Adenotonsillectomy for obstructive sleep apnea in children: outcome evaluated by pre- and postoperative polysomnography.

OBJECTIVE: To evaluate the outcome of adenotonsillectomy for obstructive sleep apnea (OSA) in children using objective data from polysomnography supplemented by subjective proxy reports from the OSA-18 quality of life instrument. STUDY
DESIGN: Prospective cohort study.
METHODS: Children 3 to 14 years of age with OSA diagnosed principally on the basis of polysomnography as having an obstructive apnea/hypopnea index (AHI) of 5 or greater underwent adenotonsillectomy. OSA was classified as mild (AHI > or = 5 < 10), moderate (AHI > or = 10 < 20), or severe (AHI > or =20). Children enrolled in the study also had postoperative polysomnography 3 to 6 months after surgery. Caregivers completed the OSA-18 survey before surgery and within 6 months after surgery. Pearson correlation was used to compare the pre- and postoperative AHI values with the pre- and postoperative OSA-18 total scores. SAS procedures (SAS Corp., Cary, NC) were used for statistical analyses. A P value less than or equal to .05 was considered significant.
RESULTS: The study population included 79 healthy children, 40 of who were male. The mean age was 6.3 (range, 3.0-14.0) years. Only tonsillar size was correlated significantly with a high preoperative AHI. For all children, the preoperative AHI value was higher than the postoperative value. The mean preoperative AHI for the study population was 27.5, whereas the mean postoperative AHI was 3.5. This change was highly significant (P < .001). The percentage of children with normal polysomnography parameters after adenotonsillectomy ranged from 71% to 90% as a function of the criteria used to define OSA. It was highest when an obstructive apnea index less than 1 was used and lowest when an AHI less than 1 was used to define resolution of OSA. Overnight respiratory parameters after adenotonsillectomy were normal for all children with mild OSA. Three (12%) children with moderate preoperative OSA, and 13 (36%) children with severe preoperative OSA had persistent OSA after adenotonsillectomy. Resolution of OSA occurred in all children with a preoperative AHI less than or equal to 10 and in 73% of children with a preoperative AHI greater than 10. The mean total OSA-18 score and the mean scores for all domains showed significant improvement after surgery (P < .001). The preoperative AHI values had a fair correlation with the preoperative total OSA-18 scores (r = 0.28), but postoperative AHI values had a poor correlation with the postoperative total OSA-18 scores (r = 0.16). Caregivers reported snoring some, most, or all of the time in 22 (28%) children; this group included all children with persistent OSA.
CONCLUSIONS: Adenotonsillectomy for OSA results in a dramatic improvement in respiratory parameters as measured by polysomnography in the majority of healthy children. Quality of life also improves significantly after adenotonsillectomy for OSA in children. However, the correlation between improvements in respiratory parameters and improvements in quality of life is poor. Severe preoperative OSA is associated with persistence of OSA after adenotonsillectomy. Postoperative reports of symptoms such as snoring and witnessed apneas correlate well with persistence of OSA after adenotonsillectomy.