Can sleep questionnaires predict outcome in children undergoing adenotonsillectomy for sleep disordered breathing?

Background: Paediatric sleep disordered breathing (SDB) affects 14% of children and is associated with adverse neurophysiological sequelae. Tonsillectomy +/− adenoidectomy (TA) is the most common first line treatment for SDB however objectively measuring outcome remains difficult. Polysomnography (PSG) is used to diagnose obstructive sleep apnoea, but is time consuming and difficult to access. This paper looks at the ability of sleep questionnaires to predict which children will benefit from TA for SDB as an alternative to PSG. Methods: A case-control study with children aged between 3 and 12 years placed on the surgical waiting list for TA secondary to SDB selected as the case group. Control participants consisted of children with no history of snoring. Forty-five children were included in the surgical group and 19 in the control. Both groups completed the sleep questionnaires OSA-18 and Paediatric Sleep Questionnaire (PSQ), as well as overnight PSG, after which the intervention group had their intended surgery. Both questionnaires and PSG were repeated postoperatively and compared. Results: The surgical group had higher preoperative scores for all three tests than the control group [apnoea-hypopnoea index (AHI): 2.5 vs. 0.7, difference (95% CI): 1.8 (1.0, 2.6), P<0.001; PSQ: 0.55 vs. 0.17, difference (95% CI): 0.38 (0.32, 0.45), P<0.001; OSA-18: 58 vs. 31, difference (95% CI): 27 (21, 33) P<0.001)]. Following surgery, scores in the surgical group all decreased [AHI (95% CI): −1.51 (−2.00, −1.02), P<0.001; PSQ (95% CI): −0.35 (−0.40, −0.30), P<0.001; OSA-18 (95% CI): −28 (−32, −33), P<0.001] but after adjusting for baseline score, only PSQ demonstrated a significant decrease in score following TA compared to controls [PSQ (95% CI): −0.14 (−0.24, −0.04), P<0.007; AHI (95% CI): 0.3 (−0.07, 0.66), P<0.106; OSA-18 (95% CI): −7 (−14, 0.5), P<0.069]. Children in the surgical group who scored less than the diagnostic cut off in PSQ showed no change in postoperative score. Conclusions: PSQ can be used to identify children with SDB who are likely to improve with TA, and of equal importance, those who are not.