As this editorial is being published, trainee anaesthetists in the United Kingdom are no longer exempt from the European Working Time Directive (EWTD). As it is predominantly trainees who provide resident anaesthetic cover in UK hospitals, this implies some substantial changes. Many have already taken place but the full effect on senior staff has yet to be felt. The Directive requires workers to have 11 h of rest in every 24 h, with a minimum 24 h rest in every 7 days or a minimum 48 h rest in every 14 days. A weekly maximum of 58 h per week is also stipulated [1].

It is quite clear that the UK Department of Health expects not only full compliance with the Directive, but also that the current level of hospital activity be maintained.

Thus the Association of Anaesthetists' guidance on the effects of fatigue on anaesthetic work and patient safety, published in July this year, is timely [2]. It draws on peer-reviewed scientific literature and other published work. There is little high-quality investigation of direct relevance to anaesthesia and, as usual, the available evidence is supplemented and interpreted by an authoritative working party.

Some definitions and basic science will help get to grips with this issue. First, it is important to make the distinction between ‘sleep deprivation’ and ‘fatigue’. These are two separate, though related, entities. Fatigue can be thought of as the inability or unwillingness to continue effective performance of a task [3]. Sleep deprivation may be one possible cause of fatigue; workload, boredom, motivation, hunger and stress are others [4]. There is a brief summary of sleep physiology in the guidance. Probably the greatest benefit of this is to lay some myths to rest. For instance, whilst sleep requirements are commonly believed to lessen with age, this is not in fact true. Over the age of 45 years, sleep quality tends to deteriorate and repaying a ‘sleep debt’ by extending sleep time is more difficult [5]. Further, anaesthetists who reassure themselves that, although they might be tired, a sudden surge of catecholamines might keep them alert during an emergency, may be fooling themselves. Fatigue can cause spontaneous ‘microsleeps’ lasting seconds or even minutes and affected individuals may not even be aware of these. During these, the individual may be unresponsive to external stimuli. In severe sleep deprivation, the individual may ‘shut down’ into sleep regardless of the circumstances.

There is extensive literature describing the effects of sleep deprivation in other settings, for example the military and transport industries. However, the haphazard transfer of findings from one setting to another may deny important differences between tasks and individuals involved [6]. The process of anaesthesia is often compared to flying an aeroplane and, although this has brought useful insights [7], the analogy has some limitations. Aeroplanes react in a consistent manner; the same is not true of patients, who may show a vast array of responses, even in identical circumstances. (One could also say that, whilst mid-air hijack is fortunately still uncommon, smooth anaesthesia is invariably disrupted by surgery). As the Association guidance suggests, a less glamorous but possibly more fitting parallel is with the railway industry. The document notes, ‘A train driver is required to maintain a high degree of vigilance often over a prolonged period of time, responding to stimuli throughout the entire journey and interpreting signals in a constant requirement to recognise malfunction, conflicts or the need for clarification…unlike aviation, an automatic pilot is not available’[2].

Train drivers, like anaesthetists, may also be more likely to work without a co-driver and controls on their working hours may not have been so strictly enforced [8]. One can also find cultural similarities between the National Health Service and the national railway network. Although the UK rail network is no longer directly state-run, its long history as a monopoly provider with a strong cultural identity of its own echoes that of the health service. Part of these identities are attitudes to fatigue and safety. In medicine the traditional ‘physician culture’ assumes that doctors are immune to the effects of tiredness [9]. However, the working party makes it clear that fatigue should be considered an inevitable physiological process and must be acknowledged and managed. It is important that doctors recognise this and stop deceiving themselves that they can function satisfactorily even when tired. Comments from senior staff such as ‘In my day we just carried on regardless’ cannot be used as justification for them (and more commonly their younger colleagues) working in conditions which are potentially unsafe to the patient and damaging to themselves.

The available knowledge about the effects of sleep deprivation in clinical staff is derived from three main approaches. Psychomotor tests use standardised procedures for measuring the motor consequences of mental events or vice versa, for example card sorting, critical flicker fusion frequency and choice reaction time [10, 11]. The choice of test may be influenced more by ease of measurement and perceived sensitivity to fatigue than relevance to work-related tasks and patient safety.

Surveys and incident reports have also been used. For instance, Gander and colleagues reported an analysis of responses from 301 anaesthetists in New Zealand to a survey of their working hours and fatigue-related errors [12]. In the preceding 6 months, 50% of trainees and 27% of specialists had worked average weekly hours which exceeded their own perceived safe limits for maintaining patient safety; for 63% of trainees and 40% of specialists, these hours exceeded their limits for maintaining their personal well-being. Fatigue-related errors were reported by 86% of respondents. Although the relationship between fatigue and error is open to bias, the perception that their working hours (means of 59.0 and 51.2 h for trainees and specialists, respectively) gave these conscientious professionals cause for concern is beyond doubt. It is precisely this feeling – that we are at, or have exceeded, the point at which health and safety become threatened – that measures such as the EWTD should eradicate. Analysis of the Australian Incident Monitoring Survey data suggested that fatigue was a causative factor in 2.7% of incidents [13].

Physiological and task-related investigations. Early studies of this type were prone to confounding from a number of factors, including the presence of the researcher, unrecognised baseline chronic sleep deprivation, circadian rhythms and training effects. More recent work has tended to be of better quality, has avoided some of the methodological pitfalls mentioned above and has been more specific to anaesthesia.

For instance, Howard et al. [14] studied daytime sleepiness in 11 volunteer trainee anaesthesiologists in three conditions: during normal work schedules, after 24 h on call, and following a 4-day period of extended sleep. Sleepiness was measured as the time taken to fall asleep when left in a quiet darkened room. This was repeated throughout the test day (‘multiple sleep latency test’). Whilst their finding that those who had had extra sleep were less sleepy is predictable, there was no significant difference in sleepiness between the baseline and postcall conditions. This suggests that the ‘normal’ work schedule for these individuals was associated with latent, chronic sleep deprivation. Furthermore, there was substantial intersubject variability and subjects' assessments of whether they had fallen asleep (defined by EEG criteria) were inaccurate. These findings have important implications for defining ‘control’ groups in future studies looking at sleep deprivation but also underline the dangers of relying on individuals' own assessment of how fatigued they are at work.

The same team subsequently performed a study to compare the performance of 12 trainee anaesthesiologists during a 4-h simulated laparoscopic procedure [15]. This time, all subjects were tested under two conditions: after 25 continuous hours working, and after 4 days in which they were allowed to maximise their sleep. Equipment checks and clinical management of peri-operative conditions were assessed. Further, three types of ‘vigilance probes’ (stimuli to assess subjects' watchfulness) were presented to the subjects at intervals during the simulated case. These consisted of illumination of a red light, a sudden change of normal arterial waveform to flat line and the gradual change of heart rate or blood pressure to outside normal range. In addition, abnormal clinical events (for instance, bronchospasm, atrial fibrillation, atelectasis and myocardial infarction) were also simulated. Again, there was significant variation between subjects, but in total, one-third of residents fell asleep when sleep-deprived, compared with none in the control condition. However, although those in the sleep-deprived groups had a longer response time, clinical management of the simulated patient did not differ significantly between the two groups. This, taken together with the results of the AIMS analysis above, suggests that although the risk of fatigue is often present, it seldom appears to lead to errors with severe patient consequences. This implies that unacknowledged protective factors are at work in the system. However, as Howard and colleagues point out, their subjects were all young (average age 32 years) and the effects of fatigue may be quite different for older anaesthetists. Furthermore, although definitive evidence relating fatigue in anaesthetists to patient injury may be lacking, the ‘precautionary principle’[16] dictates that we act to prevent possible hazards now rather than waiting until they have occurred.

One area where evidence is especially sparse is the effect of shift systems on fatigue, a topic covered by a recent editorial in this journal [17]. The strictures of the EWTD clearly imply a shift system for 24-h services such as covering the intensive care unit and labour ward. Many trainees in UK anaesthetic departments now work a full shift system, usually with a 12-h overnight duty period and one or more daytime shifts. Whilst night shift working may prevent the weariness brought on by prolonged high-intensity work, and allow for periods of uninterrupted rest in between duty periods, it also reverses the normal circadian sleep–wakefulness cycle. This in itself may cause fatigue. Furthermore, it could be argued that shift working patterns bring fragmentation and reduce continuity of care. Of particular note is Cooper's finding that one of the major stressors for anaesthetists is the perception of lack of control over their working environment [18]. As the EWTD is not universally welcomed, the demotivating effect of having working patterns imposed upon us may tend to worsen fatigue for some. Thus the net effect of shift working on patient safety must depend on the balance of a number of complex factors. As we are bound to comply with the EWTD, arguments over whether shifts provide safer care overall than what went before is largely irrelevant. However, an evidence base to draw on would be valuable if policy changes were to be considered in the future. Whilst the studies quoted above are illuminating, the study of vigilance tasks has dominated work to date. However, as Weinger suggests, ‘anaesthesia practice also requires other higher-order cognitive processes such as pattern recognition, divided attention, task prioritisation, “situation awareness” and dynamic decision making’[19]. We would welcome research into the effects of fatigue on other aspects of anaesthetists' work such as cognitive ability, communication with patients, decision-making and team working. These would be relevant not only to the operating theatre but also to the intensive care unit and in resuscitation.

Practical suggestions for managing fatigue conclude the Association document. An outline of these is shown in the box below. Further advice is contained in Howard's review [4]. It is worth remembering that although fatigue may be most obvious after long periods of work, especially overnight, chronic sleep deprivation is also common and ‘sleep promotion strategies’ might be important not only to shift workers. Other factors contributing to fatigue – workload, boredom [20], stress and hunger should also be attended to. Departments of Anaesthesia should draw up their own strategies for balancing the management of fatigue and maintaining workload and patient safety. Clearly, achieving this balance is vital for the wellbeing of anaesthetists as well as their patients. The Association's motto ‘In somno securitas’– safety in sleep – has never seemed more apposite.


  1. Top of page
  2. References
  3. Managing fatigue
  • 1 Accessed 12 February 2003.
  • 2
    Association of Anaesthetists of Great Britain and Ireland. Fatigue and Anaesthetists. London: AAGBI, 2004. Available at:
  • 3
    Weinger MB, Englund CE. Ergonomic and human factors affecting anesthetic vigilance and monitoring performance in the operating room environment. Anesthesiology 1990; 73: 9951021.
  • 4
    Howard SK, Rosekind MR, Katz JD, et al. Fatigue in anaesthesia: implications and strategies for patient and provider safety. Anesthesiology 2002; 97: 128194.
  • 5
    Kryger MH, Roth T, Dement WC, eds. Principles and Practice of Sleep Medicine, 3rd edn. Philadelphia: W.B. Saunders, 2000.
  • 6
    Sexton JB, Thomas EJ, Helmreich RL. Error, stress, and teamwork in medicine and aviation: cross sectional studies. British Medical Journal 2000; 320: 7459.
  • 7
    Rutherford W. Aviation safety: a model for health care? Quality and Safety in Health Care 2003; 12: 1623.
  • 8
    Buxton A, Bunting A, Dickson BT, et al. Visual and Mental Acuity Study. Defence Evaluation and Research Agency (DERA, Centre for Human Sciences (CHS), 2000.
  • 9
    Parker JBR. The effects of fatigue on physician performance – an underestimated cause of physician impairment and increased patient risk. Canadian Journal of Anaesthesia 1987; 34: 48995.
  • 10
    Beatty J, Ahern BK, Katz R. Sleep deprivation and the vigilance of anesthesiologists during simulated surgery. In: Vigilance: Theory, Operational Performance, and Physiological Correlates. Plenum Press, 1977.
  • 11
    Narang V, Laycock JRD. Psychomotor testing of on-call anaesthetists. Anaesthesia 1986; 41: 8689.
  • 12
    Gander PH, Merry A, Millar MM, Weller J. Hours of work and fatigue-related error. a survey of New Zealand anaesthetists. Anaesthesia and Intensive Care 2000; 28: 17883.
  • 13
    Morris GP, Morris RW. Anaesthesia and fatigue: an analysis of the first 10 years of the Australian Incident Monitoring Study 1987–97. Anaesthesia and Intensive Care 2000; 28: 3004.
  • 14
    Howard SK, Gaba DM, Rosekind MR, Zarcone VP. The risks and implications of excessive daytime sleepiness in resident physicians. Academic Medicine 2002; 77: 101925.
  • 15
    Howard SK, Gaba DM, Smith BE, et al. Simulation study of rested versus sleep deprived anesthesiologists. Anesthesiology 2002; 98: 134555.
  • 16
    McKechnie S, Davies S. Consumers and risk. In: BennetP, CalmanK., eds. Risk Communication and Public Health. Oxford: Oxford University Press, 1999.
  • 17
    D'Auria D. Shift work, the anaesthetist and Santayana's warning. Anaesthesia 2004; 59: 7357.
  • 18
    Cooper CL, Clarke S, Rowbottom AM. Occupational stress, job satisfaction and well-being in anaesthetists. Stress Medicine 1999; 15: 11526.
  • 19
    Weinger MB. Vigilance, boredom and sleepiness. Journal of Clinical Monitoring and Computing 1999; 15: 54952.
  • 20
    Galea M, Tooze A, Smith AF. Anaesthetic risk: analysis and reassurance. Clinical Risk 2004; 10: 926.

Managing fatigue

  1. Top of page
  2. References
  3. Managing fatigue
  • • Education on sleep medicine

  • • 
    Cultural change towards recognizing that fatigue is universal and physiological
  • • 
    Expansion/flexibility of staffing to allow relief for breaks
  • • 
    Formal handovers when leaving theatre for breaks and on returning
  • • 
    Careful design of shift patterns: forward rotating patterns preferred (day–evening–night). Limit night shifts to a maximum of five
  • • 
    Encourage planned ‘napping’ and provide facilities away from clinical areas to take naps.

The journal would like to apologise for omitting the author details from the August 2004 Editorial (Anaesthesia 2004; 59: 735–7). The Editorial was written by Dr D. A. D'Auria, Consultant Occupational Physician, University of Wales College of Medicine, Cardiff, UK. E-mail: