Evaluating diagnostic yield and accuracy as key performance metrics in pulmonary lung lesions

• DOI: 10.3389/fmed.2025.1572779
• Publication Date: 2025-05-07T05:14:33Z
• Authors: Junsu Choe; Hyunseung Nam; Hwan-ho Cho; Sun Hye Shin; Byeong-Ho Jeong; Sang-Won Um; Hojoong Kim; Kyungjong Lee

A conservative definition of diagnostic yields for assessing the performance guided bronchoscopy has been proposed, but it yet to be validated in practice. Patients who underwent radial endobronchial ultrasound (R-EBUS) between April 2020 and 2023 were included study. Diagnostic results classified as malignant or non-malignant based on post-lung-biopsy pathology. Non-malignant further categorized into specific benign (SB), nonspecific (NSB), atypical cells, non-diagnostic (ND). All lesions confirmed using alternative biopsy methods chest computed tomography (CT) during a follow-up over 1 year. yield accuracy calculated pre-defined (Box below). Predictors sampling success identified logistic regression analysis. Among 736 patients evaluated this study, R-EBUS-guided TBLB revealed malignancy 431 (58.6%) patients. The remaining 305 (41.4%) with SB (8.3%), NSB (21.3%), atypia (4.6%), ND (7.2%). vs. values conservative, intermediate, liberal definitions 67% 67, 88% 77, 100% 79%, respectively. Thus, definition, identical. Significant predictive factors successful lung according lesion size (> 20 mm), CT-bronchus subclassification (Ia, Ib), probe position within lesion. Our study use reliable endpoint evaluating bronchoscopy. This could serve time-saving standard prospective studies comparing effectiveness various navigation devices.