Beyond the Image: Performance of O-RADS, ADNEX, IOTA Simple Rules & RMI 4 in differentiating benign and malignant adnexal masses – An Indian perspective
To prospectively and externally validate O-RADS (Ovarian-Adnexal Reporting and Data System), RMI 4 (Risk of Malignancy Index), IOTA SR (International Ovarian Tumour Analysis Group Simple Rules) and ADNEX Rules (Assessment of Different NEoplasias in the adneXa), and compare their diagnostic accuracy in differentiating benign and malignant Adnexal Masses (AM) in the Indian population. This single-centre diagnostic accuracy study included 100 consecutive consenting women with AM presenting to the gynaecology department at a tertiary care centre in Northern India. Every AM was classified as benign or malignant using O-RADS, RMI4, IOTA SR, and the ADNEX model (with CA 125). The reference standard was histopathological diagnosis. Sensitivity, specificity, positive and negative predictive values, positive and negative likelihood ratios and overall accuracy were determined for all systems among various age groups and histopathological subtypes. Of the 100 women (mean age 42 ± 16.8 years) recruited, 45 had histopathologically benign lesions, 41 malignant and 14 borderline AM. O-RADS showed the highest sensitivity at 98 % (95 % CI: 87.4-99.5), followed by IOTA SR at 95 % (95 % CI: 83.8-98.6), RMI 4 at 88 % (95 % CI: 74.4-94.6), and ADNEX at 87.8 % (95 % CI: 73.8 %-95.9 %). However, ADNEX demonstrated the maximum specificity at 93 % (95 % CI: 81.7 %-98.6 %), followed by RMI 4 at 89 % (95 % CI: 76.5-95.1), IOTA SR at 87 % (95 % CI: 73.82-93.74), and O-RADS at 53 % (95 % CI: 39.1-67.1). Overall, ADNEX and IOTA SR showed the highest and similar diagnostic accuracy at 91 % (95 % CI: 82.7-95.2), followed by RMI 4 at 88 % (95 % CI: 79.9-93.5) and O-RADS at 74 % (95 % CI: 64.2-82.4). The O-RADS, IOTA Simple Rules and ADNEX performed well in distinguishing benign from malignant adnexal masses in an Indian population among various age groups and histopathological subtypes. While O-RADS was the most sensitive, the ADNEX model showed the best specificity. Additionally, RMI 4 is a robust diagnostic tool with strong diagnostic accuracy, outlining the importance of its application in low-resource settings.