Abstract
RATIONALE: Most bronchoscopic procedures are performed using moderate sedation achieved by combining a short-acting benzodiazepine with an opioid agent. Propofol (2.6-diisopropylphenol), a short-acting hypnotic agent, has been increasingly used to provide deep sedation in the endoscopy community with an acceptable safety profile. OBJECTIVES: To compare the impact of moderate versus deep sedation on the adequacy and diagnostic yield of endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA). METHODS: A retrospective review of prospectively collected data was performed at two academic institutions with interventional pulmonary fellowships using two methods of sedation during EBUS (deep vs. moderate sedation). Rapid on-site cytologic evaluation was used on all procedures in both groups. EBUS-TBNA nodal sampling was considered adequate if the aspirate yielded a specific diagnosis or lymphocytes. EBUS-TBNA was considered diagnostic if a lymph node aspirate yielded a specific diagnosis or if subsequent surgical sampling or prolonged radiographic surveillance revealed no nodal pathology. MEASUREMENTS AND MAIN RESULTS: No difference was observed in the indication for EBUS-TBNA between the two groups. More lymph nodes were sampled per patient in the deep sedation group (314 nodes from 163 patients; 2.2 nodes per patient) than in the moderate sedation group (181 lymph nodes from 146 patients; 1.4 nodes per patient; P < 0.01). The EBUS-TBNA diagnostic yield was higher for the deep sedation group (80% of patients) than for the moderate sedation group (66% of patients; P < 0.01). CONCLUSIONS: Diagnostic yield and number of lymph nodes sampled using deep sedation is superior to moderate sedation in patients undergoing EBUS-TBNA. Prospective studies accounting for other factors including patient selection and cost are needed.