Safety and risk factors for bleeding complications of radial probe endobronchial ultrasound-guided transbronchial biopsy.

Background: Radial probe endobronchial ultrasound (radial EBUS) is widely used to diagnose pulmonary lesions; however, the diagnostic value of radial EBUS-guided transbronchial biopsy (TBB) varies, and its complications (especially the risk of bleeding) are not properly understood.

Objectives: In this study, we evaluated the diagnostic performance and rate of complication of this procedure, and investigated the risk factors associated with the procedure-related bleeding events.

Design: A retrospective cohort study.

Methods: This was a retrospective study that included consecutive patients who underwent radial EBUS-guided TBB. Radial EBUS was performed under moderate sedation in inpatients or outpatients. The severity of bleeding was graded using the standardized definitions of bleeding.

Results: Of 133 patients (median age, 69 years; men 57.1%) included, 41 were outpatients (30.8%). The diagnostic accuracy, sensitivity, and specificity for malignancy were 76.1% (89/117), 71.1% (69/97), and 100% (20/20), respectively. The diagnostic accuracy ranged from 66.9% to 79.0%, depending on the classification of undiagnosed cases as either false negatives or true negatives. Twenty-seven patients (20.3%) developed complications (pneumothorax, 3; pneumonia, 5; complicated pleural effusion, 2; bleeding event grade 2 or higher, 21). Of the 41 outpatients, two developed complications (pneumothorax without intervention, 1; grade 2 bleeding event, 1). Of the 21 patients (15.8%) with procedure-related bleeding events, 18 had grade 2, and three had grade 3 bleeding complications. In multivariate analysis, a large size of ⩾30 mm (adjusted odds ratio (OR), 5.09; p = 0.03) and central lesion (adjusted OR, 3.67; p = 0.03) were significantly associated with the risk of grade 2 or higher bleeding events.

Conclusion: Our results suggest that radial EBUS-guided TBB is an accurate and safe method for diagnosing pulmonary lesions. Clinically significant procedure-related bleeding was rare. The central location and larger size (⩾30 mm) of pulmonary lesions were risk factors for grade 2 or higher bleeding events.