Jeffus Susanne K, Dustin-Hess Simone M, Atkins Kristen A
Department of Pathology, University of Arkansas for Medical Sciences, 4301 W. Markham Street, Little Rock, Arkansas.
Department of Pathology, University of Virginia, Charlottesville, Virginia.
J Am Soc Cytopathol. 2014 Mar-Apr;3(2):67-72. doi: 10.1016/j.jasc.2013.12.004. Epub 2013 Dec 25.
Rapid on-site evaluation (ROSE) of fine-needle aspirates is an invaluable teaching tool for a cytopathology (CyP) fellowship. The ability of fellows to accurately perform ROSEs without direct attending supervision is not well documented in the literature. This study reviewed ROSEs performed independently by CyP fellows and focused on diagnostic discrepancies with managerial implications.
All fine-needle aspirates with ROSE documentation performed at the University of Virginia from October 1, 2007 to March 31, 2013 were reviewed and compared with the final diagnosis. Cases were only included if a CyP fellow performed the ROSE. Discrepancy between ROSE and final diagnosis was categorized according to the change. Numbers of false positive (FP) and false negative diagnoses, organ site, and recurrent interpretative pitfalls were noted.
CyP fellows performed 6815 ROSEs in 6 years. An attending cytopathologist was present 8% of the time. Of ROSEs without direct attending supervision (6224 fine-needle aspirates), the preliminary and final diagnoses were identical in 95% of cases. FP rate was 1.06%. The most frequent categorical change occurred from ROSE of "atypical" to final diagnosis of "malignant." The most common sites involved in FP diagnoses were pancreas/biliary tract, lung, and lymph node. Experience gained over the fellowship year did not significantly affect the FP rate. Errors encountered are known interpretative challenges.
This is the largest study addressing discrepancies between ROSE and final diagnosis and the first study examining CyP fellow performance. Our results affirm that fellows perform extremely well when performing ROSEs independently.
