Fiber-optic sleep endoscopy in children with persistent obstructive sleep apnea: Inter-observer correlation and comparison with awake endoscopy

Objective: Evaluate the inter-observer correlation of sleep endoscopy findings in children with persistent obstructive sleep apnea (OSA) with awake office fiber-optic endoscopy. Method: Design: retrospective case series; blinded review. Setting: tertiary care children's hospital. Patients: Children with persistent obstructive sleep apnea. Interventions: Both awake and drug induced sleep endoscopy were performed. Endoscopy video recordings were mixed at random on a DVD. Two pediatric otolaryngologists and two pediatric pulmonologists independently scored each recording using an upper airway endoscopy scoring survey. Main outcome measures: reviewers scored the following parameters: each structure's contribution (nose, nasopharynx, lateral pharyngeal walls, tongue base, supraglottis) to the obstruction, the main site in which the obstruction occurs, the severity of OSA (mild, moderate, severe), the level of confidence of endoscopy findings (poor, fair, good). Results: When reviewing sleep endoscopy recordings for the upper airway obstruction site, the highest correlation among the four observers was found for the nasopharynx and the supraglottis (Kappa score: 0.6 and 0.5, respectively). Compared to awake endoscopy, sleep endoscopy demonstrated more cases of airway obstruction caused by collapse of lateral pharyngeal walls and base of tongue (McNemar test for symmetry, P value < 0.05). Level of confidence among the four observers was higher in older children and lower in children with severe OSA. Conclusions: Sleep endoscopy is a consistently reliable tool for identifying the site of obstruction in children with persistent OSA. Though anesthetic induced sleep is not a perfect model for real sleep, the technique demonstrably guides further therapy better than awake endoscopy.