Ultrasound evaluation of the airway in the ED: a feasibility study

Recognition of the difficult airway is a critical element of emergency practice. Mallampati score and body mass index (BMI) are not always predictive and they may be unavailable in critically ill patients. Ultrasonography provides high-resolution images that are rapidly obtainable, mobile, and non-invasive. Studies have shown correlation of ultrasound measurements with difficult laryngoscopy; however, none have been performed in the Emergency Department (ED) using a consistent scanning protocol.

This study seeks to determine the feasibility of ultrasound measurements of the upper airway performed in the ED by emergency physicians, the inter-rater reliability of such measurements, and their relationship with Mallampati score and BMI.

A convenience sample of volunteer ED patients and healthy volunteers with no known airway issues, aged > 18 years, had images taken of their airway using a standardized ultrasound scanning protocol by two EM ultrasound fellowship trained physicians. Measurements consisted of tongue base, tongue base-to-skin, epiglottic width and thickness, and pre-epiglottic space. Mean and standard deviation (SD) were used to summarize measurements. Inter-rater reliability was assessed by intraclass correlation coefficients (ICCs). Analysis of variance with linear contrasts was used to compare measurements with Mallampati scores and linear regression with BMI.

Of 39 participants, 50% were female, 50% white, 42% black, with median age 32.5 years (range 19-90), and BMI 26.0 (range 19-47). Mean ± SD for each measurement (mm) was as follows: tongue base (44.6 ± 5.1), tongue base-to-skin (60.9 ± 5.3), epiglottic width (15.0 ± 2.8) and thickness (2.0 ± 0.37), and pre-epiglottic space (11.4 ± 2.4). ICCs ranged from 0.76 to 0.88 for all measurements except epiglottis thickness (ICC = 0.57). Tongue base and tongue base-to-skin thickness were found to increase with increasing Mallampati score (p = .04, .01), whereas only tongue-to-skin thickness was loosely correlated with BMI (r = .38).

A standardized ultrasound scanning protocol demonstrates that the airway can be measured by emergency sonologists with good inter-operator reliability in all but epiglottic thickness. The measurements correlate with Mallampati score but not with BMI. Future investigation might focus on ultrasound evaluation of the airway in patients receiving airway management to determine whether ultrasound can predict challenging or abnormal airway anatomy prior to laryngoscopy.