Assessing the likelihood of obstructive sleep apnea: a comparison of nine screening questionnaires.

Abstract

Several questionnaires are available for the screening of obstructive sleep apnea (OSA). Herein, we compare the performance characteristics of nine available questionnaires for assessing the likelihood of OSA. Consecutive subjects who underwent polysomnography at the sleep laboratory of the unit were included. Subjects with obstructive events and apnea hypopnea index (AHI) ≥5 were considered to have OSA. The likelihood ratios (LRs) and other performance characteristics were calculated for the following nine questionnaires: Berlin, modified Berlin, STOP, STOP-Bang and OSA50 questionnaires, sleep apnea clinical score (SACS), Epworth sleepiness scale (ESS), American Society of Anesthesiologists (ASA) checklist, and the elbow sign questionnaire. Two-hundred and ten subjects (mean age, 46.5years; mean body mass index [BMI], 31.9kg/m2; 27.1% women) were included. OSA was diagnosed in 78.1% of patients; 49.5% had severe OSA (AHI ≥30). The SACS questionnaire had the highest positive LR (LR+, 5.6) and positive predictive value (95.2%). The modified Berlin questionnaire had the best negative LR (LR-, 0.2) and the highest negative predictive value (57.1%). The STOP-Bang questionnaire also had an LR- of 0.2 if BMI threshold of 25kg/m2 (like that in the modified Berlin questionnaire) was used. Among individual items of various sleep questionnaires, the highest LR+ was obtained for neck circumference >43cm (LR+, 4.9), while the best LR- was obtained for snoring and BMI >25kg/m2 (LR-, 0.2). The SACS and the STOP-Bang questionnaires (BMI threshold of 25kg/m2) were found to provide the best positive and negative LRs, respectively, for the prediction of OSA. We believe that information from these questionnaires may help in prioritizing patients for sleep studies in high-volume centers.