Staying Awake For Safety Sake: A Dream Not Yet Realized

Abstract

number of microsleeps during the driving task, steering deviation, braking reaction time and crashes all negatively correlated with the first MWT sleep latency. Using a receiver-operator characteristic curve, the authors found that first MWT sleep latency in the partial sleep deprivation plus alcohol condition significantly discriminated subjects who had a crash from those who did not. They conclude that sleep latency on the MWT is a reasonable predictor of driving simulator performance at least in sleepy, alcohol-impaired, normal subjects. While it is reassuring that impairment due to combined effects of sleepiness and alcohol can be detected by their driving simulator and that MWT tracks this impairment reasonably well, the applicability of these results is far from clear. Only 1 (albeit fairly important) driving simulator measure (braking reaction time) was significantly correlated with MWT whereas many more were brought out by the additive (and/or synergistic) effect of alcohol. It is already established that mild sleepiness is exaggerated by low dose alcohol 15 and while the results may then apply to cases where both condition apply, many motor vehicle collisions occur without any alcohol involved. Moreover, peak accident frequencies usually occur much later than 0100 hours, the time tested in this paradigm. 16