Prediction of sleep-disordered breathing by unattended overnight oximetry

Between January 1994 and July 1997, 793 patients suspected of having sleep- disordered breathing had unattended overnight oximetry in their homes follo wed by laboratory polysomnography. From the oximetry data we extracted cumu lative percentage time at SaO2 < 90% (CT90) and a saturation variability in dex (Delta Index, the sum of the differences between successive readings di vided by the number of readings - 1). CT90 was weakly correlated with polys omnographic apnea/hypopnea index (AHI), (Spearman rho = 0.36, P < 0.0001) a nd with Delta Index (rho = 0.71, P < 0.0001). Delta Index was more closely correlated with AHI (rho = 0.59, P < 0.0001). In a multivariate model, only Delta Index was significantly related to AHI, the relationship being AHI = 18.8 Delta Index +7.7. The 95% CI for the coefficient were 16.2, 21.4, and for the constant were 5.8, 9.7. The sensitivity of a Delta Index cut-off o f 0.4 for the detection of AHI greater than or equal to 15 was 88%, for det ection of AHI greater than or equal to 20 was 90% and for the detection of AHI greater than or equal to 25 was 91%. The specificity of Delta Index gre ater than or equal to 0.4 for AHI greater than or equal to 15 was 40%. In 1 13 further patients, oximetry was performed simultaneously with laboratory polysomnography. Under these circumstances Delta Index was more closely cor related with AHI (rho = 0.74, P < 0.0001), as was CT90 (rho = 0.58, P < 0.0 001). Sensitivity of Delta Index greater than or equal to 0.4 for detection of AHI greater than or equal to 15 was not improved at 88%, but specificit y was better at 70%. We concluded that oximetry using a saturation variabil ity index is sensitive but nonspecific for the detection of obstructive sle ep apnea, and that few false negative but a significant proportion of false positive results arise from night-to-night variability.