The OSA-5: Development and validation of a brief questionnaire screening tool for obstructive sleep apnea in children

Objective: To develop and test a screening tool based on the OSA-18 questionnaire for triage of referrals for sleep-disordered breathing (SDB) in children. Study design: Consecutive children aged > 2y without major comorbidities referred for polysomnography (PSG) or overnight oximetry for suspected obstructive sleep apnea (OSA) between 11 January and 31 May 2017 were included. OSA was defined by an obstructive apnea/hypopnea index (OAHI) > levent/h on PSG or an abnormal overnight oximetry (McGill Oximetry Score 2-4). An 11-item questionnaire derived from a previous validation study of the OSA-18 underwent exploratory factor analysis (EFA) with varimax rotation. ANOVA identified questions associated with the presence of OSA. This informed a 5-question, 4-category instrument, scored 0-15 (the OSA-5), that was tested prospectively on 112 children having PSG. Results: 420 children (2.0-17.9y, 43% female) met the inclusion criteria, including 366 complete questionnaires. EFA resulted in a 3-factor structure. ANOVA identified 5 questions from one factor that were independently associated with a diagnosis of OSA: snoring, breath holding, choking, mouth breathing and parental concern. Mean OSA-5 scores with and without OSA were 7.7 vs 4.5 (p < 0.001). Thirty-four percent (60/178) had a total score < 5/15, with a sensitivity at this threshold for OSA of 82% and negative predictive value (NPV) of 70%. Similar results were obtained when tested prospectively, including a sensitivity of 82% and NPV of 81% for the presence of moderate/severe OSA (OAHI > 5/h). Conclusion: The OSA-5 is a simple questionnaire that performs well as a triage screening tool to identify those children at risk of OSA among large numbers of referrals for SDB.