Novel mathematical processing method of nocturnal oximetry for screening patients with suspected sleep apnoea syndrome.

Aim: Polysomnography (PSG) is the current standard for diagnosing sleep apnea syndrome (SAS). A reliable test would be useful to reduce the considerable resources required for PSG.

Methods: We developed a new mathematical analysis, which quantifies amplitude variations of pulse oximetry (SpO(2)) and heart rate (HR) throughout the night, allowing measurement of the total time in which ΔSpO(2) >4% and presented as a new oximetric index ventilatory hypoxemic index (VHI). VHI was compared prospectively with standard PSG parameters apnea-hypopnea index (AHI) and oxygen desaturation index (ODI) in 106 patients (aged, 57 ± 14 years; BMI, 29 ± 5 kg/m(2)) suspected of having SAS. The criterion for diagnosis of SAS was AHI >15/h of sleep during PSG. The relationship between ΔSpO(2) and ΔΗR was also investigated.

Results: We observed a significant correlation between the AHI and VHI (R = 0.87, p < 0.0001). Using VHI >15 as the criterion for SpO(2), oximetry had a sensitivity of 81%, specificity of 98%, positive predictive value (PPV) of 98%, and negative predictive value (NPV) of 84% as a screening test for SAS. We also observed a clear linear relationship between ΔSpO(2) and ΔΗR, and concomitant use of VHI and the ΔSpO(2)/ΔHR slope improved the NPV to 90%.

Conclusions: The present findings from wavelet-aggregate processing of oximeter data and the relationship between ΔSpO(2) and ΔΗR show promise as a useful summary prediction of screening SAS.