Ultrasound Sliding Sign Assessed by Transvaginal or Transrectal Route for Predicting Complete Pouch of Douglas Obliteration in Patients with Ovarian Endometriomas: An Exploratory Retrospective Study

Background and Objective: The ultrasound sliding sign dynamically assesses pouch of Douglas (POD) obliteration in suspected endometriosis. Transvaginal (TVS) and transrectal (TRS) ultrasound are used in patients with and without a self-reported history of vaginal intercourse, respectively, and thus in clinically distinct populations. We evaluated sliding sign performance for predicting surgically confirmed complete POD obliteration in patients with ovarian endometriomas and described findings in route-defined TVS and TRS groups. Materials and Methods: This retrospective cohort study enrolled 318 patients with surgically and pathologically confirmed ovarian endometriomas; intraoperative POD status was evaluable in 309. Patients with and without a self-reported history of vaginal intercourse underwent TVS (n = 247) and TRS (n = 71), respectively. A negative sliding sign (absent movement between the uterus and rectum) was the positive index test result for complete POD obliteration. The unadjusted full-cohort diagnostic accuracy analysis was the primary analysis. Because no parous patient underwent TRS (a lack of overlap in examination-route assignment among parous patients), exploratory adjusted analyses were restricted to nulliparous patients, using overlap weighting (principal adjusted analysis) and 1:1 propensity score matching (sensitivity analysis). Results: In the primary unadjusted analysis of the full cohort (n = 309), the negative sliding sign showed limited to moderate discriminative ability (sensitivity 0.71, 95% CI 0.63–0.78; specificity 0.60, 95% CI 0.52–0.67; PLR 1.78, 95% CI 1.44–2.20; NLR 0.48, 95% CI 0.36–0.64) and was associated with complete POD obliteration (p < 0.001). The between-route sensitivity difference was inconclusive (0.19, 95% CI −0.01 to 0.38). In the exploratory overlap-weighted analysis of nulliparous patients (n = 217), the association remained significant (weighted OR = 2.80, 95% CI 1.17–6.69; p = 0.021), without significant sliding sign-by-route interaction (p = 0.302); findings were consistent in the matching sensitivity analysis (66 pairs). Conclusions: The sliding sign was associated with surgically confirmed complete POD obliteration in patients with ovarian endometriomas, but evidence that this association differed between route-defined groups was inconclusive. Route-specific findings are exploratory because the examination route was determined by sexual history rather than randomization. Whether TRS offers performance comparable to TVS remains undetermined, because the two techniques were not evaluated in the same population and TRS feasibility was not formally assessed.