An ultrasound evaluation of laryngeal mask airway position in pediatric patients: an observational study

Abstract

BACKGROUND: In children, the laryngeal mask airway (LMA) is frequently displaced within the hypopharynx, resulting in repositioning of the device. When the tip of the LMA is placed in the esophageal inlet, the arytenoids are moved ventrally. When the LMA is rotated or deviated, the ventral movement of the arytenoids may result in asymmetric elevation of an arytenoid cartilage, which can be detected with ultrasound (US). In this study, we sought to estimate the incidence of LMA malposition detected with US in pediatric patients. The primary end point was to compare the incidence of LMA malposition between US and fiber optic bronchoscopy (FOB). The secondary end points were to find the interrelationship between US-detected and FOB-detected malposition of the LMA and to locate the diagnostic performance of US in detecting LMA malposition.

METHODS: In this observational study, 100 consecutive children were included. After anesthetic induction, US evaluation was performed before and after LMA insertion to obtain the glottic image on the anterior neck. FOB was performed to assess LMA position (FOB LMA grade and LMA rotation grade). With a post-LMA US image, the symmetry of the arytenoid cartilages was evaluated. Asymmetrical elevation of an arytenoid cartilage in reference to the glottic midline and the opposite arytenoid cartilage was graded as 0 to 3 (US arytenoid grade). The interrelationships between US arytenoid grade and FOB LMA grade or LMA rotation grade were assessed.

RESULTS: The incidence of asymmetrical elevation of an arytenoid was 50% (95% confidence interval [CI], 40%-60%). On FOB, the incidence of LMA malposition was 78% (95% CI, 69%-86%), and that of LMA rotation was 43% (95% CI, 33%-53%). The incidence of LMA malposition was higher with FOB (P < 0.0001), but the incidence of rotation was similar (P = 0.395). US arytenoid grade did not correlate with FOB LMA grade (P = 0.611) but showed a significant correlation with LMA rotation grade (P < 0.0001; 95% CI, 60%-83%). To detect a rotated LMA, US had a sensitivity of 93% (95% CI, 81%-98%) and a specificity of 82% (95% CI, 70%-91%). The positive and negative predictive values were 80% (95% CI, 66%-90%) and 94% (95% CI, 83%-99%), respectively. The accuracy was 87% (95% CI, 79%-93%).

CONCLUSIONS: Although US could not detect the suboptimal depth of an LMA, US has promise of being an accurate tool in detecting a rotated LMA