Does This Patient Have Obstructive Sleep Apnea? The Rational Clinical Examination Systematic Review

IMPORTANCE Obstructive sleep apnea is a common disease, responsible for daytime sleepiness. Prior to referring patients for definitive testing, the likelihood of obstructive sleep apnea should be established in the clinical examination. OBJECTIVE To systematically review the clinical examination accuracy in diagnosing obstructive sleep apnea. DATA SOURCES MEDLINE and reference lists from articles were searched from 1966 to June 2013. Titles and abstracts (n = 4449) were reviewed for eligibility and appraised for evidence levels. STUDY SELECTION For inclusion, studies must have used full, attended nocturnal polysomnography for the reference standard (n = 42). MAIN OUTCOMES AND MEASURES Community and referral-based prevalence of obstructive sleep apnea; accuracy of symptoms and signs for the diagnosis of obstructive sleep apnea. RESULTS The prevalence of sleep apnea in community-screened patients is 2% to 14% (sample sizes 360-1741) and 21% to 90% (sample sizes 42-2677) for patients referred for sleep evaluation. The prevalence varies based on the apnea-hypopnea index (AHI) threshold used for the evaluation (>= 5 events/h, prevalence 14%; >= 15/h, prevalence 6%) and whether the disease definition requires symptoms in addition to an abnormal AHI (>= 5/h with symptoms, prevalence 2%-4%). Among patients referred for sleep evaluation, those with sleep apnea weighed more (summary body mass index, 31.4; 95% CI, 30.5-32.2) than those without sleep apnea (summary BMI, 28.3; 95% CI, 27.6-29.0; P < .001 for the comparison). The most useful observation for identifying patients with obstructive sleep apnea was nocturnal choking or gasping (summary likelihood ratio [LR], 3.3; 95% CI, 2.1-4.6) when the diagnosis was established by AHI >= 10/h). Snoring is common in sleep apnea patients but is not useful for establishing the diagnosis (summary LR, 1.1; 95% CI, 1.0-1.1). Patients with mild snoring and body mass index lower than 26 are unlikely to have moderate or severe obstructive sleep apnea (LR, 0.07; 95% CI, 0.03-0.19 at threshold of AHI > 15/h). CONCLUSIONS AND RELEVANCE Nocturnal gasping or choking is the most reliable indicator of obstructive sleep apnea, whereas snoring is not very specific. The clinical examination of patients with suspected obstructive sleep apnea is useful for selecting patients for more definitive testing.