The efficacy of the OSA-18 as a waiting list triage tool for OSA in children.

Abstract

Sleep disordered breathing (SDB) is common in children, resulting in extensive waiting lists for specialist clinics. There is an urgent need for a valid method of triaging patients and the OSA-18, a disease-specific tool, is an attractive candidate for this role. We aimed to examine the OSA-18 as a measurement tool in detail and to determine whether the score or aspects of it could be used as a screening tool for SDB in children. Retrospective analysis of 582 children (6months to 16.4years)-216 underwent overnight PSG and 366 overnight oximetry. Confirmatory factor analysis (CFA) and exploratory factor analysis (EFA) were conducted. Receiver operating characteristic curve analysis assessed the diagnostic accuracy of the factors for the presence of OSA. Rasch analysis was used to assess the structure of the items (1-18) and categories of response (Likert scale). The CFA with a forced five-factor structure, revealed three factors with Eigenvalues >1, and explained 73.7% of the variance. EFA resulted in a two-factor structure, explaining 60.3% of the variance. Assessment of sensitivity and specificity showed a high false-positive rate, irrespective of the factor structure tested. Rasch analysis showed poor discrimination between adjacent categories on the Likert scale. This study confirmed that the predictive value of the OSA-18 for SDB severity is weak. Some questions perform better than others statistically, and the seven categories of response introduce significant statistical noise, raising the possibility that modification of the OSA-18 may improve its performance in the prediction of OSA severity.