Psychometric Properties of Obstructive Sleep Apnea Screening Measures in Patients Referred to a Sleep Clinic

Background: Obstructive Sleep Apnea (OSA) contributes to all-cause and cardiac mortality. There are no current guidelines for OSA screening in outpatient settings. An American Academy of Sleep Medicine task force is focusing on improving detection and categorization of OSA symptoms and severity to promote screening, assessment, and diagnosis of the disorder. The purpose of this study was to identify the psychometric properties of three self-report OSA screening measures (Berlin, Epworth Sleepiness Scale (ESS), STOP Bang) and an objective portable sleep monitor (PSM) compared to apnea-hypopnea index (AHI) levels (≥5, ≥ 15, and ≥ 30) from polysomnogram (PSG). Methods: A methodological design was used. Patients referred to a sleep specialist for an OSA consultation were recruited and enrolled at initial sleep evaluation. Participants completed the three OSA self-report screening measures and those participants who met inclusion criteria were sent home with a PSM for one night measurement. Automatic scoring was used. PSGs were ordered by the physician and AHI results were obtained from the medical record. Results: Participants (N=170) were enrolled (88 male, 82 female; age 54.5, SD 5.0 years). Almost all participants completed the self-report OSA screening measures, approximately half completed PSM measurement, and the majority completed laboratory PSG. The STOP Bang had the highest levels of sensitivity; the ESS had the lowest. The ESS had the highest specificity and reliability level. The PSM measure had the highest positive predictive value (PPV). The PSM measure had the strongest psychometric properties of the screening measures. Conclusions: The STOP Bang was the preferred self-report OSA screening measure because of high sensitivity levels. A positive STOP Bang warrants assessment for OSA. The ESS is the least desirable screening measure. If a patient qualifies, further screening with a PSM is indicated. PSM measurement consistently predicted the presence of OSA but at the expense of low sensitivity at AHI levels ≥ 30. PSM results can guide the referral process from primary or specialty clinicians to sleep specialist