Category Archives: sleep apnea

Does adenotonsillectomy improve outcomes in children with obstructive sleep apnea?

Bottom Line

Early adenotonsillectomy and watchful waiting both options for children with sleep apnea.  Early surgery provides some symptomatic benefit and greater normalization of polysomnographic findings (a disease-oriented outcome) than watchful waiting for children with obstructive sleep apnea (OSA). However, many of the children in the watchful-waiting group improved during the study period, and there were no cognitive or behavioral consequences to watchful waiting. Thus, either approach is a reasonable option. (LOE = 1b)


Marcus CL, Moore RH, Rosen CL, et al, for the Childhood Adenotonsillectomy Trial (CHAT). A randomized trial of adenotonsillectomy for childhood sleep apnea. N Engl J Med 2013;368(25):2366-2376.

  • Study Design: Randomized controlled trial (single-blinded)
  • Funding: Government
  • Setting: Outpatient (specialty)
  • Allocation: Concealed


More than half a million children in the United States still get their tonsils and adenoids removed each year, with the most common indication being OSA. This multicenter trial included children, aged 5 years to 9 years, with an apnea-hypopnea index (AHI) score of 2 or more events per hour or an obstructive apnea index (OAI) score of at least 1 event per hour. The authors excluded children who had severe OSA, characterized by an AHI score greater than 30 per hour, an OAI score greater than 20 per hour, or an O2 saturation of less than 90% for 2% or more of sleep time. Children with significant obesity (the top 0.13% of body mass index for age), those taking medications for attention-deficit/hyperactivity disorder (ADHD), and those with recurrent tonsillitis were also excluded. The authors recruited a total of 464 children — who were randomized to receive adenotonsillectomy within 4 weeks (early surgery group) or watchful waiting — and followed them up for 7 months. Sixteen children in each group did not receive the assigned treatment. There were 30 patients lost to follow-up or who withdrew from the early surgery group; 23 in the watchful waiting group. Groups were similar at the beginning of the study: approximately half the children were girls, slightly less than half were obese or overweight, and slightly more than half were black. The outcomes were mixed. The primary outcome of measures of attention and executive function by masked outcome assessors showed no significant difference between groups. Unmasked assessment of behavior by parents using the Connors’ scale for ADHD found greater improvement in the early surgery group, as did the parental assessment of executive function and behavior. Unmasked assessment of symptoms and quality of life showed slightly greater improvement in the early surgery group. Polysomnographic abnormalities improved more in the early surgery group (from 4.8 to 1.3 per hour in the early surgery group and from 4.5 to 2.9 in the watchful-waiting group). However, it’s important to note that these differences were small and did not meet the usual cutoff for a clinically important difference. In general, a “clinically important difference” requires a change of at least 10% to 15% on a rating scale, and the above improvements were closer to a 5% difference between groups on the scale.