摘　　要：背景:胰性脑病(PE)是重度急性胰腺炎(SAP)的严重全身并发症之一,近年来,PE的发病率呈上升趋势。然而,早期预测SAP并发PE的工具较少。目的:从SAP患者临床检验指标和评分系统中筛选出PE的早期独立危险因素,进而构建PE早期预测评分模型并提前干预。方法:回顾性分析2016年1月—2020年9月陕西省汉中市三二〇一医院收治的130例SAP患者临床资料,采用单因素和多因素Logistic回归分析筛选出PE的早期独立危险因素,并采用加权最小二乘法构建预测评分模型。结果:单因素分析结果显示,酗酒史、乳酸、腹腔内压力(IAP)、CT严重指数(CTSI)、腹部CT胰腺外炎症评分(EPIC)、格拉斯哥昏迷评分(GCS)与PE的发生相关(P<0.05),多因素Logistic回归分析显示,酗酒史(OR=2.843,95%CI:1.759~4.595,P=0.011)、IAP(OR=1.077,95%CI:1.020~1.138,P=0.007)、EPIC评分(OR=1.768,95%CI:1.181~2.649,P=0.006)为PE的早期独立危险因素。根据构建的早期预测评分模型,将PE风险分为低危(0~3分)、中危(4~6分)、高危(>6分),三组SAP患者的PE发生率差异有统计学意义(P<0.05)。结论:构建的预测评分模型对SAP并发PE具有早期预测和评估的价值,对危险进行分层有助于提前采取干预措施降低PE的发生率。Background:Pancreatic encephalopathy(PE)is one of the severe systemic complications of severe acute pancreatitis(SAP).In recent years,the incidence of PE was on the rise.There are few tools for early prediction of SAP complicated with PE.Aims:To screen the early independent risk factors of PE from clinical testing indices and scoring system of SAP patients,and then construct an early predictive scoring model of PE and used for intervening in advance.Methods:The clinical data of 130 patients with SAP from Jan.2016 to Sept.2020 at Shaanxi Hanzhong 3201 Hospital were analyzed retrospectively.Early independent risk factors of PE was screened by univariate analysis and multivariate Logistic regression analysis.The predictive scoring model was constructed by the weighted least square method.Results:Univariate analysis showed that history of alcohol abuse,lactic acid,intra-abdominal pressure(IAP),CT severity index(CTSI),extrapancreatic inflammation on CT(EPIC)and Glasgow coma scale(GCS)score were correlated to PE(P<0.05).Multivariate Logistic regression analysis showed that history of alcohol abuse(OR=2.843,95%CI:1.759-4.595,P=0.011),IAP(OR=1.077,95%CI:1.020-1.138,P=0.007),and EPIC score(OR=1.768,95%CI:1.181-2.649,P=0.006)were independent risk factors for PE in the early stage.According to the early predictive scoring model constructed,risk of PE was divided into low risk(0-3),medium risk(4-6)and high risk(>6),and differences in the incidence of PE in SAP patients among the three groups were statistically significant(P<0.05).Conclusions:The predictive scoring model constructed has the value for early prediction and evaluation of SAP complicated with PE,and risk stratification is helpful for taking intervention measures in advance to reduce the incidence of PE.

关 键 词：重度急性胰腺炎 胰性脑病 早期预测模型 危险因素 

分 类 号：R47[医药卫生—护理学]