Prevalence and Detection of Obstructive Sleep Apnea Early after Stroke

ABSTRACT Background Obstructive sleep apnea (OSA) negatively impacts post-stroke recovery. This study’s purpose: examine the prevalence of undiagnosed OSA and describe a simple tool to identify those at-risk for OSA in the early phase of stroke recovery. Methods This was a cross-sectional descriptive study of people ∼15 days post-stroke. Adults with stroke diagnosis admitted to inpatient rehabilitation over a 3-year period were included if they were alert/arousable, able to consent/assent to participation, and excluded if they had a pre-existing OSA diagnosis, other neurologic health conditions, recent craniectomy, global aphasia, inability to ambulate 150 feet independently pre-stroke, pregnant, or inability to understand English. OSA was deemed present if oxygen desaturation index (ODI) of >=15 resulted from overnight oximetry measures. Prevalence of OSA was determined accordingly. Four participant characteristics comprised the “BASH” tool (body mass index >=35, age>=50, sex=male, hypertension=yes). A receiver operator characteristics (ROC) curve analysis was performed with BASH as test variable and OSA presence as state variable. Results Participants (n=123) were 50.4% male, averaged 64.12 years old (sd 14.08), and self-identified race as 75.6% White, 20.3% Black/African American, 2.4%>1 race, and 1.6% other; 22% had OSA. ROC analysis indicated BASH score >=3 predicts presence of OSA (sensitivity=0.778, specificity=0.656, area under the curve =0.746, p<0.001). Conclusions Prevalence of undiagnosed OSA in the early stroke recovery phase is high. With detection of OSA post-stroke, it may be possible to offset untreated OSA’s deleterious impact on post-stroke recovery of function. The BASH tool is an effective OSA screener for this application.