Construction and verification of prediction nomogram model for urinary tract infection in patients with type 2 diabetes
Objective To construction prediction nomogram model for urinary tract infection(UTI)in patients with type 2 diabetes(T2D),thus providing reference for early diagnosis for UTI.Methods This was a retrospective study involving 310 T2D patients,the patients were allocated into the test group(n=217)and the validation group(n=93).Univariate and multivariate logistic regression analyses were implemented to analyze UTI risk factors,the nomogram for risk factors was constructed for verifying the accuracy and reliability.Results Among 310 T2D patients,there were 63(20.3％)occured UTI,with UTI incidence rate of 18.9％(41/217)in the test group and 23.7％(22/93)in the validation group.The results from multivariate regression analysis demonstrated that body mass index,C-reactive protein,albumin,glycated hemoglobin were risk factors for UTI in T2D patients(P＜0.05).The verification results from prediction nomogram model showed that the C-index of test group and validation group were 0.971 and 0.917,with the area under the curve(AUC)of 0.816 and 0.788,the specificity of 80.28％and 71.59％,and the sensitivity of 77.27％and 75.61％,respectively.Conclusion A the prediction nomogram model based on the risk factors in T2D patients can predict the UTI risk,which is capable of helping medical staff to develop targeted management strategies.
type 2 diabetesurinary tract infectionrisk factornomogramprediction model
万方数据
维普
