Nocturnal oximetry: is it a screening tool for sleep disorders?

The objective of this study was to examine the usefulness of home oximetry for the screening of sleep disorders presenting with excessive daytime sleepiness (EDS). This was accomplished by blinded comparison of diagnosis by oximetry alone versus polysomnographic diagnosis carried out at a sleep disorders center at a tertiary referral hospital. This study included three hundred patients who had been referred because of EDS and suspected sleep apnea. A number of measurements were made. The arterial oxygen saturation (SaO2) data were sampled at 2 Hz and stored digitally during polysomnography (PSG). From the SaO2 data recorded onto paper six scorers calculated the number of desaturations > 3% per hour (desaturation index: DI) and then made a diagnosis [normal, DI < 5; mild obstructive sleep apnea (OSA), 5 < DI < 20; moderate OSA, 20 < DI < 40; severe OSA, DI > 40]. Upper airway resistance syndrome (UARS) was diagnosed when DI was < 5 but associated with small fluctuations in SaO2. The diagnosis made by each of six scorers was compared to the clinical diagnosis made independently using PSG. Thirty-one (10.3%) of all the records were rejected by scorers because of inadequate SaO2 signals requiring technologist intervention. Sensitivity of screening for sleep-breathing disorders was 90.0% and specificity was 75.0%. All moderate and severe OSA patients were detected by oximetry. However, among the 66 patients who were classified as normal by oximetry, 1 had mild OSA, 20 had UARS, 9 had periodic limb movements in sleep, 4 had narcolepsy and 2 had a parasomnia. In conclusion, home oximetry may not have sufficient sensitivity and specificity to detect breathing disorders reliably during sleep and is useless for other disorders of sleep.