Accuracy of Intraoperative Frozen Section Diagnosis of Borderline Ovarian Tumors by Hospital Type

Abstract Study Objective To compare the accuracy of frozen section diagnosis of borderline ovarian tumors among 3 distinct types of hospital—academic hospital with gynecologic pathologists, academic hospital with nongynecologic pathologists, and community hospital with nongynecologic pathologists—and to determine if surgical staging alters patient care or outcomes for women with a frozen section diagnosis of borderline ovarian tumor. Design Retrospective study (Canadian Task Force classification II-1). Setting Tertiary care, academic, and community hospitals. Patients Women with an intraoperative frozen section diagnosis of borderline ovarian tumor at 1 of 3 types of hospital from April 1998 through June 2016. Interventions Comparison of final pathology with intraoperative frozen section diagnosis. Measurements and Main Results Two hundred twelve women met the inclusion criteria. The frozen section diagnosis of borderline ovarian tumor correlated with the final pathologic diagnosis in 192 of 212 cases (90.6%), and the rate of correlation did not differ among the 3 hospital types (p = .82). Seven tumors (3.3%) were downgraded to benign on final pathologic analysis and 13 (6.1%) upgraded to invasive carcinoma. The 3 hospital types did not differ with respect to the proportion of tumors upgraded to invasive carcinoma (p = .62). Mucinous (odds ratio, 7.1; 95% confidence interval, 2.1–23.7; p = .002) and endometrioid borderline ovarian tumors (odds ratio, 32.4; 95% confidence interval, 1.8–595.5; p = .02) were more likely than serous ovarian tumors to be upgraded to carcinoma. Only 88 patients (41.5%) underwent lymphadenectomy, and only 1 (1.1%) had invasive carcinoma in a lymph node. Conclusions A frozen section diagnosis of borderline ovarian tumor correlates with the final pathologic diagnosis in a variety of hospital types.