Desaturation event characteristics and mortality risk in severe sleep apnea

Abstract

Obstructive sleep apnea (OSA) is a public health problem with severe health consequences. The current OSA severity estimation is based on the average number of breathing cessation and desaturation events per hour of sleep, neglecting the individual event characteristics. The aim of the current study was to evaluate desaturation event morphology in deceased and matched control patients with severe OSA. 12 deceased and 12 AHI, age, BMI and follow-up time matched alive control patients with severe OSA were analyzed. Desaturation event durations, depths, and areas of the deceased and alive control patients were compared. Also the effect of different baseline level selection in the desaturation depth analysis was investigated. Patient demographics, apnea-hypopnea-index (AHI) and oxygen-desaturation-index (ODI) did not differ statistically significantly between the groups. The average oxygen saturation levels were statistically significantly lower 89.8% vs. 93.2% (p=0.002) in the deceased patients compared to the alive controls. The median desaturation event duration 31.8s vs. 25.9s (p=0.017), depth 15.0% vs. 9.5% (p=0.006) and area 349.9s% vs. 201.4s% (p<0.001) were statistically significantly greater in the deceased patients compared to the alive control patients when using 100% saturation as baseline level for desaturation events. When the first point before desaturation onset was used as baseline no statistically significant differences (p=0.089) were found between the deceased and alive control patients in desaturation depths. Based on quantitative inspection of the distributions of individual desaturation event characteristics the desaturation events were more severe in the deceased group. Patients with similar AHI and ODI can have different individual desaturation event characteristics. Selection of the baseline for desaturation event depth analysis can affect the estimation of the event severity. The analysis of the individual desaturation event characteristics can provide supplementary information on the severity estimation of OSA and support the individual mortality risk estimation in severe OSA patients.