Evaluation of the accuracy of SNAP technology sleep Sonography in detecting obstructive sleep apnea in adults compared to standard polysomnography

Objective: To determine the accuracy of snoring and apnea analysis by SNAP (SNAP Laboratories; Glenview, IL), a technology that uses snoring recorded by a home microphone system and nasal airflow, to diagnose obstructive sleep apnea (OSA) as well as determine its severity. Methods: For all patients who had undergone a prior SNAP study and were referred to the Sleep Disorders Center of Lifespan Hospitals for polysomnography testing from January 2000 through December 2001, we compared the results of the SNAP study to standard polysomnography (polysomnography). The severity of each apnea-hypopnea index (AHI) [mild, moderate, or severe, as defined by the AHI Severity Task Force of the American Academy of Sleep Medicine] recorded by SNAP was compared to that of the polysomnography result. All polysomnography tests were scored independently and without the prior knowledge of any SNAP results. Results: For the 31 patients on whom data were available, the mean age, body mass index, and Epworth sleepiness scale scores were 50.3 years (range, 29 to 77 years), 31.6 kg/m(2) (range, 24 to 44 kg/m(2)), and 11.3 (range, 1 to 20), respectively. The mean follow-up time between the two studies was 5 months. The severity criteria indicated by the SNAP study accurately assessed the true severity confirmed by polysomnography in only 11 of 31 patients (35.5%). When the AHI severity score from the SNAP study was compared to polysomnography using a kappa statistic measure of agreement, there was overall agreement with a kappa value of 0.23 (p = 0.008), but the agreement was only fair. SNAP study severity scores were overestimated in 13 of 31 patients (41.9%) compared to the polysomnography results. In the majority of these subjects (8 of the 13 "overestimated" patients or 8 of 31 total patients [25.8%]), the SNAP study diagnosed OSA when the patient had a normal polysomnography finding. Conclusion: Although there may be some night-to-night variability in polysomnography testing, these results suggest that SNAP studies do not appear to accurately assess the severity of OSA.