Regardless of one’s views on the utility of ROSE, its application is not easy. It generally requires the physical presence of a cytologist on-site to prepare and evaluate the slides. This may be especially challenging in smaller institutions which lack dedicated resources. This has led to the development of alternatives, such as using a cytotechnologist to prepare and read the specimen,11 or dynamic telecytology.12 Another intriguing option has been described by Bonifazi et al13 in this issue of CHEST (see page 60). In this study, a pulmonologist undertook 3 months of cytology training, which consisted of taking part in weekly diagnostic cytopathology sessions for 3 h, for a total of 18 sessions, or 54 h. Additionally, the pulmonologist also acquired knowledge by referencing a cytology textbook. Once trained, this pulmonologist then attended bronchoscopy procedures where transbronchial needle aspiration (TBNA) was being performed. At these procedures, the pulmonologist stained and performed ROSE on one smear per pass, evaluating the sample for adequacy and diagnosis. Diagnosis was placed into one of the five categories, including nondiagnostic, specific benign lesion, doubtful but probably benign, doubtful but suspicious of malignancy, and malignant lesion. These slides were then sent to the cytopathologist, who read and categorized the slides using the same classification. Notably, both readers (pulmonologist and cytopathologist) were not blinded to the clinical and radiographic data of the patient. Four other slides from each pass were placed into 95% ethanol and processed at a later time for definitive evaluation. Four separate passes per node were performed, and this was not influenced by the pulmonologist’s reading of the ROSE slide. There was an 81% overall agreement between the pulmonologist and cytopathologist, with a κ of 0.73. The agreement was stronger with definitive malignant disease with 92% agreement and a κ of 0.81. More importantly, there was a substantial agreement in the ability to identify adequate vs inadequate samples (84% agreement, with a κ of 0.68). These data suggest that pulmonologists trained in cytology may not need to rely on the presence of a cytopathologist for the performance of ROSE.