The who, what, where and why of waking to fire alarms: a review

This review brings together several different strands of research: (i) the sleep arousal literature pertaining to auditory arousal thresholds (AATs), (ii) studies on factors affecting responsiveness to auditory signals during sleep, (iii) literature on responsiveness to smoke detector alarms during sleep and (iv) research on fire fatality statistics and victim characteristics. The review discusses the influence of age, sleep deprivation, signal frequency, background noise, hearing loss, time of night, stage of sleep, sex differences, dream incorporation, depression, signal meaningfulness, sleeping tablets, alcohol and marijuana on responsiveness during sleep. Studies using smoke alarms clearly suggest that an unimpaired sleeping adult will awaken quickly to a 55–60 dBA alarm (such as with a hallway installation), while the AAT literature suggests higher thresholds (most likely due to differences in signal frequency). However, it is argued that the level required to wake such adults under the ideal circumstances of an experimental situation should not be the minimum standard for pillow audibility. Such a level is unlikely to arouse children, those on sleep inducing medication, people with high frequency hearing loss (as may occur with age), those who are sleep deprived or those under the influence of alcohol or marijuana. The responsiveness of the unimpaired adults tested is not generalisable to the responsiveness of the people most likely to be the victims of fire occurring while they are asleep (the very young, elderly, intoxicated, or sick). The sound intensity of the alarm at the pillow should have the highest chance of arousing those most at risk of dying. It is, therefore, recommended that smoke detector alarms be installed in the bedrooms themselves such that the signal intensity is at the maximum level tolerable to the human ear, that is approximately 90 dBA