The accuracy of clinical evaluation of posterior vaginal wall defects.

We retrospectively reviewed 106 cases of defect-directed rectocele repair. Preoperative standardized evaluation of rectovaginal septum support was performed. Support defects were identified at surgery by visually identifying breaks in the rectovaginal fascia. Clinical and surgical findings were compared using the chi2 test. Logistic regression analysis was used to identify predictive variables. Clinical examination concurred with surgical findings in 59.4% (63/106) and differed in 40.6% (43/106). Predictors of agreement between clinical examination and intraoperative findings were the presence of multiple defects in the rectovaginal fascia at surgery (OR 4.42, 95% CI 1.89, 10.35) and stage II or III prolapse (OR 0.0007, 95% CI 0.00059, 0.003 and 0.0003, 95% CI 0.00033, 0.0017, respectively). The sensitivity and positive predictive value for all defects was less than 40%. The inaccuracy of clinical examination emphasizes the need to develop new clinical and/or imaging methods to evaluate posterior vaginal support defects.