OP06.04: Discrimination of benign and malignant adnexal masses in daily clinical routine: external validation of ADNEX and RMI in Denmark – preliminary results

To improve the quality of clinical management of adnexal masses and compare the diagnostic efficiency of ultrasound-based risk models for clinical decision-making. In this prospective multicentre study, we implemented systematic ultrasound description according to IOTA criteria in gynecological departments in all hospitals in the Central Denmark Region. A standardised ultrasound description form (including Risk-of-Malignancy-Index (RMI) and Assessment of Different Neoplasia in the adneXa (ADNEX) model calculation) was implemented in the electronic patient file. All departments had priorly been offered IOTA-certification and had regular IOTA teaching during the project period. Patients with adnexal masses at ultrasound were consecutively included. Reference standard was histopathological diagnosis or follow-up for conservatively managed adnexal masses. Inconsistencies in ultrasound descriptions were recorded. A total of 4886 systematic ultrasound scans were performed in 1800 patients during 1.5 years in 2021-2023. At least one year follow-up were available in all patients. Median age was 58.1 (Inter Quartile Range 42.9-72.3). Overall, the Area Under the receiver operating characteristics Curve (AUC) was higher for ADNEX: 0.93 (95% Confidence Interval 0.90-0.95) compared to RMI: 0.88 (0.85-0.91) (p < 0.001) for prediction of malignancy in the first recorded ultrasound classification. Pitfalls in use of IOTA criteria were observed. This large, strictly designed prospective external validation study of IOTA criteria implemented in a regional setting suggests high and superior diagnostic performance of ADNEX compared to RMI for identification of malignancy. Notably, this was observed in a low-risk patient population by mixed-level ultrasound observers, representative of general gynecological departments. The study illustrates the feasibility and advantages of implementing IOTA terminology and predictive models in a national setting. Attention to ultrasound education is needed.