Diagnostic accuracy of transvaginal ultrasound using the IDEA protocol and UBESS staging compared with laparoscopy in deep endometriosis

Deep endometriosis (DE) poses diagnostic and surgical challenges. We evaluated transvaginal sonography (TVS) performed according to the International Deep Endometriosis Analysis (IDEA) protocol, and the Ultrasound-Based Endometriosis Staging System (UBESS), against laparoscopic findings and surgical complexity. This prospective pilot study enrolled 33 women with suspected DE scheduled for operative laparoscopy at a tertiary referral centre in India. A single experienced radiologist performed IDEA-based TVS and assigned a UBESS stage preoperatively; laparoscopy was the reference standard. Diagnostic accuracy was calculated from 2 × 2 contingency tables with exact 95% confidence intervals (CIs); for UBESS, Stage III (high surgical complexity) was the positive outcome. At laparoscopy, 16 women (48.5%) had UBESS Stage III and 17 (51.5%) Stage II disease; none had Stage I. TVS-derived UBESS correctly classified 13/16 Stage III and 17/17 Stage II cases (sensitivity 81.3%, 95% CI 54.4–96.0%; specificity 100%, 80.5–100%; accuracy 90.9%, 75.7–98.1%); three Stage III cases were under-staged. All laparoscopically confirmed endometriomas were correctly identified by TVS (right 30/30, 95% CI 88.4–100%; left 29/29, 88.1–100%), although specificity rested on only 3–4 endometrioma-negative patients. Sensitivity was 85.7% (18/21; 63.7–97.0%) for pouch of Douglas (POD) obliteration and 81.3–84.2% for ovarian immobility, with no false-positive findings. In this small single-centre cohort, IDEA-based TVS identified endometriomas and POD obliteration with few errors, and UBESS under-staged 3 of 16 high-complexity cases without over-staging any case. Wide CIs, absence of Stage I disease and an expert single-operator setting preclude definitive conclusions; validation in larger multicentre cohorts is required.