SOS Score: An Optimized Score to Screen Acute Stroke Patients for Obstructive Sleep Apnea

Study Questions:

What is the accuracy of two OSA screening tools, the Berlin Questionnaire (BQ), and the Epworth Sleepiness Scale (ESS), when administered to relatives of acute stroke patients and compared against a combined screening score (SOS score)?


Ischemic stroke patients were submitted to a full polysomnography (PSG) at the first night after onset of symptoms. OSA severity was measured by apnea–hypopnea index (AHI). BQ and ESS were administered to relatives of stroke patients before the PSG and compared to the SOS score for accuracy and C-statistics.


The investigators prospectively studied 39 patients. OSA (AHI ≥10/hours) was present in 76.9%. The SOS score (area under the curve [AUC], 0.812; p = 0.005) and ESS (AUC, 0.789; p = 0.009) had good predictive value for OSA. The SOS score was the only tool with significant predictive value (AUC, 0.686; p = 0.048) for severe OSA (AHI ≥30/hours), when compared to ESS (p = 0.119) and BQ (p = 0.191). The threshold of SOS ≤10 showed high sensitivity (90%) and negative predictive value (96.2%) for OSA; SOS ≥20 showed high specificity (100%) and positive predictive value (92.5%) for severe OSA.


The authors concluded that the SOS score administered to relatives of stroke patients is a useful tool to screen for OSA, and may decrease the need for PSG in the acute stroke setting.


The results of this study confirmed the high frequency of sleep-disordered breathing in acute stroke patients. This study also showed that a structured interview of patient relatives concerning the presence of OSA-related complaints might provide useful clinical information to screen for OSA in the acute stroke setting. The SOS score, which combines elements from previous questionnaires, appears to be an appropriate tool to screen acute stroke patients for OSA, allowing stratification of these patients within categories of low, intermediate, and high risk of OSA. This risk stratification may then assist in developing therapeutic interventions to decrease the impact of OSA on stroke prognosis while decreasing the need for a formal sleep study in the acute stroke setting.

< Back to Listings