Utility of screening questionnaire and polysomnography to predict postoperative outcomes in children.

INTRODUCTION

The prevalence of pediatric obstructive sleep apnea (OSA) has increased concurrently with the increasing prevalence of obesity. We have previously validated a short questionnaire predicting the occurrence of OSA on polysomnography (PSG). This follow-up study assessed the utility of the questionnaire in predicting postoperative outcomes.

METHODS

Children undergoing surgery and completing a sleep study were prospectively screened for OSA using a short questionnaire. Procedures within 1 year of PSG were included in the analysis. Questionnaires were scored according to a cutoff previously deemed optimal for predicting OSA (apnea-hypopnea index ≥ 5) on the sleep study. Postoperative outcomes included prolonged (>60 min) length of stay (LOS) in the post-anesthesia care unit (PACU) and oxygen requirement in the PACU.

RESULTS

The study cohort included 185 patients (100/85 male/female) age 8 ± 4 years, undergoing adenotonsillectomy (n = 109), other ear, nose, and throat (ENT) procedures (n = 18), or non-ENT procedures (n = 58). There were 45 patients with OSA documented by PSG and 122 patients identified as likely to have OSA according to questionnaire responses (89% sensitivity, 41% specificity). PACU LOS was prolonged in 55/181 (30%) cases and supplemental oxygen was used in the PACU in 29/181 (16%) cases. In separate multivariable models, supplemental oxygen use in the PACU was more common if a patient scored ≥2/6 points on the short questionnaire scale (OR = 5.0; 95% CI: 1.3, 19.9; p = 0.023) or if the patient was diagnosed with OSA on PSG (OR = 4.6; 95% CI: 1.6, 13.5; p = 0.005). Neither OSA on PSG nor questionnaire score ≥2/6 were associated with prolonged PACU stay.

CONCLUSION

Both OSA diagnosis based on the AHI and the questionnaire scale achieved comparable predictive value for the need for oxygen use in the PACU. The utility of the questionnaire in predicting rare adverse events (e.g., unplanned admission or rapid response team activation) remains to be determined. Our preliminary results support using a brief questionnaire scale for preoperative risk stratification among children with suspected OSA who have not had a formal sleep study.