良性前列腺增生TURP术后复发的风险预测列线图模型构建与验证

Construction and Validation of a Line Graph Model for Predicting the Risk of Recurrence after TURP for Benign Prostatic Hyperplasia

目的构建并验证良性前列腺增生(BPH)经尿道前列腺电切(TURP)术后复发的风险预测列线图模型。方法回顾性分析医院2013年1月~2022年3月收治的1924例BPH患者的临床资料,均实施TURP治疗,且术后随访至少1年,根据是否复发分为复发组(n=61)和未复发组(n=1863)。Cox回归分析探讨患者TURP术后复发的影响因素;基于该结果建立风险预测列线图模型;Bootstrap法进行内部验证,决策曲线分析(DCA)探讨该模型的净收益,受试者工作特征(ROC)曲线分析该模型对复发的预测价值。结果复发率为3.17%(61/1924);发病年龄(HR=0.750,95%CI:0.489~0.851)、残余尿量(HR=0.755,95%CI:0.541~0.884)、病理标本炎症分级(HR=3.126,95%CI:2.765~3.988)、术后症状完全消失(HR=0.612,95%CI:0.412~0.896)、吸烟(HR=2.238,95%CI:1.755~2.967)、憋尿(HR=3.182,95%CI:1.828~3.799)均是术后复发的影响因素(P<0.05);基于危险因素分析结果建立TURP术后复发的风险预测列线图模型;Bootstrap法验证C-index分数为0.852(95%CI:0.818~0.883),风险预测列线图概率阈值为0.03~0.64时临床净收益率最高;术后复发的风险预测列线图模型预测患者复发的灵敏度、特异度和曲线下面积(AUC)分别为:86.89%、84.93%、0.901(95%CI:0.853~0.938)。结论发病年龄、残余尿量、病理标本炎症分级、术后症状完全消失、吸烟和憋尿均是BPH患者TURP术后复发的危险因素,基于此构建复发风险预测列线图模型可准确预测该不良事件发生风险。

Objective To construct and validate a line chart model for predicting the risk of recurrence after transurethral resection of the prostate(TURP)in patients with benign prostatic hyperplasia(BPH).Methods A retrospective analysis was conducted on the clinical data of 1924 patients with BPH admitted to the hospital from January 2013 to March 2022.All patients were underwent TURP treatment and were followed up for at least 1 year after surgery.They were divided into a recurrent group(n=61)and a non recurrent group(n=1863)based on whether there was recurrence.Cox regression analysis was used to explore the influencing factors of postoperative recurrence in patients undergoing TURP.A risk prediction Nomogram model based on this result was established,internal validation was carried put using Bootstrap method,and decision curve analysis(DCA)explored the net return rate of the model.The receiver operating characteristic(ROC)curve and the predictive value of this model for recurrence were analyzed.Results The recurrence rate was 3.17%(61/1924).Age of onset(HR=0.750,95%CI:0.489~0.851),residual urine volume(HR=0.755,95%CI:0.541~0.884),inflammation grade of pathological specimens(HR=3.126,95%CI:2.765~3.988),complete disappearance of postoperative symptoms(HR=0.612,95%CI:0.412~0.896),smoking(HR=2.238,95%CI:1.755~2.967),and holding urine(HR=3.182,95%CI:1.828~3.799)are all influencing factors for postoperative recurrence(P<0.05).A risk prediction column chart model for postoperative recurrence of TURP was established based on the results of risk factor analysis.The Bootstrap method verified that the C-index score was 0.852(95%CI:0.818~0.883),and the clinical net return rate was the highest when the probability threshold of the risk prediction column chart was 0.03-0.64.The sensitivity,specificity,and area under curve(AUC)of the line graph model for predicting postoperative recurrence in patients were 86.89%,84.93%,and 0.901,respectively(95%CI:0.853~0.938).Conclusion Age of onset,residual urine volume,inflammation grade of pathological specimens,complete disappearance of postoperative symptoms,smoking,and holding urine are all risk factors for postoperative recurrence in BPH patients undergoing TURP.Based on this,constructing a recurrence risk prediction column chart model can accurately predict the risk of adverse events.