Predicting the Airway Patency using the Parameters of Soft-tissue Lateral Neck Radiography in Adult Patients with Acute Epiglottitis

ABSTRACT

PURPOSE: We wanted to predict the high risk group that requires urgent airway intervention by using the parameters of the soft-tissue lateral neck radiographs of adult acute epiglottitis patients. METHODS: This retrospective study was conducted in two teaching hospitals. The patients who were diagnosed with acute epiglottitis from June, 2007 to May, 2009 were enrolled and their medical records and x-ray films were reviewed. The width of the epiglottis at the widest point (EW), the width of the arytenoid at the widest point (AW), the prevertebral soft tissue distance at the third cervical spine (PSTD), the shortest distance from the epiglottis to the hypopharyngeal wall (EHD) and the shortest distance from the epiglottic root to the arytenoids'tip (EAD) were investigated and we performed regression analyses of these parameters of the patients in the high risk group that required urgent airway intervention. RESULTS: A total of 42 patients were enrolled. Dyspnea and hoarseness were more frequent in the high risk group that required urgent airway intervention (p=0.008, 0.040, respectively). The EW was significantly longer (p=0.001) in the high risk group. The EHD and EAD were significantly shorter (p=0.012, <0.001, respectively) in the high risk group. Only the EAD showed significant correlation with the percent of airway patency on linear regression analysis (p=0.003) and the EAD was the only significant predictor for the high risk group on multivariate logistic regression analysis (p=0.043). The receiver operating characteristics curve of the EW/EAD for the high risk group was obtained and it showed the best predictive power (AUC 0.977, p<0.001). CONCLUSION: The EAD noted on soft-tissue lateral neck radiography is an important predictor of high risk patients who require urgent airway intervention. The cut-off value of the EW/EAD for the predicting the high risk group is 2.44 (sensitivity 100%, specificity 85.7%).