Introduction: Airway assessment is an important step in evaluating feasibility and prognosis of extubation in critically ill patients. However, the common clinical screening tests lack sensitivity and specificity. Ultrasound is a quick. non-invasive, and inexpensive tool. The aim of this study is to determine whether measurements of airway structures by ultrasound can predict failure on extubation. Methods: ICU patients who were intubated were eligible for inclusion. Exclusion criteria were prone positioning, extubation within 24 hours, presence of cervical collar, neck surgery or masses. Data were obtain using a linear or curvilinear ultrasound probe. Measurements included neck circumference, skin-to-epiglottis depth, epiglottis width, and tongue width. Measurements were taken at three timepoints for each patient. Pairwise associations were measured using logistic regression between binary outcome (reintubation, prolonged intubation, and tracheostomy) and continuous features (US measurements), and OLS regression between continuous outcomes (fluid balance) and features (gender, BMI, etc). Bonferroni corrections were conducted for multiple measurements. The primary endpoint was reintubation. Secondary endpoint was the patient’s fluid balance between each day’s measurements. Results: 50 patients were included in the analysis. The primary endpoint of re-intubation occurred in 8% of patients. Overall, there were no clear trends between the various neck measurements and the re-intubation frequency. Associations between neck circumference (mean=42.6cm (SD 7.1)) on day three and fluid balance on day three (mean=536ml (SD 2335)) trended positively when controlling for age (p< 0.04), BMI (p< 0.01), and SOFA score (p< 0.02). Odds ratios analyzing various independent factors for re-intubation, prolonged ventillation, or need for tracheostomy, did not demonstrate statistical significance. Conclusions: Assessment of airway anatomy to predict the success of extubation is crucial in critical care clinical scenarios. This pilot study ultimately lacked the statistical power to identify clear associations; however, it did demonstrate that neck circumference can be used to indicate positive fluid balance on day three. Further investigation remains to develop such a predictive tool utilizing point-of-care ultrasound.
Sweetser et al. (Sun,) studied this question.