Establishment and predictive efficiency of model for the effect and safety of endoscopic lithotripsy on pediatric upper urinary tract calculi
10.3760/cma.j.cn112330-20191216-00557
- VernacularTitle:儿童上尿路结石腔内手术治疗有效性和安全性预测模型的构建及其预测效能
- Author:
Yu ZHANG
1
;
Fangzhou ZHAO
;
Jun LI
;
Lei TANG
;
Chunming LI
;
Jianwei JIAO
;
Ye TIAN
;
Wenying WANG
Author Information
1. 首都医科大学附属北京友谊医院泌尿外科,北京 100050
- From:
Chinese Journal of Urology
2020;41(9):681-686
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To establish a predictive model that can predict the effectiveness and safety of endoscopic surgery for pediatric upper urinary tract calculi, and evaluate its diagnostic efficacy through internal validation.Methods:The data was selected from the prospective database of pediatric upper tract calculi constructed by Beijing Friendship Hospital Affiliated to Capital Medical University from June 2014 to April 2019. A total of 348 children were recruited in this investigation including 250 boys and 98 girls, with a median age of 3.0 years(Interquartile Range 1.4, 7.0 years). Totally 375 endoscopic surgeries were performed, with an overall stone free rate (SFR) of 88.0% (330/375) and complication rate (CR) of 23.2% (87/375). This research used univariate and multivariate logistic regression analysis to evaluate and screen the predictors of SFR and CR. The nomogram of SFR and CR was established by using the selected predictive factors. The differentiation degree of the model was evaluated by the area under the curve (AUC), the consistency between the prediction probability and the actual risk was evaluated by the calibration curve, and the clinical benefits of the model application were assessed by the decision curve.Results:The results of multivariate analysis demonstrated that stone burden, operation duration, intraoperative perfusion, stone location and operative options were the postoperative predictors of SFR and CR for children. Besides, BMI was also a predictor of postoperative CR. The AUC of the model (model 1: SFR; model 2: CR) based on the above predictive factors was 0.81 and 0.73, respectively. The predictive probability of the calibration curve showed that the model had a good consistency with the actual value. The decision curve showed that the application of model 1 to predict SFR had significant clinical benefits when the threshold value was greater than 20%, and the application of model 2 to predict the postoperative CR had a significant clinical benefit when the threshold was greater than 10%.Conclusions:Nomogram based on stone burden, operation time, intraoperative perfusion, stone location and operative options was validated internally with preferable predictive power on the effectiveness and safety of pediatric endoscopic surgery. This model can be used to predict the clinical effects of pediatric endoscopic lithotripsy. BMI was also an important factor for the safety of pediatric endoscopic procedures.