结合盆底彩超测量数据及相关的临床资料，构建压力性尿失禁的多因素Logistic回归模型，探讨压力性尿失禁发病的危险因素。

选取2022年10月至2023年10月所有就诊于中山大学附属第三医院泌尿外科门诊需要完善盆底彩超检查的女性患者238例，其中压力性尿失禁组123例，无压力性尿失禁组115例。收集所有患者的临床资料以及盆底彩超测量数据。采用t检验或卡方检验进行差异性分析筛选压力性尿失禁的危险因素，运用多因素Logistic回归进一步筛选压力性尿失禁的独立危险因素，根据独立危险因素构建压力性尿失禁的预测模型并绘制该模型的受试者工作特征曲线加以验证。此外，根据尿失禁咨询委员会国际评分标准（ICI-Q-SF）将压力性尿失禁患者分为轻、中、重度三组，利用卡方检验对三组患者的盆底独立危险因素进行差异性分析。

多因素Logistic回归显示，患者BMI≥24 kg/m²、有阴道分娩史、分娩胎儿体质量大、膀胱颈移动度增大、尿道内口漏斗形成、肛提肌裂孔扩张、膀胱膨出及会阴体运动过度与压力性尿失禁的发生密切相关，差异具有统计学意义（P<0.05）。基于多因素分析结果，构建出压力性尿失禁的预测模型，ROC曲线下面积为0.928（95%CI：0.890～10.966，P<0.001），模型的最佳截断值为0.382，灵敏度为92.0%，特异度为84.3%。并且，与轻度压力性尿失禁相比，中、重度压力性尿失禁患者膀胱颈移动度增大率、尿道内口漏斗的形成率、肛提肌裂孔扩张率、膀胱膨出率及会阴体过度运动率更高，差异有统计学意义（P<0.05）。

临床资料结合盆底彩超构建的压力性尿失禁预测模型，可以为临床筛查女性压力性尿失禁高危人群提供简便、科学的工具。盆底彩超数据有望成为压力性尿失禁分度的客观依据。

To explore the risk factors of stress incontinence with pelvic floor ultrasound data and related clinical data.

A total of 238 female patients who were admitted to Urology Department of the Third Affiliated Hospital of Sun Yat-sen University from October 2022 to October 2023 in need of perfect pelvic floor ultrasound examination were selected, including 123 in the stress incontinence group and 115 in the non-stress incontinence group. Clinical data and pelvic floor ultrasound data of all patients were collected. T test or Chi-square test were used for difference analysis to screen the risk factors of stress urinary incontinence. Secondly, multiple Logistic regression was used to further screen the independent risk factors of stress urinary incontinence. A prediction model of stress urinary incontinence was constructed according to the independent risk factors and the receiver operator characteristic curve of the model was drawn for verification. In addition, patients with stress urinary incontinence were divided into mild, moderate and severe groups according to the Incontinence Advisory Committee International Scoring Criteria (ICI Q-SF), and Chi-square test was used to analyze the differences in independent risk factors of pelvic floor among the three groups.

Multivariate Logistic regression showed that BMI≥24 kg/m², history of vaginal delivery, large fetal weight during delivery, increased bladder neck mobility, infundibular formation of internal urethral orifice, expansion of anal divider, cystocele and excessive movement of perineal body were closely related to the occurrence of stress incontinence, with statistical significance (P<0.05). Based on the results of multi-factor analysis, a prediction model for stress urinary incontinence was constructed. The area under ROC curve was 0.928(95%CI: 0.890-10.966, P<0.001), the optimal cut-off value of the model was 0.382, the sensitivity was 92.0%, and the specificity was 84.3%. In addition, compared with mild stress incontinence, the increased rate of bladder neck movement, the formation rate of inner urethral orifice funnel, the expansion rate of anal levator hiatus, the cystocele rate and the overmovement rate of perineal body were higher in patients with moderate and severe stress incontinence, and the differences were statistically significant (P<0.05).

The prediction model of stress urinary incontinence constructed by the combination of clinical data and pelvic floor ultrasound can provide a simple and scientific tool for clinical screening of females at high risk of stress urinary incontinence. The ultrasound data of pelvic floor may be the objective reference for the classification of stress urinary incontinence.