Levator ani muscle avulsion: Digital palpation versus tomographic ultrasound imaging.

Abstract

To determine the diagnostic performance of digital palpation of levator ani muscle (LAM) avulsion compared with translabial tomographic ultrasound imaging (TUI). A cross-sectional study, incorporating 195 women involved in a longitudinal cohort study. Palpation for levator integrity was performed, followed by a four-dimensional translabial ultrasound. LAM avulsion defects were diagnosed in the presence of puborectalis muscle detachment from its insertion. Post-processing analysis of ultrasound volumes for LAM integrity on TUI was performed blinded against palpation findings. Agreement between methods was assessed using Cohen's κ. In all, 388 paired assessments of LAM bilaterally, were available. Sixteen (8.2%) unilateral avulsion defects were detected on palpation. Sonographically, 31 (16%) were diagnosed with avulsions: 4.6% bilateral and 11.3% unilateral. An overall agreement of 91% was observed between digital palpation and TUI, yielding a Cohen's κ of 0.32 (95% confidence interval 0.15-0.48) demonstrating "fair agreement": and implying 25% sensitivity, 98% specificity, 63% positive predictive value, and 92% negative predictive value. Analysis of the first and last 20 palpations showed no change in performance during the 13-day study period. Assessment of LAM avulsion defects by digital palpation is feasible but may require substantial training. Confirmation by imaging is crucial, especially if the diagnosis of avulsion may influence clinical management.