Fatigue: time to recognise and deal with an old problem

Being awake at times that are biologically programmed for sleep, prolonged wakefulness, and having had restricted sleep over a period of time result in fatigue and sleepiness. These conditions undoubtedly adversely affect human performance. This week's issue includes another reminder of the dangers of sleepiness while driving,¹ yet we continue to fail to treat fatigue with the seriousness it deserves.

The impact of sleep loss on performance is well documented by laboratory research.² Even modest amounts of sleep loss over short periods (about two hours a night over one week) accumulate and manifest themselves as an irresistible tendency to fall asleep during inappropriate or dangerous situations, like driving.³ The vulnerability of performance to circadian rhythm in alertness and sleepiness, even in well rested individuals, is similarly well documented.² Data on accidents from a variety of sources worldwide confirm the impact of time of day on the occurrence of accidents.⁴ Working at night, and working hours that restrict sleep opportunity, have long been implicated in compromised safety at work.⁵

The size of the problem is significant. A survey of car drivers in the United Kingdom found that 29% admitted to having felt close to falling asleep while driving in the previous year.⁶ Among New York drivers about a quarter reported having at some time fallen asleep at the wheel.⁷ About a third of truck drivers responding to a national survey in Australia reported that fatigue was a substantial problem.⁸ This week's study by Philip et al shows that 10% of almost 68 000 serious road crashes in good conditions affecting only one vehicle were related to fatigue (p 829).¹ The effects of fatigue are not limited to drivers. Over a third of a sample of junior hospital doctors reported that their hours of duty were always or often long enough to impair their work.⁹ A survey of over 3000 high school students in the United States showed that adolescents aged 13-19 do not get enough sleep, the extent of sleep loss increases with age, and sleep loss interferes with daytime functioning.¹⁰

Yet our attitude to fatigue is inconsistent with the concern that this knowledge should elicit. Lack of sleep is not seen as a risk and rest is not given high priority in the face of competing activities. Extended periods of wakefulness can even be seen as beneficial. A study of truck drivers using continuous electroencephalography convincingly showed that drivers obtained less sleep than they needed for alertness on the job, although they had enough time available in their schedules to do so.¹¹ Despite ample evidence showing that performance of hospital doctors is impaired by sleep deprivation,¹² long hours of work are seen by some as an integral part of the profession and training for it.¹³

Why, in the face of considerable evidence, is the potential for harm underestimated? Perhaps the answer lies in the fact that fatigue is a common experience. Clearly an adverse event does not accompany every occasion of fatigue. This ignores the fact that events causing injuries, on the road or at work, are multifactorial. The presence of fatigue, like the presence of alcohol, increases the risk of, rather than guarantees the occurrence of, an injury due to decreased performance capacity.

Secondly, the nature of the experience gives the illusion of control. During the development of fatigue, alertness waxes and wanes, so that the overall and inevitable decline in performance capacity is not necessarily recognised. Changes in stimulation (increasing ventilation, going for a walk, etc) appear to restore alertness, when in fact they are temporary interruptions of a continuing decline in alertness. People do not necessarily associate fatigue and sleepiness with falling asleep.¹⁴

Thirdly, there is no simple objective test of fatigue, equivalent to a breath analyser for alcohol, that can be applied after an injury has occurred. The contribution of fatigue needs to be inferred (as Philip et al have done¹). The inference is based on well established causal factors implicating fatigue in performance impairment such as time awake, prior wake-rest schedule, time of day, and characteristics of the crash or other injury-causing event.

There is also the practical issue of determining the level of fatigue at which performance poses a real risk. How do we set standards for fatigue? How much fatigue is too much? We recently compared the effects of sleep deprivation and alcohol intoxication and found that after 17-19 hours without sleep, starting from waking at about 0600 hours, individuals' performance was equivalent to or worse than at 0.05% blood alcohol concentration.¹⁵ In other words, commonly experienced levels of sleep deprivation—one extended day for a well rested individual—had a profound effect on performance. At around 2230-2430, well before reaching the circadian trough in alertness, performance levels were low enough to be considered incompatible with safe driving in many countries.

Fatigue is not new. Nor is knowledge about its potential for harm. Convincing evidence about the size of the risk and actual consequences has been slower to accumulate. While the evidence base needs to be strengthened, we already know enough to issue some cautions. Driving and working after extended wakefulness, after a night without sleep, after sleep has been restricted, or at vulnerable times of the day and night all contribute to fatigue. The effects of such conditions are exacerbated by alcohol.¹⁴ Public awareness of the potential hazards of fatigue and its causes needs to be raised in general, and among drivers in particular. Employers need to understand, and take responsibility for, the impact of work-rest schedules on performance at work and on performance when driving to and from work. Lack of sleep needs to stop being regarded as a badge of honour and seen for the serious hazard that it actually is.

Papers p 829