Evaluation of Four Risk of Malignancy Indices (RMI) in the Preoperative Diagnosis of Ovarian Malignancy at Rajavithi Hospital

• Main Authors: Putsarat Insin, Nisa Prueksaritanond
• Format: Article
• Language: English
• Published: The Royal Thai College of Obstetricians and Gynaecologists 2013-12-01
• Series: Thai Journal of Obstetrics and Gynaecology
• Subjects: Ovarian tumor; Risk of malignancy index (RMI); serum levels of CA-125; Ultrasound score; Menopausal status
• Online Access: https://tci-thaijo.org/index.php/tjog/article/download/11007/14165/

Objective:To evaluate the ability of four types of the risk of malignancy indices (RMI) based on serum levels of CA-125, ultrasound score, and menopausal status to discriminate between benign and malignant ovarian tumors.Material and Method: This is a retrospective study of 255 women admitted at Rajavithi Hospital between January 2012 and December 2012 for elective laparotomy of ovarian tumor. The sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of four types of the risk of malignancy indices were calculated. And the Receiver Operating Characteristic (ROC) curves for RMI 1, RMI 2, RMI 3, and RMI 4 were calculated to compare the accuracy.Results:Using a cut-off level of 200 to indicate malignancy for RMI 1, RMI 2 and RMI 3, and using a cut-off level of 450 to indicate malignancy for RMI 4. The RMI 2 gave the highest sensitivity (71%) while the RMI 1, RMI 3 and RMI 4 gave the sensitivity of 62–69%. The RMI 1 gave the highest specificity (80%) while the RMI 2, RMI 3 and RMI 4 gave the specificity of 71–78%. The positive predictive value of the four methods was 61-66% and the negative predictive value of the four methods was 66-80%. For the ROC curve, the greatest area under curve (AUC) was associated with the RMI 4 values (0.801) as compared to the ROC values for the RMI 1 (0.785), RMI 2 (0.782), and RMI 3 (0.778).Conclusion: The RMI is able to discriminate between benign and malignant ovarian tumors. The RMI 4 was the most reliable in predicting malignancy in terms of area under the curves. It is a simple method that can be incorporated into clinical practice easily to enable the selection of patients for referral to a gynecologic oncologist.