ObjectiveTo evaluate the role of rapid on-site evaluation (ROSE) in the diagnosis of lung space-occupying lesions.MethodsFrom June 1, 2017 to December 30, 2017, transbronchial biopsies were performed in patients with lung space-occupying lesions on chest CT, and biopsies were taken for ROSE and subsequent routine pathological examination. ROSE interpretation was performed by a pulmonologist who had been trained in cytopathology for 3 months. The interpretation was correlated with the follow-up routine pathological examination. The contemporary cases with lung space-occupying lesions who underwent transbronchial biopsies without ROSE were retrospectively reviewed. The quality assessment of biopsy specimens recorded in pathological reports were compared between cases with and without ROSE.ResultsA total of 101 patients underwent transbronchial biopsies in parallel with ROSE. The interpretation results of the pulmonologist were compared with the follow-up routine pathology, which showed that the consistency rate of malignant lesions was 84.1% and the consistency rate of benign lesions was 93.8%. Consistency test showed good agreement between the ROSE conducted by the pulmonologist and the routine pathological examination by pathologists (κ=0.66, P<0.01). The quality assessment of biopsy specimens showed that there was no significant difference on rate of satisfied biopsy specimens between cases with and without ROSE (98.0%vs 94.5%, P=0.14).ConclusionsThe use of ROSE combining with bronchoscopy allows good preliminary assessment of lung space-occupying lesions. Pulmonologists trained in short-term formal cytopathology are fully capable of performing ROSE, thereby obviating the need for cytopathologists to participate in on-site evaluation.