Berlin questionnaire performance for detecting sleep apnea in the general population

Introduction: Berlin questionnaire (BQ) has been proposed as a screening tool for identifying patients at risk for obstructive sleep apnea (OSA). The aim of our study is to evaluate the performance of this questionnaire for detecting OSA in a large sample of middle-aged general population. Methods: 469 subjects (46.4% women, 50.6±7.4 years old, BMI 25.1±4.9 kg/m 2 ) participating in an ongoing population-based cohort study (HypnoLaus, Lausanne, Switzerland) underwent a complete polysomnographic recording at home and an extensive clinical workup including BQ. This instrument includes 3 categories: 1) witnessed apnea and snoring 2) daytime sleepiness and 3) obesity or hypertension. A positive score in 2 or more categories was considered suggestive of OSA. Results: Mean AHI was 6.3±10.6/h. Mean 4%ODI was 5.8±10.0. Prevalence of OSA defined as an AHI >5/h, >15/h and >30/h was 33.3%, 10.3% and 3.8%, respectively in our population. Prevalence of positive BQ score was 24.4% (29.2 in men, 18.81 in women). BQ sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) to detect OSA were 36.1%, 82.4%, 51.8% and 72.9% respectively for an AHI >5/h; 52.1%, 78.8%, 21.9% and 93.5% for an AHI>15; and 72.2%, 77.6%, 11.4% and 98.6% for an AHI>30/h. Positive BQ was associated with higher 4% ODI (10.8/h vs 4.2/h p Conclusion: BQ questionnaire performance for identifying OSA is lower in a middle-aged general population than previously reported in a clinical population. Our results do not support its use as a screening tool for OSA in an unselected population.