Background: Endosonography is being implemented rapidly in pulmonary medicine for the diagnosis and staging of lung cancer, the assessment of sarcoidosis, and the assessment of mediastinal lesions. Although serious adverse events (SAEs) have been described, safety data outside cohort studies are scarce.
Objective: To assess the SAE and mortality rate of EUS-guided FNA (EUS-FNA) and endobronchial ultrasound guided transbronchial needle aspiration (EBUS-TBNA) for mediastinal and/or hilar analysis.
Design: Nationwide, retrospective survey by using questionnaires.
Setting: All hospitals in the Netherlands.
Patients: All patients undergoing EUS-FNA and EBUS-TBNA for intrathoracic analysis in the period 1999 to 2011.
Interventions: EUS-FNA and EBUS-TBNA.
Main Outcome Measurements: Occurrence of fatal outcomes and SAEs. Detailed information was obtained for each reported case, and all cases were reviewed independently by 2 investigators, including identification of risk factors.
Results: All 89 hospitals (100%) responded. An estimated 14,075 EUS-FNA and 2675 EBUS procedures were performed. Seven patients died after endosonography (5 EUS-FNA, 2 EBUS [mortality rate 0.04%]). All fatalities occurred in patients of poor performance status (American Society of Anesthesiologists Physical Status Classification System score of III/IV). Twenty-five SAEs were reported (22 EUS-FNA, 3 EBUS [SAE rate of 0.15%; EUS-FNA 0.16%, EBUS 0.11%]). SAEs were mostly (64%) of infectious origin. No specific risk factors for infectious adverse events could be identified.
Limitations: Retrospective study, possible recall bias, overrepresentation of EUS-FNA cases.
Conclusion: Endosonography appears to be a safe technique for the analysis of mediastinal and/or hilar lesions. Poor performance status is a risk factor for fatal outcomes. Mediastinitis and/or mediastinal abscess formation is rare but is a potential and dangerous adverse event of endosonography. (Gastrointest Endosc 2015;82:1009-15.)