Frozen section diagnosis: an audit.

A consecutive series of 1000 operative frozen section diagnoses was reviewed. Correct diagnosis was made at the time in 96.5% of the cases. Clinically relevant errors were found in 1.3% of the cases and unimportant errors in 0.9%. Diagnosis was deferred, to await subsequent paraffin sections, in a further 1.3%. All the errors and provisional diagnoses in the deferred cases were conservative false negative results; no false positive diagnosis of malignancy was made. The cases of incorrect or deferred diagnosis were analysed to ascertain the origin of the difficulties, which comprised: technical imperfection (three cases); the focal nature of the lesion (14); and pathological misinterpretation (28). More than one of these factors played a part in eight cases. Further retrospective assessment indicated that the factors leading to error or deferred diagnosis were avoidable in 57% and potentially avoidable in 43% of cases. Misinterpretation was the single factor responsible for all avoidable misdiagnoses or deferred diagnosis. None the less, unavoidable factors led to erroneous or deferred diagnosis in about 2% of operative frozen section requests. Using present methods this seems to be the irreducible minimum of failures to make the correct diagnosis when frozen sections are assessed.