Airway Obstruction during Drug-Induced Sleep Endoscopy Correlates with Apnea-Hypopnea Index and Oxygen Nadir in Children

Objective: To determine if standardized intraoperative scoring of anatomic obstruction in children with obstructive sleep apnea correlates with the apnea-hypopnea index (AHI) and lowest oxygen saturation on preprocedural polysomnogram (PSG). A secondary objective was to determine if age, presence of a syndrome, or previous adenotonsillectomy affect this correlation.

Study design: Case series with chart review.

Setting: Two tertiary care children's hospitals.

Subjects: Patients with a preprocedural PSG who underwent drug-induced sleep endoscopy (DISE) over a 4-year period.

Methods: All DISEs were graded in a systematic manner with the Chan-Parikh (C-P) scoring system. AHI and nadir oxygen saturations were extracted from preprocedural PSG. Data were analyzed with a multivariate linear regression model that controlled for age at time of sleep endoscopy, syndrome diagnosis, and previous adenotonsillectomy.

Results: A total of 127 children underwent PSG prior to DISE: 56 were syndromic, and 21 had a previous adenotonsillectomy. Mean AHI was 13.6 ± 19.6 (± SD), and mean oxygen nadir was 85.4% ± 9.4%. Mean C-P score was 5.9 ± 2.7. DISE score positively correlated with preoperative AHI (r = 0.36, P < .0001) and negatively correlated with oxygen nadir (r = -0.26, P = .004). The multivariate linear regression models estimated that for every 1-point increase in C-P score, there is a 2.6-point increase in AHI (95% confidence interval: 1.4-3.8, P < .001) and a 1.1% decrease in the lowest oxygen saturation (95% confidence interval: -1.7 to -0.6, P < .001).

Conclusion: The C-P scoring system for pediatric DISE correlates with both AHI and lowest oxygen saturation on preprocedural PSG.