Sample adequacy in bronchoscopic ROSE: comparison between laboratory specialist and pathologist

Rapid on-site evaluation (ROSE) during bronchoscopy provides the opportunity to assess the adequacy of the sampling site, thus enabling diagnosis. The rapid evaluation of smears is typically performed by cytopathologists, but this is not always feasible. At our institution, a university-trained clinical biochemist performs the evaluation of ROSE smears. Our aim was to compare ROSE evaluations conducted by a laboratory specialist and a cytopathologist. We assessed the adequacy of lymph node samples from 78 patients using ROSE: 37 samples were obtained via EBUS-TBNA and 41 via TBNA. All smears were examined by the laboratory specialist, and simultaneously prepared parallel smears from the same needle passes were sent to the Department of Pathology for evaluation. Of the 78 samples, both the laboratory specialist and the pathologist team deemed 63 samples adequate and 8 samples inadequate. In 7 cases, there was a disagreement between the pathologist team and laboratory specialist. This resulted in a 92.3% agreement. The Cohen's kappa value was 0.71, indicating strong and Gwet's AC1 value was 0.90 corresponding to almost perfect agreement. The diagnostic performance was also excellent. Our conclusion is that ROSE performed by a laboratory specialist is a suitable alternative to on-site evaluation by a pathologist. It may help to overcome the resource shortage of interventional pulmonologists and cytopathologists.