住院2型糖尿病患者合并尿路感染的危险因素及预测模型构建

Risk Factors and Prediction Model Construction of Urinary Tract Infection in Hospitalized Patients with Type 2 Diabetes Mellitus

目的分析住院2型糖尿病患者合并尿路感染的危险因素,构建列线图预测模型。方法选取2018年6月至2023年6月于某医院住院治疗的457例2型糖尿病患者为研究对象,以7∶3的比例随机分为模型集(n=320)和验证集(n=137);根据有无合并尿路感染,将模型集分为感染组(n=108)和未感染组(n=212)。比较模型集和验证集的临床资料,筛选患者尿路感染的危险因素,构建列线图风险预测模型。结果模型集中尿路感染的发生率为33.75%,未感染组与感染组在年龄、性别、病程、住院时间、抗生素使用、导尿管留置、并发症种类、空腹血糖(FPG)、糖化血红蛋白(HbA1c)、血清白蛋白(ALB)、淀粉样蛋白A(SAA)等方面比较,差异有统计学意义(P<0.05)。高龄、女性、病程长、留置导尿管、并发症种类多、血糖高、低蛋白血症和高水平SAA是住院2型糖尿病患者并发尿路感染的危险因素(P<0.05)。对建立的模型进行评估和验证,模型集的ROC曲线下面积(AUC)=0.889(95%CI:0.852~0.927),验证集的AUC=0.895(95%CI:0.842~0.948),模型集、验证集的预测发生率与实际发生率接近,均靠近理想曲线,准确度和一致性良好。结论高龄、病程长、血糖高、低蛋白血症和高水平SAA等是住院2型糖尿病患者并发尿路感染的危险因素,构建的列线图预测模型效果良好,可预防尿路感染的发生。

Objective To analyze the risk factors for urinary tract infection(UTI)in hospitalized patients with type 2 diabetes mellitus(T2DM)and construct a nomogram prediction model.Methods A total of 457 patients with T2DM who were hospitalized in a hospital from June 2018 to June 2023 were selected as the research objects.They were randomly divided into a model set(n=320)and a verification set(n=137)at a ratio of 7∶3.And patients in the model set were further divided into an infected group(n=108)and an uninfected group(n=212)based on the presence or absence of UTI.The clinical data were compared between the model set and the verification set.The risk factors for UTI in patients were screened.And a nomogram risk prediction model was constructed.Results The incidence of UTI in the model set was 33.75%.There were statistically significant differences between the uninfected group and the infected group in terms of age,gender,disease course,length of hospital stay,antibiotic use,catheter indwelling,types of complications,fasting plasma glucose(FPG),glycosylated hemoglobin(HbA1c),serum albumin(ALB),serum amyloid A(SAA),etc(P<0.05).Advanced age,female,long course of T2DM,catheter indwelling,multiple complications,hyperglycemia,hypoproteinemia and high level of SAA were the risk factors for UTI in hospitalized T2DM patients(P<0.05).The model established was evaluated and verified,with AUC=0.889(95%CI=0.852-0.927)for the model set and AUC=0.895(95%CI=0.842-0.948)for the verification set.The predicted incidence of the model set amd the verification set was close to the actual incidence and the ideal curve,indicating good accuracy and consistency.Conclusion Advanced age,long course of T2DM,hyperglycemia,hypoproteinemia and high level of SAA are the risk factors for UTI in hospitalized patients with T2DM.The nomogram prediction model constructed in the study is effective and can be applied to prevention of UTI.