Comparative accuracy of telecytology and in-person rapid on-site evaluation by specimen source: A multi-campus analysis

Abstract

Rapid on-site evaluation (ROSE) of fine-needle aspiration (FNA) specimens improves specimen adequacy, diagnostic accuracy, and patient management. Telecytology (TC) has emerged as a potential alternative to in-person ROSE, particularly in multi-campus healthcare systems where cytopathologist coverage is limited; however, its comparative accuracy across different specimen sources remains uncertain. In this retrospective study, all FNAs with ROSE preliminary diagnoses performed across three campuses in our health system from 2022 to 2023 were reviewed. ROSE interpretations rendered via in-person evaluation or TC were compared with final diagnoses. Major discrepancies were defined as benign-to-malignant or malignant-to-benign discordance. Diagnostic accuracy was analyzed by specimen source, and chi-square tests were used to assess statistical significance. Of 2836 FNAs performed, 1697 underwent ROSE, including 1459 in-person and 238 TC cases. Major discrepancies occurred in 3% of in-person ROSE cases and 5% of TC cases, with no statistically significant difference (p = 0.2429). TC demonstrated numerically higher concordance than in-person ROSE for several visceral sites, including liver, kidney/adrenal, pancreatobiliary, and upper gastrointestinal FNAs, though none reached statistical significance. In contrast, in-person ROSE showed significantly higher accuracy for lymph node, head and neck, and bone/soft tissue FNAs. Overall, TC demonstrated accuracy comparable to in-person ROSE, with performance varying by specimen source. These findings support the feasibility of TC while underscoring the importance of specimen-specific evaluation, ongoing quality assurance, and prospective validation in multi-campus cytopathology practices.