Interinstitutional comparison of performance in breast fine-needle aspiration cytology. A Q-probe quality indicator study.

In 1989, the College of American Pathologists, Northfield, Ill, instituted a voluntary quality assurance program, called "Q-Probes," that utilized nationwide interinstitutional peer comparison. One of the anatomic pathology modules retrospectively assessed performance in fine-needle aspiration cytology (FNAC) of the breast from cytohistologic correlations that were made in 294 institutions by 988 pathologists on their own cases that were accessioned during a 6-month period. Of the 13,066 cases of FNAC, 10,751 (82%) were satisfactory for evaluation. Of these satisfactory aspirates, 3471 cases (33%) had histologic correlation, which formed the basis for determining diagnostic accuracy. Of breast aspirates, 2254 (17%) were unsatisfactory for evaluation, with the mean frequency of unsatisfactory aspirates obtained by nonpathologists (18%), ie, more than twice that of unsatisfactory aspirates obtained by pathologists (7.2%). In the diagnosis of breast cancer by FNAC, the following performance results were derived with the use of the aggregate data: 82% sensitivity of the FNAC procedure, 97% sensitivity of diagnosis, 97% specificity, 95% positive predictive value, 86% negative predictive value, and 90% efficiency. No significant difference in performance was detected when institutions were stratified by six peer group characteristics. We made the following conclusions: (1) the aggregate and median breast FNAC performance values obtained from this interinstitutional comparison of data from routine procedures performed in diverse settings in North America compared very favorably with performance from single institutions published in studies of similar design in the literature, and (2) these data may provide a reference point for participant institutions to measure future quality improvement in fine-needle aspiration of breast.