Prediction of obstructive sleep apnea using GOAL questionnaire in adults with or without excessive daytime sleepiness: A cross-sectional study

Abstract

ObjectiveObstructive sleep apnea (OSA) is a prevalent disorder and excessive daytime sleepiness (EDS) is a frequently associated complaint. Our objectives were (1) to evaluate the predictive performance of the GOAL questionnaire in patients with or without EDS and (2) to compare its discrimination with that of 3 other validated instruments: STOP-Bang, No-Apnea, and NoSAS. MethodsA cross-sectional study from July 2019 to March 2020 was designed with consecutive adults undergoing full polysomnography. Participants were grouped into 2 cohorts according to the Epworth Sleepiness Scale (ESS): without EDS (ESS <11) or with EDS (ESS ≥11). Discrimination was assessed by area under the curve, while predictive parameters were calculated using contingency tables. OSA severity was classified based on apnea/hypopnea index thresholds: ≥ 5.0/h (OSA≥5), ≥ 15.0/h (OSA≥15), and ≥ 30.0/h (OSA≥30). ResultsOverall, 2627 patients were enrolled. In 1477 individuals without EDS, for the screening of OSA≥5, OSA≥15, and OSA≥30, GOAL showed sensitivities ranging from 79.3% to 91.5% and specificities ranging from 60.6% to 40.2%. In 1150 individuals with EDS, for the screening of OSA≥5, OSA≥15, and OSA≥30, GOAL reported sensitivities ranging from 86.2% to 94.5% and specificities ranging from 63.9% to 36.0%. In both cohorts, GOAL showed similar discrimination to STOP-Bang, No-Apnea, and NoSAS for predicting OSA≥5, OSA≥15, and OSA≥30 (all P values >.05). ConclusionsThe GOAL questionnaire, a practical 4-item instrument, showed adequate predictive performance for the prediction of OSA. Moreover, its discrimination was satisfactory and non-inferior to that of STOP-Bang, No-Apnea, and NoSAS.