Detecting Aspiration During FEES with Narrow Band Imaging in a Clinical Setting

The use of narrow band imaging (NBI) during flexible endoscopic evaluation of swallowing (FEES) is recognised as an emerging technology to improve the contrast of the test fluid during endoscopic dysphagia evaluation. This study tested the hypothesis that the use of NBI in FEES would improve the detection of laryngeal penetration and aspiration in patients with unilateral vocal fold paralysis/paresis (UVFP), a typically difficult population in which to detect the presence of aspiration with FEES. Twenty-one consecutive outpatients with UVFP were evaluated with FEES using white light (WL) and NBI under 150 test conditions (75 WL & 75 NBI). Three speech pathologists, highly experienced in FEES using WL but novices to using NBI, rated laryngeal penetration and aspiration for green dyed thin fluid (5 ml and 90 ml) and mildly thick fluid (5 ml) milk, and were compared to two raters more experienced in using NBI during FEES. Laryngeal penetration and aspiration were significantly higher for larger volumes (90 ml) (p < 0.05). With NBI-naïve raters, there was a trend towards lower intra-rater and inter-rater reliability compared to WL on all bolus trials reaching significance on mildly thick fluid (p < 0.01). There was lower rater confidence when using NBI compared to WL in NBI-naïve raters to detect aspiration (p < 0.01). Sensitivity was lower regardless of NBI experience; 80.77-84.21% with WL compared to 46.15-50.00% with NBI. Findings indicate that the improved contrast of a dyed opaque milk trial under WL may negate the potential benefits of using NBI to increase the contrast of the test fluid and supports the use of an opaque test fluid such as milk. NBI may also not be as useful to clinicians with no experience with the altered light condition, and can result in lower sensitivity in even the experienced user.