Characteristics Predictive of Successful Pelvic Floor Muscle Training Outcomes Among Women with Stress Urinary Incontinence

Abstract

Physiotherapist-supervised pelvic floor muscle (PFM) training is the recommended first-line treatment for women with stress urinary incontinence (SUI); however, only up to 50% of women are cured with this treatment. The primary objective of this study was to develop a predictive model of successful physiotherapy intervention outcomes among women with SUI using predictors that are accessible to clinicians. The secondary objective of this study was to run a preliminary investigation of morphologic differences between women with SUI who were cured with a physiotherapy intervention and those who were not cured, using a subset of ultrasound imaging data. Seventy-nine women with SUI were assessed at baseline on measures of demographic data (i.e. age, body mass index, etc.), 3-day bladder diary, 30-minute standardized pad test, clinical assessments of PFM strength and tone, and transperineal ultrasound assessments of PFM morphology. Women then attended a 12-week physiotherapy intervention and returned for a follow-up assessment. The multivariate logistic regression model was significant (p < .001) with two predictors: baseline ICIQ-FLUTS UI subscale (SUI severity; p = .01) and parity (p = .06). A significant ROC curve for the ICIQ-FLUTS UI subscale (p ˂ .01) predicts physiotherapy intervention outcomes with 55.6% sensitivity and 80.8% specificity at a cut-off score of 7.50. Women most likely to be cured with a physiotherapy intervention were those with lower scores on the ICIQ-FLUTS UI subscale and those who have given birth to fewer children. Significant differences were found in a subset of data between women with SUI who are cured with the physiotherapy intervention and women who are not cured on morphologic measures in standing of bladder neck height at peak cough (p = .03), descent of the bladder neck during maximal Valsalva maneuver (MVM; p = .04), levator hiatus circumference at rest (p = .03) and at maximal voluntary contraction in both standing and supine (MVC; p = .01; p = .03). Variables that were trending towards significance included bladder neck height in standing at rest, levator plate length (LPL) at maximal excursion during a cough and MVM in standing, and mid-urethral wall cross-sectional area. These significant differences indicate potential value in using ultrasound imaging outcomes as predictors of a cure with physiotherapy intervention in future models, and a combination of demographic, clinical, and morphologic variables may build a more robust predictive model.