Diagnosis of Obstructive Sleep Apnea in Adults

My candle burns at both ends; It will not last the night; But ah, my foes, and oh, my friends It gives a lovely light. Edna St. Vincent Millay TO THE EDITOR: We read with concern the ACP clinical guideline about the diagnosis of OSA (1). This guideline offers PSG as a solitary tool for sleep-related symptoms; the guideline's broad application could result in unnecessary testing and treatment. Polysomnography performed with current technologies and scored using the criteria recommended by the AASM in 2012 will yield an average apneahypopnea index that is 3-fold higher than that obtained with the equipment and scoring criteria available at the time of most of the studies cited to support this guideline. Flow changes are currently graded using a pressure-transduced air flow monitorwhich is far more sensitive than the thermistor used in prior studiesand the new AASM guideline does not require oxygen desaturation to be present for a breathing event to be scored (2). In fact, a recent trial showed that the prevalence of an apneahypopnea index of 5 or more events per hour using the current criteria was 94.6% in a population with a mild-moderate pretest probability of OSA (3). This condition exists on a spectrum, and apneahypopnea index cutoffs are largely arbitrary. Given the changes in diagnosis, are we measuring clinically meaningful disease? What are the costs of overdiagnosis? Furthermore, the term unexplained sleepiness (which is pivotal in the guideline's first recommendation) is meaningful only when clinicians thoroughly understand the causes of sleepiness. The average physician receives approximately 2 hours of formal medical education on the evaluation of sleep disorders (4). This guideline fails to acknowledge that behaviorally induced insufficient sleep, insomnia, mood disorders, the restless legs syndrome, and many other problems cannot be measured by PSG or treated with continuous positive airway pressure. We know from survey data that the average person in the United States effectively burns the candle at both ends, obtaining 6 hours and 40 minutes of sleep on the average workday (5). Recommending PSG for every patient whose sleepiness is unexplainedwithout also recommending qualitative and quantitative assessment of sleep durationis ill-advised. We suggest that the ACP develop a comprehensive sleepsymptom guideline that encourages a more holistic evaluation of the patient who presents with sleepiness and focuses more attention on the limitations of PSG.