There is but one randomized controlled trial that actually evaluated a simplified model of care for OSA in a primary care setting from Australia.15 Although as-yet only presented in abstract form, both my debater and I had a chance to view the poster presentation at the American Thoracic Society 2012 International Conference in San Francisco, California. The sleep-specialist primary investigators introduced a screening program to a group of PCPs using the Epworth sleepiness scale (ESS), reviewed symptoms suggestive of OSA with them, and then instructed them with regard to use of an HST to confirm OSA. Qualifying patients with OSA were randomized to compare management with either usual care with their management guided by a sleep physician in a specialty sleep center with laboratory-based testing or management led by their PCP with a community-based nurse and autotitrating CPAP done in the home after positive HST. The outcome measures included symptom reevaluation with the ESS (primary outcome) as well as change in functional outcomes of the sleep questionnaire, and the objective CPAP compliance was recorded after 6 months. After 155 patients were randomized and 137 participants were reevaluated at 6 months, the mean improvement in ESS score was near 5 and not different for the two groups. The mean change in functional outcomes of the sleep questionnaire score was near 2 and similarly not different for both groups. The authors also noted that the CPAP compliance in the specialist group was not significantly better at 5.4 (± 1.8) hours vs 4.8 (± 2.1) hours in the primary care-based group (P = .1). The within-study costs for primary care management were significantly lower with a savings of $2,157 per patient (in Australian dollars). The investigators concluded that “a simplified, ambulatory approach which utilizes the skills of appropriately trained PCPs and community-based nurses are not clinically inferior to usual management in a specialist sleep center.”15 This study might seem to support an equally efficacious and cheaper PCP model, but this seems plausible only when done within a highly structured program; the savings primarily came from not doing HST in both groups, not because of a sleep specialty-based diagnosis.