Ultrasound assessment of pelvic floor muscle contraction: reliability and development of an ultrasound-based contraction scale

Objectives To determine intra- and interrater reliability and agreement for ultrasound measurements of pelvic floor muscle contraction and to assess the correlation between ultrasound and vaginal palpation. We also aimed to develop an ultrasound scale for assessment of pelvic floor muscle contraction. Methods This was a cross-sectional study of 195 women scheduled for stress urinary incontinence (n = 65) or prolapse (n = 65) surgery or who were primigravid (n = 65). Pelvic floor muscle contraction was assessed by vaginal palpation using the Modified Oxford Scale (MOS) and by two- and three-dimensional (2D/3D) transperineal ultrasound. Proportional change in 2D and 3D levator hiatal anteroposterior (AP) diameter and 3D levator hiatal area between rest and contraction were used as measures of pelvic floor muscle contraction. One rater repeated all ultrasound measurements on stored volumes, which were used for intrarater reliability and agreement analysis, and three independent raters analyzed 60 ultrasound volumes for interrater reliability and agreement analysis. Reliability was assessed using the intraclass correlation coefficient (ICC) and agreement using Bland-Altman analysis. Tomographic ultrasound was used to identify women with major levator injury. Spearman's rank correlation coefficient (r(S)) was used to assess the correlation between ultrasound measurements of pelvic floor muscle contraction and MOS score. The proportion of women allocated to each category of muscle contraction (absent, weak, moderate or strong) by palpation was used to determine the cut-offs for the ultrasound scale. Results Intrarater ICC was 0.81 (95% CI, 0.74-0.85) for proportional change in 2D levator hiatal AP diameter. Interrater ICC was 0.82 (95% CI, 0.72-0.89) for proportional change in 2D AP diameter, 0.80 (95% CI, 0.69-0.88) for proportional change in 3D AP diameter and 0.72 (95% CI, 0.56-0.83) for proportional change in hiatal area. The prevalence of major levator injury was 22.6%. The strength of correlation (r(S)) between ultrasound measurements and MOS score was 0.52 for 2D AP diameter, 0.62 for 3D AP diameter and 0.47 for hiatal area (P < 0.001 for all). On the ultrasound contraction scale, proportional change in 2D levator hiatal AP diameter of < 1% corresponds to absent, 2-14% to weak, 15-29% to normal and > 30% to strong contraction. Conclusions Ultrasound seems to be an objective and reliable method for evaluation of pelvic floor muscle contraction. Proportional change in 2D levator hiatal AP diameter had the highest ICC and moderate correlation with MOS score assessed by vaginal palpation, and we constructed an ultrasound scale for assessment of pelvic floor muscle contraction based on this measure. Copyright (c) 2019 ISUOG. Published by John Wiley & Sons Ltd.