The utility of WhatsApp-based off-site evaluation for rapid cytology of EBUS-TBNA samples

Whether off-site evaluation of slides by a cytologist viewing the images shared by WhatsApp improves the on-site evaluation by a pulmonologist (P-ROSE) remains unknown. This study's objective was to compare the sensitivity of P-ROSE and WHOSE for adequacy and diagnosis of cytology specimens obtained by endobronchial ultrasound (EBUS)-guided transbronchial needle aspiration (TBNA).

We retrospectively reviewed our bronchoscopy database to identify subjects who underwent EBUS-TBNA for lymph node sampling and had reports of P-ROSE and WHOSE. We collected data on the adequacy of samples as reported by the pulmonologist (P-ROSE), remotely by the cytologist (WHOSE), and finally after detailed cytologic evaluation. The study's primary outcome was to assess the increment in sensitivity for adequacy and diagnostic category (using the final cytology report as reference) by incorporating WHOSE.

We included 264 (P-ROSE, n = 184; WHOSE, n = 80) subjects. The sensitivity (95% CI) for sample adequacy by P-ROSE and WHOSE was 65.3% (57.9%-72%) and 92% (83.6%-96.2%), respectively. There was a 26.6% (95% CI, 16%-35.2%) increment in the sensitivity for adequacy. The sensitivity (95% CI) for diagnosis by P-ROSE and WHOSE was 53.9% (46%-61.1%) and 89.8% (79.5%-95.3%), respectively. There was a 35.9% (95% CI, 23.4%-45%) increment in the sensitivity for diagnosis with WHOSE. The agreement between P-ROSE and final cytology in adequacy was poor (κ = -0.023, p = 0.616). The agreement between WHOSE and final cytology was moderate for adequacy (κ = 0.491, p = <0.001).

We found WHOSE significantly improves the performance of P-ROSE for rapid assessment of cytology specimens obtained by EBUS-TBNA.