OC12.02: *Diagnostic test accuracy of each of the four IDEA ultrasound steps

Robust guidelines are established on performing and reporting transvaginal ultrasound (TVS) for 'deep endometriosis' (DE) in the IDEA consensus 2016 and ESHRE 2022 endometriosis guideline outlining imaging as the 'gold standard' for endometriosis diagnosis. As access to tertiary ultrasound may be limited, we aim to assess the ability to of each cumulative subsequent step of the four step IDEA TVS for deep endometriosis. This retrospective international multicentre study used data previously collected to assess diagnostic accuracy of predicting and developing UBESS and validating the IDEA consensus methodology with results published previously. Patients scheduled for laparoscopy underwent DE ultrasound within 12 months of their intended surgery. Each individual step is analysed in correlation with prediction of 'any endometriosis' inclusive of superficial and deep. Subgroup analysis is performed to exclude patients with superficial endometriosis (SE) only. 640 participants were included. Endometriosis confirmed in 562 patients. 212 had SE only, 272 had DE with or without SE and 78 patients had no endometriosis at laparoscopy. With each additional step the sensitivity and NPV increased. A four step IDEA TVS for DE having sensitivity 0.96 and NPV 0.95 although specificity of 0.50. See table 1. Greatest interval increase in sensitivity was 11% with the addition of step 2. If inclusive of SE and DE the four-step scan had sensitivity of 0.73 and NPV 0.23. Incorporating the features of all four steps increases the sensitivity of detecting DE to a sensitivity of 96% from 0.78 with step 1. The high NPV value 95% of the cumulative four steps supports that a negative IDEA scan strongly supports the absence of deep endometriosis.