Behavioural observation as a means of assessing sleepiness related driving impairment in obstructive sleep apnoea

This study compared the temporal pattern of non-verbal behaviours (actions not directly related to task performance) in obstructive sleep apnoea (OSA) drivers under monotonous driving conditions following sleep restriction versus normal sleep. Seventeen patients with untreated severe OSA completed a 90-minute driving simulator task during mid-afternoon under two experimental conditions: prior normal habitual sleep (~8h) and prior sleep restriction (4h time in bed). Steering deviation and crash events were identified using a driving simulator. Non-verbal behaviours (self-centred gestures, non-verbal facial activities, postural adjustments, non-self-centred gestures and eye closures) were identified using video recording and a behavioural ethogram. Participants demonstrated increased steering deviation over the drive (p<0.001) and following sleep restriction (p<0.001). All non-verbal behaviours, except non-verbal facial activities, increased over the drive (all p<0.01). Compared to normal sleep condition, the incidence rate was 2.1 times higher for eye closures (95%CI 1.75-2.60) and 1.5 times higher for postural adjustments (95%CI 1.29-1.72) following sleep restriction, while non-self-centred gestures reduced by 50% (IRR 0.53, 95%CI 0.36-0.78), all p<0.01. In the 10 minute period prior to simulator crash events, eye closure frequency increased compared to equivalent periods without a crash event 2.1 (95%CI 1.4-3.8, p<0.01). Kaplain-Meier analyses showed a progressive cessation of non-verbal facial activities leading up to crash events (X²=6.2, p=0.013). Although eye-closure appears to be a more sensitive marker of poor vigilance, behavioural observation could provide a novel further method for assessing vigilance failure in OSA and could assist in the development of novel video-based in-car devices for detection of driver sleepiness/ fatigue.