Development and validation of a prediction model for bothersome stress urinary incontinence after prolapse surgery: a retrospective cohort study

Objective To develop and validate a prediction model for bothersome stress urinary incontinence after prolapse surgery and to compare it with an existing clinical prediction model (CUPIDO model). Design Retrospective cohort study. Setting Two tertiary hospitals in South Korea. Population A total of 1142 patients who underwent prolapse surgery with or without a concomitant midurethral sling. Methods To construct a prediction model, we performed logistic regression using both exhaustive and stepwise variable selection, validating the model both internally and externally. Main outcome measures Bothersome stress urinary incontinence defined as the presence of bothersome symptoms of stress urinary incontinence and/or subsequent continence procedure one year after surgery. Results Postoperative bothersome stress urinary incontinence occurred in 10% of patients. A model containing six predictors (age, diabetes mellitus, subjective urinary incontinence, prolapse reduction stress test result, type of prolapse surgery, and a concomitant midurethral sling) showed excellent performance for predicting bothersome stress urinary incontinence (area under the curve 0.74, 95% confidence interval 0.62-0.86) and outperformed the CUPIDO model (area under the curve 0.63, 95% confidence interval 0.49-0.76; DeLong's test P = 0.014). Conclusions This prediction model might be a useful tool to guide patient decision making regarding a concomitant continence procedure at the time of prolapse surgery. The predictive value of this model needs to be validated further in cohorts with different characteristics. Tweetable abstract The proposed prediction model for bothersome stress urinary incontinence after prolapse surgery outperforms an existing model.

Automated summary

Bothersome stress urinary incontinence occurred in 10% of patients after prolapse surgery. The prediction model containing six predictors showed superior performance compared to the existing CUPIDO model, suggesting its potential usefulness in guiding patient decision making. Further validation in diverse cohorts is necessary to confirm the predictive value of this model.