Construction and validation of risk prediction model for postoperative uri-nary retention in patients with cervical cancer
Aim:To explore the risk factors of postoperative urinary retention in patients with cervical cancer and build a risk prediction model.Methods:A retrospective study was conducted to collect 1 968 patients with cervical cancer who were admitted to a class A tertiary hospital from February 2020 to July 2023.Patients were allocated into a training set(1 320 cases)and a validation set(648 cases)randomly.Logistic regression was used to analyze the influencing factors of urinary retention after cervical cancer surgery.According to the results,a risk prediction model was constructed and a nomo-gram was drawn.ROC curves,calibration curves,and decision curve analysis(DCA)were used to evaluate the predictive value of the model.Results;The risk of urinary retention was higher in patients who was older than 60 years,who had un-dergone extensive hysterectomy,suffered from urinary tract infection after surgery,using urinary catheter for more than 14 days after surgery,and used pain relief pump after surgery,and the OR(95％CI)was 1.759(1.225-2.526),5.557(3.386-9.121),2.503(1.857-3.374),8.909(6.224-12.754),1.521(1.041-2.223),respectively.In training and valida-tion sets,the area under the ROC curve(95％CI)was 0.831(0.806-0.855)and 0.822(0.785-0.859).The calibration curve was close to the ideal curve,while the DCA curve was far from the baseline,indicating good consistency between the predicted result and the actual outcome.Conclusion:The prediction model of urinary retention after cervical cancer surgery shows good predictive value.
cervical cancerpostoperative urinary retentionprediction model
万方数据
维普
