OBJECTIVE: To optimize referral to specialized gynaecologists for surgical treatment of ovarian cancer by improving preoperative discrimination between benign and malignant pelvic tumours. STUDY DESIGN: In a prospective multicentre study 143 patients with a pelvic mass were included. At several occasions during the diagnostic work-up the gynaecologist estimated the chance of malignancy (educated guess/expert opinion). MRI in the local setting was suggested for uncertain cases. All MRI images were reviewed by an expert radiologist. The datasheet designed for the study further allowed for determining the risk of malignancy index (RMI). RESULTS: The diagnostic accuracy of the gynaecologist's final estimation of the chance of malignancy and the calculated RMI were comparable (area under the ROC curve of 0.87 and 0.86). MRI did not improve the accuracy of the diagnostic work-up for the study population as a whole. Subgroup analysis did however show improved diagnostic accuracy in cases with an estimated chance of malignancy between 20 and 80% when the MRI was read by an expert radiologist. CONCLUSION: Patient selection for surgery of a pelvic mass should be based on the chance of malignancy as assigned by the referring gynaecologists. In case of uncertainty MRI improves diagnostic accuracy, when judged by an expert.
OBJECTIVE: To optimize referral to specialized gynaecologists for surgical treatment of ovarian cancer by improving preoperative discrimination between benign and malignant pelvic tumours. STUDY DESIGN: In a prospective multicentre study 143 patients with a pelvic mass were included. At several occasions during the diagnostic work-up the gynaecologist estimated the chance of malignancy (educated guess/expert opinion). MRI in the local setting was suggested for uncertain cases. All MRI images were reviewed by an expert radiologist. The datasheet designed for the study further allowed for determining the risk of malignancy index (RMI). RESULTS: The diagnostic accuracy of the gynaecologist's final estimation of the chance of malignancy and the calculated RMI were comparable (area under the ROC curve of 0.87 and 0.86). MRI did not improve the accuracy of the diagnostic work-up for the study population as a whole. Subgroup analysis did however show improved diagnostic accuracy in cases with an estimated chance of malignancy between 20 and 80% when the MRI was read by an expert radiologist. CONCLUSION:Patient selection for surgery of a pelvic mass should be based on the chance of malignancy as assigned by the referring gynaecologists. In case of uncertainty MRI improves diagnostic accuracy, when judged by an expert.