Establishment and validation of nomographic prediction model for urinary tract infection in patients with double-J tube indwelling after urological surgery
Objective To establish a clinical prediction model of urinary tract infection(UTI)in patients with indwelling double J stents after urological surgery to improve the diagnosis and treatment of such patients.Methods A total of 137 patients who underwent urological surgery and indwelling double J stents in our hospital from January 2018 to July 2022 were selected as the research subjects.After medical record screening,a total of 137 subjects were included in the study,and patients with UTI were included in the UTI group,and those without UTI were included in the non-UTI group.The gender,age,body mass index(BMI),indwelling time of double J stents,diabetes mellitus,his-tory of urinary tract infection,preoperative nutritional status score(COUNT),operation time,side of double J stents,and combined renal dysfunction were collected.The multi-factor Logistic regression model was established using the"RMS"function package of R-studio 4.0.2 software,and the line chart was plotted.The diagnostic value of the model and single factors for double J stent infection was evalu-ated using the receiver operating characteristic curve(ROC).Results There were no significant differ-ences in gender,age,BMI,diabetes mellitus,operation time,side of double J stents,and combined renal dysfunction between the UTI group and the non-UTI group(P>0.05).The indwelling time of double J stents,preoperative COUNT score,and proportion of urinary tract infection history in the UTI group were higher than those in the non-UTI group(P<0.05).Multivariate Logistic regression analysis showed that the indwelling time of double J stents,urinary tract infection history,and preop-erative COUNT score were independent risk factors for postoperative urinary tract infection after uro-logical surgery with indwelling double J stents(P<0.05).The internal validation of the prediction model line chart was performed using the Bootstrap method.The Hosmer-Lemeshow test showed that χ2=6.753 and P =0.325,indicating that the model had good calibration.ROC analysis showed that the AUCs of double J stent indwelling time,urinary tract infection history,and preoperative COUNT score for predicting postoperative UTI were 0.793,0.747,and 0.750,respectively.The AUC of the prediction model established by these three individual factors for predicting postoperative UTI was 0.925,indicating good predictive accuracy(P<0.05).Conclusion The prediction model based on doub-le J stent indwelling time,urinary tract infection history,and preoperative COUNT score has good predictive ability for the occurrence probability of urinary tract infection in patients with indwelling double J stents after urological surgery,which can be used to identify high-risk groups of urinary tract infection in the early stage.
万方数据
维普
