Search for articles
The accuracy of frozen section of uterine lesions in the practice of gynecologic surgery. A retrospective assessment study in a government tertiary training hospital
Jimmy A. Billod, MD, FPOGS; Efren J. Domingo, PHD, MD, FPOGS and Nelson T. Geraldino, MD, FPOGS
Department of Obstetrics and Gynecology, Philippine General Hospital, University of the Philippines-Manila
Methodology. This is a retrospective validation study from 2004-2015 involving cases of uterine lesions from gynecologic surgeries. All histopathologic results of frozen and paraffin sections were retrieved and reviewed. Chi square test with 2x2 Fischer Exact test adjustment was used to check for associations. Accuracy indices of FS tool were estimated such as sensitivity, specificity, likelihood ratios, negative and positive predictive values, and overall accuracy. A p-value of < 0.05 alpha is considered significant.
Results. A total of 143 uterine specimens were submitted for frozen section analysis. The utilization rate of FS is 1% per year. The FS results were correlated with the final histopathologic diagnosis with 96% agreement rate. Utilizing a median number of 3 sections per specimen provides an overall accuracy rate of 97%. The accuracy rate of FS is equal between combined benign-premalignant and malignant cases at 96%. The accuracy rate is not statistically affected by the procedure by which the specimen was taken, as well as the source and gross morphology of the specimen. Moreover, a minimum of 11 sections per specimen is needed to obtain an accuracy rate of 99-100%. The accuracy rate particularly for endometrial lesions is between 94 and 100%.
Conclusion. Accuracy rates of frozen section on uterine lesions are high regardless of the sampling procedure and source of the specimen. Increasing the number of sections during FS parallels that of the final histopathologic diagnosis. FS for uterine lesions is a vital and cost-effective intraoperative decision tool to maximize care of patients.