The most appropriate next step is in-laboratory polysomnography. This patient has risk factors for both obstructive sleep apnea (OSA) (male gender, snoring, overweight) and central sleep apnea (CSA) (heart failure and atrial fibrillation). The lack of history from a bed partner hampers determination of pretest probability. Although out-of-center (performed outside of a sleep laboratory or at home) sleep testing is appropriate for patients with a high pretest probability of uncomplicated moderate to severe OSA, those with heart failure or advanced pulmonary disease or those at risk for CSA should undergo in-laboratory polysomnographic diagnostic testing. In-laboratory polysomnography allows more detailed analysis of the possible underlying disorder than out-of-center methods. Once the type of apnea is clarified during the diagnostic portion of the in-laboratory study, the technician may then utilize the most appropriate mode of positive airway pressure (PAP) therapy and assess the response to treatment. Treatment options include continuous positive airway pressure (CPAP) for OSA, adaptive servoventilation (ASV) for CSA, and the addition of supplemental oxygen, if needed, for gas exchange abnormalities due to chronic heart failure or lung disease.
In this patient, auto-titrating positive airway pressure (APAP) would not be appropriate before diagnostic testing is performed to clarify the predominant type of apnea. APAP is indicated for treatment of OSA but is unlikely to be an effective treatment for CSA or complex sleep apnea, if present.
Out-of-center sleep testing should be used in patients without comorbid cardiopulmonary disease who are likely to have OSA of at least moderate severity, such as in an obese middle-aged man who snores loudly, pauses breathing, and gasps during sleep.
This patient has symptoms of interrupted sleep and daytime sleepiness, which are strong indications for treatment of sleep-disordered breathing. Further screening with overnight pulse oximetry is unlikely to add reliably important diagnostic information, and it will not alter the decision to treat.