There is no meaningful difference in the respiratory event index calculated using the actigraph estimated sleep time versus total recording time

Abstract

Out-of-center sleep tests (OCSTs) are increasingly being used to diagnose obstructive sleep apnea (OSA). Several OCST devices include an embedded actigraph that helps to limit the analysis times to those most likely to represent sleep. The respiratory event index (REI) can thus be determined either by dividing the sum of apneas and hyponeas by the total recording time (REI-TRT) or by dividing by the actigraphic estimate of sleep time (REI-ACTI). This study examines how closely the REI-TRT compares to the REI-ACTI for the diagnosis of OSA. A retrospective analysis was conducted for 129 OCSTs performed at the Mayo Clinic from October 2009 to May 2012 using the portable Embletta™ system. The REI-TRT was compared to the REI-ACTI by a paired-t test and concordance correlation coefficient (CCC). The clinical endpoint of each study, namely the presence and severity of OSA, was determined using the actigraphic estimate of sleep and the total recording time to verify if there was a difference in the clinical outcome between the REI-TRT and REI-ACTI, respectively. There was a significant difference between the actigraphic sleep time and the total recording time ( p value < 0.0001, CCC 0.809) with a mean difference of 18.5 min (CI 15.2–21.8) between them. Despite a statistically significant difference between the REI-TRT and REI-ACTI ( p value < 0.001), they were highly concordant with a CCC of 0.997 and the mean difference between the REI-TRT and REI-ACTI was only 0.8 (CI 0.59–0.98). This difference did not change the clinical diagnosis of OSA nor the severity designation for any of the 129 studies. These results suggest that the difference between the REI-TRT and REI-ACTI is clinically negligible and that the REI-TRT is a reasonable surrogate for the REI-ACTI. Rather than estimating the total sleep time, the actigraph seemed most useful at determining the beginning and end of the total recording time.