Objective: To explore the value of pulse oximetry in evaluating the severity of obstructive sleep apnea hypopnea syndrome (OSAS).
Methods: A total of 129 outpatients with a complaint of snoring visiting doctors at Beijing Tongren Hospital were recruited between October 2012 and August 2013. All outpatients received an pulse oxygen saturation (SpO2) test and a polysomnography (PSG) test at the same night. The results of all PSG tests were analyzed by a technician while the results of all SpO2 tests were automatically analyzed by software. They were divided into two groups according to the apnea hypopnea index (AHI) of PSG. One group consisted of patients with simple snoring or mild OSAS (AHI<5 events/h) while another group had moderate-to-severe OSAS (AHI ≥ 15 events/h). The differences between the values of SpO2 variables and the relationship between SpO2 variables and AHI were all analyzed. A Logistic regression analysis of age, body mass index (BMI) and all SpO2 variables was performed. A receiver operating characteristics (ROC) test was also performed to evaluate the SpO2 variables in predicting moderate-to-severe OSAS.
Results: There were 115 males and 14 females. Their age and body mass index (BMI) were 41.8 ± 10.4 years and (27.7 ± 3.6) kg/m(2) respectively. And 27 subjects (20.9%) suffered from simple snoring or mild OSAS (AHI<15 events/h) while another 102 patients (79.1%) had moderate-to-severe OSAS (AHI ≥ 15 events/h). The oxygen desaturation index of ≥ 2%, ≥ 3%, ≥ 4% in patients with simple snoring or mild OSAS were all significantly lower than those in those with moderate-to-severe OSAS ((8.4 ± 5.6), (4.6 ± 4.1), (3.2 ± 3.3) and (42.1 ± 22.1), (36.0 ± 23.0), (31.8 ± 23.0) events/h). And the average and the lowest SpO2 were both significantly higher in this group than those with moderate to severe OSAS (all P < 0.001). All SpO2 variables were significantly correlated with AHI (all P < 0.001). The correlation coefficient between ODI2 and AHI could reach 0.949. ODI2 was the only significant predictor parameter for moderate-to-severe OSAS (OR = 1.450, 95%CI: 1.205-1.745, P < 0.001). The area under ROC curve for diagnostic performance could reach 0.966. If the ODI2 values of 6.7 and 23.8 events/h were used as cutoff points, 74 patients (57.4%) with an ODI2 of ≥ 23.8 events/h suffered from moderate-to-severe OSAS, 14 patients (10.9%) with an ODI2 of <6.7 events/h had simple snoring or mild OSAS while the disease severity of 41 remainders (6.7 events/h ≤ ODI2<23.8 events/h) were hard to predict.
Conclusions: By setting proper cutoff values, the results of oximetry may be used to evaluating preliminarily the severity of OSAS. Therefore it may be used as a screening test for guiding the treatments of OSAS.