ROSE, as a complementary technique in interventional pulmonology, has been the target of much controversy over the years due to its cost-effectiveness. Finding a pathologist during the procedure represents another challenge. This study aims to observe the learning curve in an intensive training program in ROSE for a resident in the last year of pulmonology and compare the agreement against an experienced pathologist.

A Prospective study was carried out at the University Clinical Hospital of Valladolid. A final-year Pneumology resident received 2 months of cytopathology training together with a pathologist(25h). The cytological samples were performed in 50 selected patients through brushing and TBNA of different lung lesions(mass, nodules, Lymph nodes). Two rapid stains were carried out: Papanicolaou and Diff-Quick. Finally, the sample was analyzed, first by the pulmonologist and then by the pathologist.

There was an 84% overall substantial agreement between the pulmonologist and pathologist in ROSE(k 0.682,p <0.001), which increased by 3% during the second month of the study. Brushing was the technique that most frequently led to the diagnosis(43%). There were no significant differences in diagnostic cost-effectiveness between the two stains. The best inter-observer agreement (80.5%) was obtained from samples of lung masses. The pulmonologist's agreement to verify if samples were from ganglion cells was 100%.

The high inter-observer agreement opens the possibility of performing ROSE by pulmonologists in places where pathologists are not available.

Excellent inter-observer agreement when it comes to knowing that the lymphadenopathy sample is adequate in blind punctures or EBUS.

Only one stain could be dispensed.