基于盆底超声参数构建的列线图模型预测盆底重建术后进展性压力性尿失禁的价值

Value of nomogram model based on pelvic floor ultrasound parameters in predicting progressive stress urinary incontinence after pelvic floor reconstruction

目的 基于术前盆底超声参数构建列线图模型,探讨其预测盆底脏器脱垂(POP)患者盆底重建术后发生进展性压力性尿失禁(SUI)的价值。方法 选取2019年5月至2022年4月我院因国际泌尿科学会POP-Q分度≥Ⅱ度行盆底重建术女性患者495例建立预测模型作为模型组,其中术前POP合并SUI 295例;根据术前、术后SUI发生情况分为术后进展组76例(包括新发46例,加重30例)与SUI改善组219例;对模型组样本采取Bootstrap法重复抽样1000次作为内部验证组。应用单因素分析、多因素Logistic回归分析POP患者盆底重建术后发生进展性SUI的独立危险因素,并构建列线图模型。绘制受试者工作特征曲线、校准曲线、临床决策曲线、临床影响曲线对该模型区分度、校准度及临床适用性进行验证。结果 (1)单因素分析显示,SUI改善组与术后进展组孕次、产次、Ba点及指压诱发试验阳性、高血压史、糖尿病史、术前混合性尿失禁史、联合尿道中段无张力悬吊术(TVT-O)占比,以及盆底超声所测残余尿量(PVR)、尿道旋转角(Uret.rot)、尿道内口呈漏斗样改变、肛提肌损伤比较差异均有统计学意义(均P<0.05);多因素Logistic回归分析显示,指压诱发试验阳性、Ba点,以及盆底超声所测尿道内口呈漏斗样改变及肛提肌损伤均为术后发生进展性SUI的独立危险因素(OR=6.974、1.400、3.651、11.137,均P<0.05);(2)受试者工作特征曲线分析显示,列线图模型预测模型组术后发生进展性SUI的曲线下面积为0.885(95%可信区间:0.737~0.936),预测内部验证组发生进展性SUI的曲线下面积为0.823(95%可信区间:0.800~0.849),提示模型区分度良好;(3)校准曲线显示预测曲线与理想曲线较为贴合,Brier score为0.185,校准曲线的平均绝对误差为0.026,提示模型具有较好的准确性与一致性;临床决策曲线和临床影响曲线显示,模型组与内部验证组分别在阈概率值为0.07~0.84与0.11~0.67时临床最大净收益较高,提示模型具有较好的临床适用性。结论 基于术前盆底超声参数构建的列线图模型在预测POP患者盆底重建术后发生进展性SUI中有一定的价值,可为临床术前个体化诊治提供参考。

ObjectiveTo explore the value of nomogram model based on preoperative pelvic floor ultrasound parameters in predicting the occurrence of progressive urinary incontinence(SUI) after pelvic floor reconstruction in patients with pelvic organ prolapse(POP).MethodsFrom May 2019 to April 2022,495 patients who underwent pelvic floor reconstruction surgery due to ICS-POP Q score≥Ⅱ degree POP in our hospital were selected to establish a prediction model as the model group,among whom 295 cases of preoperative POP accompanied by SUI.According to the occurrence of SUI before and after surgery,76 patients in postoperative progress group(including 46 cases of new occurrence,30 cases of aggravation),and 219 cases in SUI improvement group.The model group was internally validated by repeated sampling 1000 times by the Bootstrap method as verification group.Univariate analysis and multivariate Logistic regression were used to analyze the independent risk factors for SUI after pelvic floor reconstruction surgery in POP patients,and a nomogram was constructed.Receiver operating characteristic curve,calibration curve,clinical decision curve and clinical impact curve were drawn to verify the model's discrimination,calibration and clinical applicability.Results(1) Univariate analysis showed that there were statistically significant differences in pregnancy,parity,Ba point and positive acupressure test,history of hypertension,history of diabetes,history of mixed urinary incontinence before operation,proportion of combined TVT-O operation,the residual urine volume(PVR),urethral rotation angle(Uret.rot),funnel-like change of the internal opening of the urethra,and levator ani muscle injury measured by pelvic floor ultrasound between the SUI improvement group and postoperative progress group(all P<0.05).Multivariate Logistic regression analysis showed that positive acupressure test,Ba point,and funnel-like changes of internal opening of the urethra,levator ani muscle injury measured by pelvic floor ultrasound were all independent risk factors for postoperative progressive SUI(OR=6.974,1.400,3.651,11.137,all P<0.05).(2)Receiver operating characteristic curve analysis showed that the area under the curve of the nomogram model for predicting postoperative progressive SUI in the model group was0.885(95% confidence interval:0.737~0.936),and the area under the curve of the verification group was 0.823(95% confidence interval:0.800~0.849),indicating that the model had good discrimination.(3)Calibration curve showed that the prediction curve of the model group was more consistent with the ideal curve,Brier score was 0.185,and the mean absolute error of the calibration curve was 0.026,indicating that the model had good accuracy and consistency.Clinical decision curve and clinical impact curve showed that the model group and verification group had higher clinical maximum net returns when the threshold probability values were 0.07~0.84 and 0.11~0.67,respectively,which had good clinical practicality within this threshold probability range.ConclusionThe nomogram model constructed based on preoperative pelvic floor ultrasound parameters has certain value in predicting SUI after pelvic floor reconstruction in POP patients,which can provide reference for clinical preoperative individualized diagnosis and treatment.