Objective: Pulse oximeter is a useful screening device for SAS. We studied whether measurement of SaO2 could identify patients with SAS and evaluate the severity of SAS.
Methods: 174 snorers were assessed clinically and then underwent formal PSG and oximetry test at the same time. From the oximetry data, the percentage of time spent at SaO2 below 90% (SIT90%), waking SaO2(H SaO2%), the lowest SaO2(L SaO2%) during sleep, the mean SaO2(M SaO2%) and the number of oxygen desaturation > or = 4% per hour (DI4) were calculated. We also divided 100 cases among them into four groups by AHI, which were G0(AHI < 5), G1(AHI 5-19), G2(AHI 20-39), G3(AHI > or = 40), and evaluated whether the SIT90, ISaO2, MSaO2 and DI4 are different among the four groups.
Results: There was an statistically significant correlation (r = 0.91 P < 0.001) between DI4 and AHI. DI4 > or = 5 per hour identified patients with AHI > 5 with a sensitivity of 94%; DI4 > or = 15 per hour identified patients with AHI > 5 with a specificity of 98%, and for patients with AHI > or = 20, the sensitivity was 100%. The SIT90 and DI4 were significantly different among the four groups (P < 0.01).
Conclusions: Oximetry with DI4 < 5 practically excludes clinically significant SAS, and DI4 > or = 15 identified almost all of them. SIT90 and DI4 could be used as parameters in evaluating the severity of SAS.