Conventional flexible bronchoscopy versus radial endobronchial ultrasound (R-EBUS)-guided bronchoscopy under fluoroscopy: A retrospective study at Phrapokklao Hospital, Chanthaburi, Thailand

Abstract

Background: Conventional flexible bronchoscopy is widely used for evaluating peripheral pulmonary lesions; however, its diagnostic yield remains limited. Although radial endobronchial ultrasound (R-EBUS) combined with fluoroscopic guidance may improve lesion localization and tissue acquisition, real-world comparative data from provincial tertiary-care centers are scarce.

Objectives: To compare the diagnostic yield of conventional flexible bronchoscopy and R-EBUS–guided bronchoscopy under fluoroscopic guidance for peripheral pulmonary lesions and identify patient-, lesion-, and procedure-related factors associated with diagnostic success.

Methods: We conducted a retrospective observational study at a provincial tertiary-care hospital in Eastern Thailand. Adult patients undergoing diagnostic bronchoscopy for peripheral pulmonary lesions between October 2023 and October 2025 were included. Diagnostic yield was defined according to contemporary ATS/ ACCP Delphi consensus criteria. Univariable and multivariable logistic regression analyses were performed to evaluate factors associated with diagnostic yield, and Firth penalized logistic regression used as confirmatory analysis to address potential sparse-data bias.

Results: A total of 102 patients were analyzed, including 66 who underwent conventional bronchoscopy and 36 who underwent R-EBUS–guided bronchoscopy under fluoroscopic guidance. The final procedure-level diagnostic yield was higher, but not significantly different, in the R-EBUS with fluoroscopy group than in the conventional group (75.0% vs. 60.6%; P = 0.14). After adjusting for clinically relevant lesion-related factors, REBUS–guided bronchoscopy under fluoroscopic guidance was independently associated with higher diagnostic yield (adjusted odds ratio [OR] 3.72, 95% confidence interval [CI] 1.12–12.38, P = 0.03). Presence of a computed tomography (CT) bronchus sign was the strongest, most consistent predictor of diagnostic success across all models (adjusted OR 31.33, 95% CI 8.43–116.52, P < 0.001). Confirmatory Firth penalized regression yielded consistent estimates in direction and magnitude.

Conclusion: In this real-world provincial tertiary-care setting, R-EBUS–guided bronchoscopy under fluoroscopic guidance was independently associated with higher diagnostic yield after accounting for lesion accessibility, particularly presence of a CT bronchus sign. Lesion-related factors, especially bronchial anatomy, play a central role in determining diagnostic success and should be carefully assessed when selecting bronchoscopic techniques in resource-limited settings.