Unsupervised clustering analysis of treatment strategies for elite female athletes with severe stress urinary incontinence: focusing on competition return and SUI improvement.

Abstract

Purpose: To evaluate two primary outcomes in elite female athletes (EFAs) with severe stress urinary incontinence (SUI) 24 months post-intervention: return to elite-level competition and improvement in SUI symptoms. Clustering analysis was conducted to identify subgroups within the patient population and explore treatment efficacy.

Methods: A retrospective analysis was performed on 183 EFAs with severe SUI who underwent treatments including pelvic floor muscle training (PFMT), vaginal and urethral erbium laser (Fotona Laser), and mid-urethral sling (MUS) surgery. Clustering analysis was conducted using K-means to categorize patients, followed by multivariate regression and Random Forest to determine predictors.

Results: Three distinct clusters were identified. PFMT frequency was the most significant predictor of both return to sports and SUI improvement across clusters. Cluster 0, characterized by younger participants, required interventions like MUS surgery and Fotona Laser for significant improvement. Cluster 2, with high PFMT adherence, showed the best improvement in pad test results (4.9 g) and the highest return to sports rate (85.9%). Fotona Laser was particularly effective in Cluster 2, with 91.3% of patients returning to sports within one year. In contrast, Cluster 1, which included older participants with more severe symptoms, demonstrated the least improvement and lowest return to sports rate (2.8%), likely due to lower PFMT frequency and inconsistent training.

Conclusion: Clustering analysis effectively categorized EFAs with SUI, highlighting the critical role of personalized, intensive PFMT in achieving sports return and symptom improvement. Future research should validate findings in larger cohorts and integrate machine learning to refine personalized medicine.