Very Severe Obstructive Sleep Apnea in Children: Outcomes of Adenotonsillectomy and Risk Factors for Persistence.

Objectives (1) To describe the clinical, demographic and polysomnographic (PSG) characteristics of children with very severe obstructive sleep apnea (OSA) without significant comorbidities; (2) to assess the outcomes following tonsillectomy and adenoidectomy (T&A); and (3) to determine predictors of persistence of OSA after T&A. Study Design Case series with chart review. Setting Tertiary-level freestanding children's hospital. Subjects and Methods Seventy-four children aged 2 to 12 years who underwent T&A for very severe OSA (obstructive apnea-hypopnea index [AHI] >30) were included. Children with significant comorbidities were excluded. PSG variables were compared pre- and post-T&A using statistical tests. Factors affecting OSA resolution and persistence were studied. Results The mean (95% confidence interval) age was 4.3 (3.8-4.7) years with the majority black or Hispanic (64/74, 86%). The mean decrease in AHI after T&A was 49 (43-58) ( P < .001). Complete resolution of OSA, defined by an AHI <1, or an AHI <5 was seen in 32% (24/74) and 80% (59/74), respectively. Total sleep time (TST) greater than 5 minutes with end-tidal CO2 >50 mm Hg was strongly associated with persistent OSA. The decrease in AHI post-T&A was best predicted by higher preoperative oxygen saturation (SpO2) nadir and lower TST with SpO2 <90% ( R2 = 0.24, P < .001). Conclusions T&A is associated with a significant improvement but not resolution of very severe OSA. The severity of baseline hypercapnia and hypoxemia may best predict persistent OSA after T&A. The study supports obtaining routine post-T&A PSG in children with very severe OSA.