The standard diagnostic test of OSA, i.e. overnight polysomnography (PSG), is complex and difficult to perform outside the sleep laboratory. However, the value of electrophysiological and respiratory monitoring has not been assessed critically in unattended settings. In this study we evaluated whether full PSG is necessary to establish a diagnosis of OSA in unattended studies.
Full unattended PSG studies were conducted in the home place (25 tetraplegic patients)or in the hospital (25 surgical patients studied preoperatively)using a standard montage. Installation of the biosensors was performed by a trained sleep technician. Recording was initiated at a preset time and all signals stored on a digital storage media (Flash card) and subsequently downloaded to a desktop computer for analysis. No overnight monitoring was possible. The PSGs were first scored using standard criterions by one of 3 trained sleep technicians and an apnea/hypopnea index (AHI) calculated as the number of respiratory events per hour of sleep as determined by sleep staging. The analysis was then repeated on a separate day and in a blind fashion using respiratory variables only (i.e. without EEG and EMG signals) and a respiratory disturbance index (RDI) calculated as the number of respiratory events per hour of recording time.
Sleep efficiency was 77.1±13.8%. Recording time exceeded sleep time by 117±79 minutes (p<.03) but RDI (20.5±21.3) was not ≥significantly different from AHI (22.9±24.2; p=.116). Both were highly correlated (r2=.82). Using a AHI diagnostic cutoff value ≥15 events/h, the diagnostic sensitivity and specificity of RDI were both 86%. If a RDI cutoff value of ≥10 events/h was adopted instead, sensitivity would be 100%.
Because of frequent arousal, the Rechtschaffen and Kales method for scoring sleep markedly underestimate sleep time as well as the number of respiratory events. We conclude that bioelectric signals are not necessary for the diagnosis of OSA.
Because PSGs are much simpler to perform without bioelectric signals, this simplified approach should improve access to and cost/benefit of diagnosis and treatment of OSA.
Pierre Mayer, None.