Diagnostic outcomes of radial endobronchial ultrasound bronchoscopy guided by manual navigation in the evaluation of peripheral pulmonary lesions: An observational study.

Abstract

Manual navigation (MN), drawing a bronchoscopic road map simply by looking at the consecutive computed tomography (CT), is feasible and economical. However, scant data about the use of MN in radial endobronchial ultrasound (r-EBUS) bronchoscopy have been documented till now. We aimed to evaluate the diagnostic performance of r-EBUS bronchoscopy guided by MN for diagnosing peripheral pulmonary lesions (PPLs) and to determine clinical factors affecting the diagnostic yield. We performed a retrospective, cohort study of consecutive patients with PPLs who underwent r-EBUS bronchoscopic biopsy via guidance of MN from May 2020 to June 2021 in our Respiratory Endoscopic Division. The overall diagnostic yield of MN-guided r-EBUS, the factors affecting the yield, and the diagnostic performance for malignancy were evaluated. A total of 102 patients (103 lesions) were evaluated. The overall diagnostic yield of MN-guided r-EBUS was 82.0%, and it ranged from 79.6% to 82.5%, assuming the undermined cases were all positive cases (79.6%) or negatives (82.5%). The sensitivity of MN-guided r-EBUS for malignancy was 71.4%, ranging from 68.2% to 71.4%, the specificity was 100%, the positive predictive value was 100%, and the negative predictive value was 67.3%, ranging from 63.8% to 69.0%. The multivariate logistic regression showed that "bronchus sign on CT" was the only predictor of the overall diagnostic yield (odds ratio = 11.5, 95% confidence interval: 1.9-70.9, P = 0.009). MN-guided r-EBUS is feasible in diagnosing PPLs, especially for lesions with bronchus sign on CT.