Analysis of risk factors and establishment of decision tree model for multi-drug resistant urinary tract infection in type 2 diabetes mellitus patients
10.3969/j.issn.1006-6187.2025.07.007
- VernacularTitle:2型糖尿病患者泌尿系统多重耐药菌感染危险因素分析及建立决策树模型的研究
- Author:
Suqiong ZHANG
1
;
Xianzhong WANG
;
Zhaoyan SHUAI
Author Information
1. 628000 广元市中心医院泌尿外科
- Publication Type:Journal Article
- Keywords:
Diabetes mellitus,type 2;
Multi-drug resistant bacteria infection of urinary system;
Influencing factors;
Decision tree
- From:
Chinese Journal of Diabetes
2025;33(7):512-517
- CountryChina
- Language:Chinese
-
Abstract:
Objective To analyze the risk factors of multi-drug resistant bacterial infection of urinary system in type 2 diabetes mellitus(T2DM)patients and establish a decision tree model.Methods A total of 371 patients with T2DM were selected and divided into infected(IF,n=47)group and non-infected(NIF,n=324)group.The influencing factors of multi-drug resistant urinary system infection were analyzed in patients with T2DM.IBM SPSS Modeler was used to establish a decision tree model for predicting multi-drug resistant urinary system infection.Results The infection rate of multidrug resistant bacteria in urinary system was 12.67%.Age,female proportion,DM duration,use of antibiotics before admission,length of hospital stay,and hemoglobin A1c(HbA1c)in IF group were higher than that in NIF group(P<0.05).Logistic regression analysis showed that age≥60 years old(OR 2.629,95%CI 1.349～5.125,P=0.005),woman(OR 3.044,95%CI 1.506～6.151,P=0.002),DM duration≥10 years(OR 3.320,95%CI 1.671～6.593,P=0.001),length of hospital stay>15 days(OR 2.406,95%CI 1.237～4.682,P=0.010)and HbA1c>8.02%(OR 2.363,95%CI 1.210～4.617,P=0.012)were influencing factors for multidrug-resistant bacteria infection of urinary system.Decision tree model showed that sex was the most important factor in urinary multidrug-resistant infections.Receiver operating characteristic curve results showed that the area under the curve value of Logistic regression model was slightly higher than that of decision tree(Z=1.465,P=0.143).Conclusions Age,sex,DM duration,length of hospital stay,and HbA1c are the influencing factors of T2DM patients complicated with urinary multi-drug resistant bacteria infection,the decision tree model established in this study has good risk prediction efficiency,which is helpful for early clinical identification of high-risk patients with multi-drug resistant bacteria infection of urinary system.
