The correlation of the diagnosis made at intraoperative consultation (IC) with the final diagnosis is one of the cornerstones of quality assurance in the anatomic pathology laboratory.
To review correlation of IC diagnoses with final diagnoses during a 1-year period in a regionalized, multisite hospital setting in a major Canadian city.
One pathologist reviewed all surgical pathology cases at Calgary Laboratory Services from June 2004 through May 2005 that had an IC to extract the following data points: intraoperative diagnoses, final diagnoses, correlation between the two, anatomic site of the tissue on which the IC was requested, pathologic procedure requested of the IC, types of disagreements encountered, reasons for disagreement, and deferrals.
Intraoperative consultations occurred for 2812 specimens, of which 87 were discordant and 135 were deferred. Percent agreement was 96.75% (95% confidence interval, 96.08-97.42) with a kappa statistic of 0.94 (95% confidence interval, 0.92-0.95). Lymph nodes for evaluation for metastases (427), thyroid/parathyroid (401), and central/peripheral nervous system (378) specimens were sent most frequently for IC, and the latter 2 tissue types accounted for the greatest number of disagreements. The most common assessments requested were the presence/typing of a neoplasm (1161) and assessment of margins (730), which also accounted for the largest number of disagreements. Disagreements were most frequently due to interpretive (53) and gross sampling (23) errors; false-negative disagreements were nearly 3 times as common as false positives.
The IC was an excellent diagnostic test. Agreement and deferral rates varied by specimen site and by type of assessment requested.