The value of screening questionnaires/scoring scales for obstructive sleep apnoea in patients with atrial fibrillation

Abstract

Obstructive sleep apnoea (OSA) is an important modifiable risk factor for atrial fibrillation (AF) but is underdiagnosed in this population. Currently, polysomnography is the gold standard for diagnosing OSA, but is expensive and requires overnight examination. Alternatively, home sleep apnoea testing can be used as a diagnostic tool, but also requires a complete data review. Therefore, these OSA diagnostic modalities are not ideal screening methods. Several OSA screening tools exist, but their value in patients with AF remains unclear. To test the performance of existing screening questionnaires/scales for clinically relevant OSA in patients with AF referred for diagnostic polysomnography. This prospective study compared the performance of seven screening tools (Epworth Sleepiness Scale, Berlin Questionnaire, Sleep Apnea Clinical Score, NoSAS, OSA50, STOP-Bang and MOODS) with polysomnography in the detection of clinically relevant OSA in consecutive patients with AF referred to two sleep clinics. A total of 100 patients referred for polysomnography and known previous AF were included. Polysomnography indicated at least clinically relevant OSA (i.e., apnoea-hypopnoea index≥15 events/hour) in 69% of cases, and 33% had severe OSA (apnoea-hypopnoea index>30 events/hour). In screening for clinically relevant OSA, only the SACS and NoSAS scores had fair areas under the curve (0.704 and 0.712, respectively). None of the seven screening tools was performant enough (i.e., had a fair area under the curve>0.7) in the detection of severe OSA. In this AF cohort referred for polysomnography, clinically relevant OSA was prevalent. None of the selected screening tools showed sufficient performance as a good discriminative screening tool for clinically relevant OSA in patients with AF. Given these findings, other screening modalities for OSA should be considered in the work-up of patients with AF.