Intraprocedural evaluation of fine-needle aspiration smears: how good are we?

A significant percentage of fine-needle aspirations (FNAs) is obtained using radiographic image guidance. Such image-guided aspirates are associated with increased costs, and it is highly desirable that adequate material for diagnosis is obtained at the time of the initial procedure. Hence, intraoperative evaluation of smear material is frequently performed to evaluate adequacy and to triage the aspirate to specialized tests such as flow cytometry and cytogenetics for appropriate work-up. Although FNA under radiologic guidance has been shown to be useful, little data exist as to the accuracy of such intraprocedural evaluation. The quality-assurance records between May 2009 and May 2010 were electronically searched for all FNA cases with intraprocedural evaluation. The reports were obtained and reviewed to compare the intraprocedural diagnosis and the final cytopathologic diagnosis. Percentage of deferred cases and percentage of disagreement were calculated. From May 2009 to May 2010, a total of 775 FNAs were performed with intraprocedural review. These cases were reviewed by seven board-certified cytopathologists. In 738 cases, the preliminary interpretation agreed with the final diagnosis. In 2.2% (17 of 775) of cases, the interpretation was deferred until-cell block, and/or cytospin preparations were obtained. In 2.6% (20 of 775) of cases, a disagreement in diagnosis existed between the initial interpretation and the final sign-out diagnosis. Intraprocedural interpretation of FNA material appears to be highly reliable with disagreements occurring in only 2.6% of cases. This degree of disagreement is slightly higher than that reported in some frozen section studies.