RESUMO
INTRODUCTION: Endobronchial ultrasound with transbronchial needle aspiration (EBUS-TBNA) has become an important diagnostic tool for the pulmonologist. Learning this procedure and maintaining technical skills requires continuous practice and evaluation. OBJECTIVES: The aims of the study were a retrospective evaluation of the diagnostic quality of the EBUS-technique and the learning profile of the endoscopy team during the first years (2007-2013) of experience in an unselected population. METHODS: EBUS-TBNA procedures were analysed for clinical data, including results from surgery or clinical/radiological follow-up for at least 6 months. Rapid on-site cytological evaluation (ROSE) was introduced on regular basis the forth year. RESULTS: A total of 711 EBUS-TBNA from 635 patients were included. The percentage of representative EBUS-TBNA initially decreased the first years (minimum 60,9%), before increased to a final result of 82,4%. There was a lower proportion of representative EBUS-TBNA in the benign group (76,8%) vs the malignant group (85,8%). A significant increase in the proportion of representative EBUS-TBNA was seen after ROSE had been introduced. The major indications were diagnosing/staging of lung cancer (54%) and mediastinal lymphadenopathy of unknown cause (25,7%). The sensitivity detecting malignancy was 94,9%, negative predictive value 81,2% and diagnostic accuracy 95,8%. During the study period the percentage of re-examinations with EBUS-TBNA declined from 18,0% to 8,2%. CONCLUSION: After an initial run-in period with declining results, the overall diagnostic yield of EBUS-TBNA increased and reached acceptable levels. These results underline the importance of continuously evaluation of our own results when new methods are implemented in clinical practice.
