Predictors of improvement of radial-endobronchial ultrasonography findings from "adjacent to" to "within" in endobronchial ultrasonography using a guide sheath: a retrospective cohort study

Background: In flexible bronchoscopy, endobronchial ultrasonography using a guide sheath (EBUS-GS) has varying diagnostic yield depending on the findings of radial-endobronchial ultrasonography (R-EBUS). The diagnosis rate is lower when the ultrasound probe is "adjacent to", than when it is "within" the lesion. However, these findings are inconsistent, and the imaging status may change from "adjacent to" to "within" as examination progresses. In this study, we analyzed the predictive factors for this change, which remain unexplored till date. Methods: Patients who underwent flexible bronchoscopic biopsy with EBUS-GS at Kameda Medical Centre between 1 April 2014 and 31 March 2019 were included in this retrospective cohort study. Patients without "adjacent to" lesions were excluded. The appearance of "A to W" (the change from "adjacent to" to "within" imaging status) was the primary outcome. Based on multivariate regression and receiver operating characteristic curve analysis, we evaluated the discriminative properties of the factors strongly correlated with "A to W". Results: In total, 260 patients were included in this study. In 84 cases, the R-EBUS findings were "A to W". No such findings were observed in 176 cases. The mean lesion diameter was significantly larger (P=0.021) in the group with "A to W" than in the group without. The odds ratio [1.023 (1.003-1.046)] for lesion diameter showed statistical significance in the multivariable regression model. The sensitivity and specificity were 0.346 and 0.852, respectively, at the optimal threshold (29.25 mm) set using the Youden index. Conclusions: We found that lesion diameter was a significant factor in predicting "A to W", with a cut-off value of 29.25 mm and high specificity (0.852).