Using a modified version of the “STOP-BANG” questionnaire and nocturnal oxygen desaturation to predict obstructive sleep apnea after stroke or TIA

Obstructive sleep apnea (OSA) is a risk factor and common morbidity for stroke and transient ischemic attack (TIA). However, screening for OSA in patients with stroke or TIA is uncommonly performed, due in part to difficulties associated with conducting polysomnography (PSG) and Home Sleep Apnea Tests (HSATs). The 8-point “STOP-BANG” questionnaire has been shown to have high methodological quality in screening for OSA. This study examined the clinical utility of a modified version of the “STOP-BANG” questionnaire, which removed neck circumference and included nocturnal oxygen desaturation in diagnosing OSA (ie, the “STOP-BAG-O” tool), with the goal of improving uptake and accuracy in diagnosing OSA. In total, 231 participants completed both the STOP-BAG questionnaire and PSG or HSAT within 12 months of stroke/TIA. Using receiver-operating curves, scores on the “STOP-BAG-O” and “STOP-BAG” questionnaires were assessed for their ability to predict a diagnosis of OSA and classify at least 50% of the study population. Compared to an OSA diagnosis of AHI≥10, the STOP-BAG (using cut-offs of ≤3 and ≥4) had a sensitivity and specificity of 83.5% and 67.2%, respectively. The STOP-BAG-O (using cut-offs of ≤3 and ≥5) had a sensitivity and specificity of 95.9% and 78.4%, respectively. For all AHI cut-offs used, the area under the curve for the STOP-BAG-O was greater and statistically different (p