Abstract
Objectives:
(1) Determine the association of a sleep endoscopy rating scale with pediatric obstructive sleep apnea (OSA) severity. (2) Assess the reliability of this rating scale.
Methods:
This was a retrospective cohort study of pediatric patients who underwent drug-induced sleep endoscopy (DISE) from January to December 2013 at a tertiary care children’s hospital. DISE recordings were reviewed by 2 blinded senior pediatric otolaryngologists. Severity of obstruction was scored 0, 1, or 2 for no, partial, or complete obstruction at 6 levels: nasal airway, nasopharynx, velum, oropharynx, tongue base, and arytenoids. Ratings at each level were summed for a total score from 0-12. Associations of DISE scores with obstructive apnea-hypopnea index (OAHI) and OSA-18 scores were assessed using linear regression. Interrater and intrarater reliability were calculated using a kappa statistic with linear weighting.
Results:
Twenty patients were included (mean age 8.4 ± 5.1 years, 55% obese, mean OAHI 16.5 ± 15.6). Associations between OAHI and DISE scores were strongest for the nasopharynx (beta = 12.3, P = .09) and total score (beta = 4.7, P = .06). Associations between OSA-18 and DISE ratings were strongest for the nasopharynx (beta = 11.1, P = .08) and velum (beta = 12.8, P = .07). There was fair-moderate interrater reliability (kappa 0.24-0.63) and moderate-almost perfect intrarater reliability (kappa 0.52-0.88) at individual anatomic sites.
Conclusions:
Although observed associations did not reach statistical significance, this pilot study suggests clinically important associations between DISE ratings and OSA severity and showed promising interrater and intrarater reliability.