Frozen-section diagnosis in surgical pathology. A prospective analysis of 526 frozen sections.

Five hundred eighty-six consecutive frozen-section consultations performed during a 1-year period were studied prospectively in order to assess the accuracy of the method and develop a quality control mechanism. The overall accuracy was 97.1%. The accuracy of the method with breast lesions was 97.9%. Specimens from the gastrointestinal tract and thyroid were incorrectly interpreted in 5% of the cases. The accuracy for lymph node specimens was 96.2%, with more than 50% consulted out of curiosity. The authors conclude that frozen section of lymph node is not recommended. Most of the errors were sampling errors made by the pathologist. The authors therefore conclude that in clinically suspected malignancy, more than one sample must be examined in order to decrease the false-negative diagnosis in frozen section.