Establishment and validation of predictive model for postoperative pulmonary complications in patients undergoing robot-assisted laparoscopic urological surgery
10.3760/cma.j.cn131073-20250618-00904
- VernacularTitle:机器人辅助腹腔镜泌尿外科手术患者PPCs预测模型的构建和验证
- Author:
Baoli CHENG
1
;
Yumeng FU
;
Shuting YANG
;
Yan WANG
;
Dan XIA
;
Shilong WEI
;
Qianqian ZHAO
;
Yongqian YUAN
Author Information
1. 浙江大学医学院附属第一医院麻醉科，杭州 311100
- Publication Type:Journal Article
- Keywords:
Postoperative complications;
Robotic surgical procedures;
Laparoscopy;
Urologic surgical procedures;
Postoperative pulmonary complications;
Prediction mode
- From:
Chinese Journal of Anesthesiology
2025;45(9):1104-1109
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To construct and validate a predictive model for postoperative pulmonary complications (PPCs) in patients undergoing robot-assisted laparoscopic urological surgery.Methods:This retrospective study included the medical records of 932 patients who underwent robot-assisted laparoscopic urological surgery at the First Affiliated Hospital of Zhejiang University School of Medicine from January 2020 to February 2022. The patients were divided into a training group ( n=559) and a validation group ( n=373) at a 6∶4 ratio. Logistic regression analysis was used to determine the independent risk factors for PPCs, and a nomogram prediction model was constructed based on these factors. The performance of the model was evaluated using the receiver operating characteristic curve and calibration curve, and the clinical benefit was assessed using the clinical decision curve analysis. Results:The independent risk factors for PPCs included advanced age (>60 yr), smoking history, respiratory tract infection within 1 month, preoperative low SpO 2 (<96%), and prolonged length of postoperative hospital stay ( P<0.05), and the body mass index (18.5-<28.0 kg/m 2) was a protective factor. The nomogram prediction model developed based on the aforementioned 6 influencing factors had an area under the receiver operating characteristic curve of 0.81 (95% confidence interval 0.76-0.86) in training group and 0.80 (95% confidence interval 0.75-0.86) in validation group. The calibration curve indicated a good consistency between the predicted and actual occurrence curves, and the clinical decision curve analysis showed good accuracy and net benefit of the prediction model. Conclusions:The predictive model for PPCs is successfully constructed based on age, low body mass index, smoking history, history of respiratory tract infection within 1 month, preoperative low SpO 2 and prolonged length of postoperative hospital stay and has good predictive performance in patients undergoing robot-assisted laparoscopic urological surgery.
