AbstractBackground: Frozen section is an important tool in the intraoperative diagnosis of ovarian tumours which guides the surgeon in tailoring the surgical therapy, especially in younger women. Aim: To determine the accuracy of frozen section in ovarian neoplasms and to study the possible sources of discordance. Methods: Retrospective study of 78 ovarian tumours sent for frozen section from July 2016 to July 2018. The frozen and paraffin section reports were compared and accuracy, sensitivity, specificity, positive predictive value and negative predictive value of frozen section was determined individually for benign, borderline and malignant tumours. Statistical Analysis: Sensitivity, specificity, positive predictive value and negative predictive value were calculated for benign, borderline and malignant tumours separately. Results: The overall accuracy of frozen section in diagnosis of ovarian tumours was 97.4%. The sensitivity and specificity for benign tumours was 100%; for borderline tumours it was 100% and 97% respectively; for malignant tumours 92% and 100% respectively. The major source of discordance was found to be in regard to borderline mucinous tumours, with 28.5% showing invasion on final paraffin section. Conclusion: Accuracy of frozen section is high in benign and malignant ovarian tumours. The accuracy rate declines in cases of borderline tumours especially the mucinous type. Deferring the frozen section report to paraffin section is indicated when the diagnosis of borderline versus invasion is difficult or suspected.
Keywords: Accuracy; Borderline; Frozen Section; Mucinous; Ovarian tumors.