RT Journal Article SR Electronic T1 Accuracy of frozen-section analysis in the diagnosis of ovarian tumors: a systematic quantitative review JF International Journal of Gynecologic Cancer JO Int J Gynecol Cancer FD BMJ Publishing Group Ltd SP 192 OP 202 DO 10.1136/ijgc-00009577-200503000-00002 VO 15 IS 2 A1 L. R. Medeiros A1 D. D. Rosa A1 M. I. Edelweiss A1 A. T. Stein A1 M. C. Bozzetti A1 A. Zelmanowicz A1 P. R. Pohlmann A1 L. Meurer A1 M. T. Carballo YR 2005 UL http://ijgc.bmj.com/content/15/2/192.abstract AB A quantitative systematic review was performed to estimate the diagnostic accuracy of frozen sections in ovarian tumors. Studies that compared frozen sections and paraffin sections within subjects for diagnosis of ovarian tumors were included. Fourteen primary studies were analyzed, which included 3 659 women. For benign ovarian vs borderline/malignant tumor cases, the occurrence of a positive frozen-section result for benignity (pooled likelihood ratio [LR], 8.7; 95% confidence interval [CI], 7.3–10.4) and posttest probability for benign diagnosis was 95% (95% CI, 94–96%). A positive frozen-section result for malignant vs benign diagnosis (pooled LR, 303; 95% CI, 101–605) increased the probability of ovarian cancer to 98% (95% CI, 97–99%). In borderline vs benign ovarian tumor cases, a positive frozen-section result (pooled LR, 69; 95% CI, 45–106) increased the probability of borderline tumors to 79% (95% CI, 71–85%). In borderline vs malignant ovarian tumor cases, a positive frozen-section result (pooled LR, 18; 95% CI, 13–26) increased the probability of borderline tumors to 51% (95% CI, 42–60%). We conclude that diagnostic accuracy rates for frozen-section analysis is high for malignant and benign ovarian tumors, but the accuracy rates in borderline tumors remain relatively low.