预测粘连性小肠梗阻患者手术治疗的危险因素判别分析  

Discriminant analysis on the risk factors for predictingthe surgery requirement in patients with adhesive small bowel obstruction

作　　者：任龙 张云[1] 张淼[1] REN Long;ZHANG Yun;ZHANG Miao(Department of General Surgery,the Affiliated Yixing Hospital of Jiangsu University,214200 Yixing,Jiangsu China)

机构地区：[1]江苏大学附属宜兴医院普外科,江苏宜兴214200

出　　处：《临床消化病杂志》2022年第4期269-272,共4页Chinese Journal of Clinical Gastroenterology

摘　　要：[目的]建立可以预测粘连性小肠梗阻手术治疗的Logistic回归分析模型,提供临床治疗参考。[方法]回顾性分析184例粘连性小肠梗阻患者临床资料,通过统计学方法研究手术治疗相关的危险因素,选取有统计学意义的因素完成多因素Logistic回归,建立Logistic回归预测模型,预判粘连性小肠梗阻病例采取手术治疗方案的特异性、准确度、敏感性。[结果]184例中手术治疗组86例(46.74%),非手术治疗组98例(53.26%)。在单因素分析中13个危险因素组间比较差异有统计学意义,多因素回归分析显示,腹膜炎、腹痛持续、CT提示肠壁增厚水肿、非手术治疗时间、白细胞计数是预测手术的独立因素,并纳入Logistic回归预测模型。模型预测准确度为87%(160/184),敏感度为89.74%(70/78),特异度为91.84%(90/98)[结论]腹膜炎、腹痛持续、CT提示肠壁增厚水肿、非手术治疗时间、白细胞计数是粘连性小肠梗阻手术决策的独立因素,得出的Logistic回归模型可以较准确预判患者是否需要手术治疗。[Objective]To establish a logistic regression model to predict the surgical requirement of adhesive small bowel obstruction,and provide basis for clinical treatment.[Methods]The clinical data of patients with adhesive small bowel obstruction were analyzed retrospectively.Through statistical methods to study the risk factors related to the surgical treatment of adhesive small bowel obstruction,select the statistically significant factors to complete the multi factor Logistic regression,and establish the logistic regression prediction model to judge the specificity,accuracy and sensitivity of the surgical treatment of adhesive small bowel obstruction.[Results]A total of 184 adult patients with adhesive small bowel obstruction were included in the study.Among them,there were 86 patients(46.74%)in the surgical treatment group and 98 patients(53.26%)in the non-surgical treatment group.Single factor analysis showed that there was statistical significance among 13 risk factors in the groups.Multiple factors regression analysis showed that peritonitis,persistent abdominal pain,intestinal wall thickening and edema shown by CT findings,non-surgical treatment time,and leukocyte count were independent factors in predicting surgery,and the logistic regression prediction model was included.The accuracy of the model prediction was 87%(160/184),the sensitivity was 89.74%(70/78),and the specificity was 91.84%(90/98).[Conclusion]Peritonitis,persistent abdominal pain,intestinal wall thickening and edema shown by CT findings,non-surgical treatment time,and leukocyte count are independent factors in the surgical decision-making of adhesive small bowel obstruction.The established logistic regression model can better predict whether small intestinal obstruction requires surgical treatment.

关 键 词：粘连性小肠梗阻 手术 预测 危险因素 

分 类 号：R574.2[医药卫生—消化系统]