Pulseoximetry: sufficient to diagnose severe sleep apnea.

OBJECTIVES: To assess the clinical value of pulseoximetry in the diagnosis of sleep apnea when satisfactory agreement with polysomnography is obtained.

METHODS: This was a prospective clinical study, set in the Department of Otorhinolaryngology, Ullevaal University Hospital, Oslo, Norway. One hundred consecutive patients were investigated for sleep related breathing disorders. The main outcome measurements were: measurement success rate, oxygen desaturation thresholds, sensitivity and specificity at apnea-hypopnea-index (AHI) thresholds of 5 and 15.

RESULTS: Pulseoximetry was successfully performed in 93%. When different oxygen desaturation thresholds were calculated, optimal agreement with polysomnography was found at a 3% oxygen desaturation level. The sensitivity and specificity of diagnosing moderate/severe sleep apnea (AHI above 15) were 0.86 and 0.88, respectively. The corresponding figures for milder sleep apnea (AHI above 5) were 0.91 and 0.67. Good agreement was found between the AHI and the oxygen desaturation index (ODI) at the 3% level, with a mean AHI-ODI difference of 2.6 (SD, 7.3), a Pearson correlation of 0.95 and a weighted kappa of 0.86. The best agreement was found for AHI values below 15, where the estimated AHI-ODI difference was only -0.4 (SD, 3.3).

CONCLUSIONS: Pulseoximetry is a simple, non-invasive procedure, which is easy to perform and well suited for outpatient registration. When adjusted to polysomnography with high sensitivity of hypopnea registrations, an ODI at the 3% level is optimal to diagnose sleep apnea. In patients with moderate/severe sleep apnea with AHI values above 15, it is sufficient to establish the diagnosis and subsequent treatment. A negative pulseoximetry does not rule out sleep disorders; the patients should complete a full examination.