Clinical validation and reproducibility of endoscopic airway measurement in pediatric aerodigestive evaluation

Background: Endoscopic airway measurement (EAM) combines optical endoscopic instruments with open source image processing to accurately obtain airway dimensions. Preclinical models have demonstrated EAM as an accurate technique of airway measurement with the added advantage of characterizing multilevel stenosis, non-circular lesions, and distal obstruction. The aim of this prospective clinical study was to compare EAM to airway measurements obtained from endotracheal tube approximation (ETTA) during pediatric aerodigestive evaluation and to evaluate reproducibility of EAM across practitioners.

Methods: Thirty-seven pediatric patients undergoing routine microlaryngoscopy and bronchoscopy at a single tertiary care children's hospital were prospectively recruited. Patients undergoing emergent procedures were excluded. Two blinded reviewers performed airway measurements using ImageJ (NIH) as previously described and average values were compared to ETTA measurements. Additional EAMs were obtained from an ex vivo airway model by 28 separate clinicians and were analyzed by the same reviewers to evaluate reproducibility.

Results: EAM and ETTA measurements were themselves significantly different (p = 0.0003); however, the average absolute difference between the two methods was small (Mean: 0.5 mm, 95%CI: -2.6-1.6 mm). There were notable differences between raters such that estimates of raters with more experience were more similar to ETTA. Despite observed differences between EAM and ETTA, endoscopic airway measurement was highly correlated with ETTA (p = 0.0002, Spearman r = 0.4185), and strong agreement was observed (Bias: -0.4974 ± 1.083 mm, 95% LOA: -2.62-1.625 mm).

Conclusion: Clinical use of EAM is a valid and precise approach for quantification of airway luminal dimensions. This method may provide advantages over traditional ETTAs for evaluation of asymmetric airway morphology in the pediatric population.