Risk of failure of adenotonsillectomy for obstructive sleep apnea in obese pediatric patients.

INTRODUCTION: Pediatric obesity is a leading risk factor for obstructive sleep apnea (OSA), a condition commonly treated with adenotonsillectomy (T&A). It has been hypothesized that obesity increases a child's risk of failing T&A for OSA, however this relationship has not yet been quantified. The primary objective of this study was to investigate the relationship between obesity as measured by perioperative Body Mass Index (BMI) and persistent OSA following T&A as measured by polysomnography (PSG).

STUDY DESIGN: Retrospective chart review.

METHODS: Pediatric patients who underwent T&A between Jan. 2004 and Jan. 2016 were included. We recruited both obese and non-obese patients to compare caregiver/self reported improvement. Obese patients were recruited from a weight management clinic and included if they had a BMI z-score >1.65 and had pre- and post-operative polysomnograms (PSGs). Control patients included those undergoing T&A for OSA at our institution with BMI <1.65. These patients were age matched to the obese patient population. Age, gender, perioperative BMI z-score, caregiver/self reported improvement, total Apnea-Hypopnea Index (AHI), and O2 saturation nadir were collected where available. Univariate linear regressions were calculated between perioperative BMI z-score and PSG data.

RESULTS: 26 obese study and 47 control subjects were identified for analysis. T&A resulted in statistically significant improvements in total AHI (p = 0.030) and nadir O2 saturation (p = 0.013) in obese subjects. There was no significant difference between the rate of caregiver/self reported improvement in the two groups. There was a statistically significant correlation between perioperative BMI z-score and the change in total AHI (p = 0.049). Within our population, for every increase by 0.1 in perioperative BMI z-score, the improvement in total AHI post-operatively decreased by 1.63 events/hr. Further, patients with BMI more than 3 standard deviations away from the age-derived normative mean received essentially no benefit from T&A alone.

CONCLUSIONS: Our study established an inverse linear relationship between perioperative BMI z-score and improvement in total AHI with essentially no improvement in patients with BMI z-scores >3. Further studies are required to further elucidate this relationship and investigate the role of additional procedures in the initial management of OSA in obese children.