Guiding principles for determining work shift duration and addressing the effects of work shift duration on performance, safety, and health: guidance from the American Academy of Sleep Medicine and the Sleep Research Society

Abstract

Risks associated with fatigue that accumulates during work shifts have historically been managed through working time arrangements that specify fixed maximum durations of work shifts and minimum durations of time off. By themselves, such arrangements are not sufficient to curb risks to performance, safety, and health caused by misalignment between work schedules and the biological regulation of waking alertness and sleep. Science-based approaches for determining shift duration and mitigating associated risks, while addressing operational needs, require: (1) a recognition of the factors contributing to fatigue and fatigue-related risks; (2) an understanding of evidence-based countermeasures that may reduce fatigue and/or fatigue-related risks; and (3) an informed approach to selecting workplace-specific strategies for managing work hours. We propose a series of guiding principles to assist stakeholders with designing a shift duration decision-making process that effectively balances the need to meet operational demands with the need to manage fatigue-related risks.

Appendices

APPENDIX A

Biological Processes of Sleep/Wake Regulation

Mental fatigue, as defined earlier in this paper, is influenced strongly by sleepiness, and thereby by time of day and by the durations of wakefulness and prior sleep. This relationship is governed by two primary biological processes: the “circadian” wake drive and the “homeostatic” sleep drive.131–133 During the day, the circadian process produces rising pressure for wakefulness, which counteracts sleepiness. During the night, the circadian process withdraws this wake pressure, which promotes sleepiness and tendency to fall asleep. Simultaneously, during periods of wakefulness, the homeostatic process builds up pressure for sleep, which promotes sleepiness. During sleep periods, homeostatic pressure for sleep dissipates; the longer the sleep period, the less the remaining homeostatic sleep drive. Awakening from sleep occurs naturally when the remaining homeostatic sleep drive is overcome by the circadian wake drive. At any given time during wakefulness, sleepiness—and thereby mental fatigue as defined in this paper—is thus influenced by the interplay between the two processes, as illustrated in Figure A1.

Figure 2

Regulation of sleepiness by two biological processes.

A circadian process (red) produces a rising pressure for wakefulness, which counteracts fatigue, during the day; and a withdrawal of that pressure for wakefulness, thereby promoting sleepiness, during the night. Simultaneously, a homeostatic process (blue) builds up a pressure for sleep, thereby promoting sleepiness, during periods of wakefulness; and dissipated that pressure for sleep (dark gray), thereby providing recovery, during periods of sleep (top panels).131 The combined effect of the circadian and homeostatic processes on sleepiness may be calculated as the net difference between the homeostatic pressure for sleep and the circadian pressure for wakefulness (bottom panels, green),132 as illustrated here for a scenario with a daytime duty period (light blue) and an 8-h nighttime sleep opportunity (left panels) and for a scenario with a nighttime duty period (light blue) with an 8-h daytime sleep opportunity (right panels). Note that in the day work scenario, the sleep opportunity is ended (e.g. through use of an alarm clock) somewhat prematurely, as there is still some homeostatic pressure for sleep left to be dissipated (top left). In the night work scenario, however, sleep is curtailed much more, with the rising circadian wake pressure causing awakening from daytime sleep well before the end of the sleep opportunity. The early awakening leaves a higher level of homeostatic sleep pressure at the end of the shortened sleep period (dark gray) and causes a portion of time available for sleep in this scenario (light gray) to remain unutilized (top right). The combined effect of the two processes—sleepiness calculated as the net difference between the homeostatic pressure for sleep and the circadian pressure for wakefulness—is that sleepiness is low and stable throughout the duty period in the daytime duty scenario (bottom left), whereas sleepiness increases and peaks toward the end of the duty period in the nighttime duty scenario (bottom right). Note that the transient cognitive impairment immediately after awakening known as sleep inertia115 is not depicted in this figure.

Because the circadian process is a function of time of day, while the homeostatic process is a function of time awake and time asleep, the interplay between the two processes depends critically on the timing of periods of wakefulness and sleep. Figure A1 illustrates this for a healthy young adult in a day work scenario with daytime wakefulness and nighttime sleep (left panels), and in a night work scenario with nighttime wakefulness and daytime sleep (right panels). In the day work schedule, the interplay between the circadian and homeostatic processes maintains a stable, low level of sleepiness during most of the day, followed by a rapidly rising level of sleepiness in the late evening leading to the initiation of sleep. Furthermore, during nighttime sleep, the interplay between the two processes produces a consolidated sleep period. If the sleep period is not long enough, dissipation of the homeostatic pressure for sleep may be insufficient, and use of an alarm clock may be needed to wake up in time for work or other responsibilities.

By contrast, in a night work schedule, the interplay between the circadian and homeostatic processes produces a steady increase in sleepiness through most of the night. Furthermore, the rising circadian wake drive during daytime sleep causes early awakening and incomplete dissipation of the homeostatic pressure for sleep. High circadian wake drive during the late afternoon and early evening (the so-called “forbidden zone for sleep”134 or “wake maintenance zone”135) makes it difficult to obtain additional sleep during the afternoon. Thus, compared to a day work scenario, a night work scenario tends to produce sustained sleep loss and dynamically changing, higher levels of sleepiness.13

Given this biological regulation of sleepiness and its impact on mental fatigue, a prescriptive limit on work hours would not, by itself, prevent high fatigue levels during a night work schedule. (In fact, a prescriptive limit on work hours could inadvertently place the commute home at the time of greatest sleepiness, just before the rising circadian wake drive would partially reduce sleepiness again and mitigate fatigue.11)

APPENDIX B

Example Risk and Safety Management Policy Frameworks

Shift duration is a safety and health issue with legal implications, and policies and procedures pertaining to shift duration are best embedded within organizational risk and safety management systems.121 Many industries have pre-existing policy frameworks from which to draw, such as workplace health and safety policies, Safety Management System (SMS) policies, or Fatigue Risk Management (FRM) policies. Here we provide a few links to example frameworks and free resources (shared with permission), which may serve as a starting point for organizations to establish their own policies:

• The “Fatigue Risk Management Toolkit for Residents, Leaders, and Policy Makers in Canadian Postgraduate Medical Education” document contains practical tools and an implementation guide for managing fatigue in Canadian medical education. In large part it readily applies to U.S. medical education settings. It can be downloaded from: http://www.residentfatigue.ca/mssites/frm/PDF/Fatigue-Risk-Management-Toolkit-WEB5.pdf (last accessed on March 30, 2021).

• The “FRM Standard Operating Procedure for Utilities” document provides a straightforward but comprehensive standard operating procedure for a fatigue risk management system (FRMS) currently used in Australia. It may be adapted for U.S. operations and can be downloaded from: https://www.sleepresearchsociety.org/wp-content/uploads/2021/03/FRM-Standard-Operating-Procedure-for-Utilities.pdf (last accessed on March 30, 2021).

• The Fatigue Risk Management Systems Implementation Guide for Operators, 1st edition6 developed by the International Air Transport Association (IATA), the International Civil Aviation Organization (ICAO), and the International Federation of Airline Pilots’ Associations (IFALPA) is widely seen as a model for FRMS in larger operations. It can be downloaded from: https://www.icao.int/safety/fatiguemanagement/FRMS%20Tools/FRMS%20Implementation%20Guide%20for%20Operators%20July%202011.pdf (last accessed on March 30, 2021).

• The 2018 Fatigue Risk Management Guidelines for Emergency Medical Services Implementation Guidebook helps EMS administrators with the implementation of rigorously developed, evidence-based guidelines52 for fatigue risk management in EMS, with condensed summaries of recommendations and sample policy statements that may be tailored to the needs of local agencies. It can be downloaded from: https://nasemso.org/wp-content/uploads/Fatigue-Guidebook-FINAL-2018Oct.pdf (last accessed on March 30, 2021).

In addition, the “Fatigue at Work Employer Toolkit” developed by the National Safety Council (NSC) contains educational materials for human resources personnel, supervisors, and employees pertaining to fatigue-related safety risks. It can be downloaded from: https://safety.nsc.org/fatigue-risk-management-toolkit (last accessed on March 30, 2021).

APPENDIX C

Recommendations for the Development of Systems to Monitor the Outcomes of Changes in Shift Duration

After making changes in shift duration, organizations should monitor and respond to leading indicators of potential problems and both positive and negative outcomes to ensure the effectiveness of their shift duration policies. Monitoring systems should identify, report, quantify and manage existing and emerging risks. For the development of such systems, the following recommendations deserve consideration:

• (1) The degree (and likely cost) of monitoring should be proportional to the anticipated risk level. Where the risk level is low, monitoring may be relatively minimal. Where the risk level is high, however, monitoring should be more comprehensive, evidence-based and, in larger operations, bolstered by third-party oversight.

• (2) Monitoring systems should identify and report key performance indicators. These indicators should enable the organization to demonstrate evidence for the following.

  • (a) There is appropriate knowledge of relevant policies, including personal and organizational roles and responsibilities in identifying, reporting, quantifying and mitigating the risks associated with extended shifts.

  • (b) There is appropriate training and education of staff in how to identify, report, quantify and manage the risks associated with extended shifts.

  • (c) Methodologies used for measuring, reporting and mitigating risks are evidence-based, subject to regular evaluation and, where appropriate, modified in light of local experience and/or relevant or emerging scientific evidence. These could include, and are not limited to, measures of

    • ∘ planned and actual working time arrangements for all at-risk employees, including planned and unplanned overtime, on-call work, secondary employment, or other activities likely to interfere with sleep

    • ∘ sleep-wake behavior of individualemployees, as part of a shared-responsibility framework in which employer and employees have joint responsibility for ensuring adequate sleep

    • ∘ fatigue/sleepiness-related changes in behavior or task performance

    • ∘ self-report measures of fatigue, sleepiness, alertness, or fitness-for-duty

    • ∘ the efficacy and utilization rates of in situ countermeasures at the organizational level.

• (3) Based on data obtained in (2), organizations should respond appropriately to opportunities to apply corrective actions that will reduce risks associated with shift duration.

• (4) Monitoring systems should be cost-effective, so that the costs of implementation, or of the application of corrective actions, does not exceed the likely benefits. This guidance should, however, not be interpreted as advice to minimize resources dedicated to monitoring.

• (5) Organizations should implement a system that will enable them to identify and respond to any meaningful relationships between shift duration and risk outcomes including, but not limited to, productivity, performance, safety, health, and community and environmental consequences.

APPENDIX D

Background Information on Countermeasures for Risks Associated with Shift Duration

This Appendix provides background information and references that pertain to selected countermeasures for risks associated with shift duration as shown in Table 1. Implementation of these countermeasures can be complex, and consultation with an expert is generally recommended.

D1. Scheduling Improvements for Shift Work and Extended Shift Operations

Thoughtful and data-driven design of working time arrangements can contribute to risk mitigation by maximizing and protecting sleep opportunities, aligning work schedules with the circadian rhythm of the biological clock, and/or increasing time for recovery after extended duties or multiple shifts. While the primary focus of this paper is on shift duration, the timing of shift starts and ends, and other aspects of the working time arrangements are also important to consider in this regard.136,137 Scientific evidence pertaining to these issues is limited, and conclusive studies of scheduling improvements are largely lacking due to a wide range of possible confounds.138 Nonetheless, the following guidelines provide some insight into what kinds of scheduling practices affect performance, safety, and health risks and therefore present potential opportunities for improvement.

• Shift systems with backward rotation, in which the start times of consecutive shifts is advancing (i.e. the next block of shift starts at an earlier time of day than the current block), are generally less well tolerated than shift systems with forward rotation or fixed shift times.81 The speed of rotation (i.e. how many shifts in a block with the same start time) also plays a role in rotating shift systems, but the evidence on its effects is mixed.139

• Both night shifts and early morning starts curtail nighttime sleep opportunities, which increases risk levels relative to late morning and afternoon/evening shifts.77,140

• Shift durations beyond 12 h in duration tend to be associated with increased risk levels.80,140 Shift durations of 24 h or more without protected opportunities to nap while on duty are not recommended.51 Evidence is mixed, however, on 12-h shifts compared to 8-h shifts.51,141

• Risk levels tend to accumulate across consecutive shifts without days off.142,143

  Double (i.e. back-to-back) shifts and quick returns, overtime, and second jobs increase risk levels.22,144,145

• Irregular and unpredictable work hours and on-call duty schedules are often perceived as stressors and may also interfere with the ability to obtain adequate sleep, thereby increasing risk levels.9,146

• Whereas the impact of workload (or task load) is not well established,147,148 high workload may interfere with control over the pacing of work tasks and restrict time for rest breaks, which may increase risk levels.

D2. Napping

Taking a nap—a relatively short sleep period that may be (loosely) defined as less than half the duration of an individual’s typical nocturnal sleep length—is an effective way to supplement the daily amount of sleep and a powerful countermeasure to sleepiness and fatigue.149–153 Naps as short as 15 min and as long as several hours can be effective, whether before work (pre-emptive or prophylactic naps), during work (on-the-job or strategic naps), or after work (catch-up naps).112,154–156 Naps may facilitate adaptation to a shift work schedule and ease the return to daytime activity,157,158 and may offer cardiovascular health benefits.159,160 Split sleep schedules, in which a person takes a nap after a work shift and another nap before the next shift, have been associated with increased sleepiness,161 but few differences in performance relative to a consolidated post-shift sleep bout.162,163 For on-the-job napping to be implemented successfully as a countermeasure strategy, it is important that it be sanctioned164 and that there is access to a safe and quiet place to rest while on break.109,127

Importantly, napping may produce post-nap sleep inertia, a transient feeling of grogginess and impaired performance immediately upon awakening.115,165 Sleep inertia can be particularly problematic in on-call settings146 and may require a worker to delay the return to work for up to about 30 min after waking.166,167 Sleep inertia may be less intense after shorter (approximately 10–30 min) naps, but scientific evidence on this matter is inconclusive.168 Caffeine consumed just prior to a nap appears to be an effective countermeasure to performance impairment due to sleep inertia immediately after the nap.129 (see Appendix D3).

D3. Caffeine Use

There is abundant evidence that caffeine reduces sleepiness and fatigue and mitigates deficits in cognitive performance.116,169 When caffeine is used in conjunction with a nap, it may also reduce the time to overcome post-nap sleep inertia.170–173 Caffeine is widely available and found in many drinks and foods, either naturally occurring or as an additive; furthermore, caffeine can be obtained in the form of chewing gum and various over-the-counter medications. Available evidence suggests that moderate use of caffeine is compatible with a healthy lifestyle.174

The pharmacodynamics of caffeine are poorly understood, and large inter-individual differences in caffeine sensitivity, effectiveness, habituation, and tolerance exist.175–177 This limits the ability to provide tailored advice regarding dosing of caffeine to mitigate fatigue - although given the widespread presence of caffeine in foods and drinks, precise dosing could be difficult in practice regardless. Even so, the preponderance of scientific evidence indicates that caffeine is an effective fatigue countermeasure, and access to caffeine as part of a comprehensive fatigue risk management program is recommended.111

Caffeine present in the bloodstream just before bedtime may delay sleep onset and reduce the quality and quantity of subsequent sleep, although individuals differ considerably in their sensitivity to these effects.178 Sustained use of high-dose caffeine can cause additional undesirable effects, including anxiety, tremor, arrhythmias, insomnia, dehydration, and withdrawal headaches.179,180 For healthy adults, caffeine consumption up to about 400 mg per day (300 mg per day in pregnant women) is generally considered safe;181 however, the decision to use caffeine and the amount and frequency of use should be based on individual assessment of benefits versus undesirable side effects.

D4. Sleep Hygiene and Treatment of Sleep Disorders

Sleep disturbances that are not necessarily related to the work environment contribute to work-related errors and injuries.182 Sleep hygiene, which refers to a set of behavioral and environmental recommendations intended to promote good sleep,183 can help to obtain adequate duration of quality sleep. The recommended amount of sleep for the average adult is 7 h per night or more.105,106 Achieving this on a regular basis provides some degree of resilience against the adverse cognitive effects of subsequent sleep loss.184,185

Sleep hygiene environmental recommendations, which pertain to personal sleep spaces as well as any workplace sleep facilities, include ensuring a comfortable, appropriately sized bed; minimal light exposure and noise; comfortable temperature and humidity; and no sleep interruptions (unless there is an emergency).186,187 Sleep hygiene behavioral recommendations found in the literature, which are primarily focused on the habitual sleeping environment, include maintaining regular bed and wake times; avoiding daytime naps; avoiding bright light exposure during the 2–3 h prior to sleep; avoiding large meals or alcohol consumption for at least 2 h prior to bedtime; avoiding strenuous exercise immediately before bedtime; and avoiding caffeine, nicotine, and other stimulants for at least 6 h before bedtime.187,188

In operational settings, some or all of the behavioral recommendations may not be feasible or practical to implement (e.g. because irregular work schedules may interfere with maintaining regular bed and wake time) and may actually be at odds with effective fatigue risk management (e.g. because taking a daytime nap may be needed to mitigate fatigue). That is, some of the behavioral recommendations could be counterproductive for getting enough sleep or ensuring optimal performance and safety (especially in shift work settings). Also, although they are worthy sleep health recommendations in their own right, evidence of their effectiveness with regard to workplace performance, safety and health is limited.188 The behavioral recommendations should therefore not be seen as advice against pursuing catch-up sleep or naps, or using bright light or caffeine prior to bedtime, when doing so would be the better choice for safety.

Sleep disorders (e.g. obstructive sleep apnea, insomnia), medical conditions (e.g. diabetes mellitus, gastroesophageal reflux disease, back pain), and psychiatric disorders (e.g. depression, post-traumatic stress disorder) can transiently or chronically reduce the quality and quantity of sleep.189 For affected workers this can result in poor sleep quality, insufficient sleep, or excessive sleepiness, which can then negatively affect work performance. Additionally, some medications (e.g. hypnotics, antidepressants) incur side-effects that contribute to increased risk of fatigue during work hours, either through direct effects on sleepiness, or indirectly by worsening sleep quality or decreasing sleep duration.190 Workplace education regarding sleep disorders, treatment, and implications for safety and well-being has been found to reduce risk of occupational injuries.191 Furthermore, an employer-supported sleep disorder (obstructive sleep apnea) treatment program in the U.S. trucking industry has been shown to yield significant benefits in terms of crash risk, driver retention, and medical insurance costs.192,193

D5. Wake- and Sleep-Promoting Medications

Wake-promoting (stimulant) and sleep-promoting (hypnotic) medications should be used in consultation with and under supervision of a medical provider with expertise in management of sleep/wake disturbance associated with non-traditional work hours. A discussion of their pharmacological and clinical specifications is beyond the scope of this paper but can be found in the literature.194,195 The advantages and disadvantages of wake- and sleep-promoting medications and the legal and ethical considerations for their use in operational settings have also been discussed in the literature.123,196 Certain occupations have regulations or policies that prohibit the use of some or all of these medications.

Use of wake-promoting medications may interfere with the ability to obtain adequate sleep after bedtime.197 Use of sleep- promoting medications may result in next-day residual sedation, which could impair performance while at work or commuting to work,198 although caffeine intake may be helpful to mitigate this effect.199 Long-term use of wake- or sleep-promoting medications may have unintended effects on sleep, mood, and health.195,200 Interactions with other medications or with alcohol201 may increase the risk of side effects from wake- or sleep promoting medications. Alcohol, which some individuals use as a sleep aid, may degrade sleep quality, exacerbate sleep apnea, and cause next-day sleepiness.202

Melatonin, which in the United States is available over the counter as a dietary supplement, is usually marketed and used as a sleep aid. However, the primary effect of melatonin—a hormone that is also naturally produced by the pineal gland during the evening and night—is that it can shift the timing of the biological clock.203 As such, melatonin may be used to help realign sleep-wake timing and facilitate daytime sleep for night shift workers204 or to help overcome jet lag after travel across time zones.205 Optimizing the timing of melatonin administration is critical to achieving the desired effect, with morning administration leading to delays of the biological clock (shifting sleep later) and evening administration leading to advances of the biological clock (shifting sleep earlier).206 This makes achieving optimal effectiveness complicated in practice, and mistimed melatonin administration may even result in the opposite effect of what is desired. Melatonin use is generally considered safe and side effects are minimal,207 but it is not regulated in the United States and may contain additives with adverse health effects. As with other sleep-promoting medications, a medical provider with expertise in management of sleep/wake disturbance associated with non-traditional work hours should be consulted if melatonin is considered as a potential aid for sleep difficulties in the context of working time arrangements.

D6. Bright or Blue Light Exposure

Light has the potential to shift the biological clock, and it also affects alertness.208 Morning light exposure causes the biological clock to advance (shift earlier), and evening light exposure causes the clock to delay (shift later).209,210 The magnitude of these effects depends on the duration, brightness, and color of light exposure. The brighter the light and the longer the exposure, the greater the shifting,211,212 and light that is blue or blue-enriched is particularly effective for shifting the biological clock.213 Based on these principles, manipulation of light exposure can be used to shift the biological clock by some desired amount of time, for example to facilitate adaptation to a shift work schedule.214 Optimizing the timing of light exposure is critical to achieving the desired effect.208

The biological clock tends to shift no more than a few hours per day. This approach to shift work adaptation would therefore only work well for fixed or slowly rotating shift schedules, in which shift start or end times are expected to occur at approximately the same time for several consecutive days. Even then, additional measures may be required, such as wearing dark goggles or using technological solutions to reduce bright or blue light exposure at certain times,215 making the desired effect difficult to achieve.216 Strategies for mitigating risks in shift workers that rely on minimization of extended wakefulness and sleep loss, rather than shifting the biological clock, may be more effective in practice.

Light exposure has an acute alerting effect.217 To some extent the magnitude of this effect depends on the brightness level of the light,218 and light that is blue or blue-enriched is particularly effective for promoting alertness.219 Little is known about how long the alerting effect of light exposure lasts after the exposure has ended.220 That said, use of blue- enriched white light in the workplace has been reported to improve alertness, performance, and sleep quality.221 The acute alerting effect of light, however, cannot be separated from its effects on the biological clock. Especially in shift work settings, therefore, individuals exposed to bright or blue light to improve alertness may also experience a shift of the biological clock, which may or may not be problematic depending on the situation. Furthermore, inappropriate light exposure has the potential to adversely affect sleep. In many situations, therefore, light may not be a suitable countermeasure for risks associated with shift duration.

Exposure to light at night associated with shift work has been found to increase the risk of cancer.222–224 Additionally, chronic exposure to bright or blue light therapy has been linked to retinal damage later in life.225,226 This finding awaits corroboration with additional clinical studies.

D7. Exercise and Activity Breaks

In addition to the well-known benefits of physical activity from exercise and activity breaks with regard to skeletomuscular and overall health and wellness, physical activity has some potential to improve sleep.227,228 However, exercise less than 1 h before bedtime may delay sleep onset.229 Exercise also has the potential to shift the biological clock, with morning exercise facilitating advances and evening exercise facilitating delays of the biological clock.230 In the laboratory, nightly bouts of exercise have been found to promote adjustment of the biological clock to a night shift schedule.231

During sleep deprivation, physical activity may produce a modest short-term reduction in subjective sleepiness, but there may not be any concomitant improvement in cognitive performance.232 A review of continuous exercise at active workstations found minimal evidence for increased workplace performance or productivity.233 The extent to which exercise can be used effectively as a fatigue or risk countermeasure in operational settings remains to be investigated.

D8. Fatigue Prediction, Detection and Warning Technologies

Technologies available to help manage risks from fatigue can be broadly categorized as biomathematical models of fatigue,234 tools for monitoring sleep and sleep debt,235 tools for detecting and warning about fatigue,236 and fitness-for-duty tests.237 Biomathematical models of fatigue make use of equations describing the regulation of sleep and wakefulness and the biological clock to provide predictions of sleepiness or performance impairment for a (hypothetical) average individual, based on a given sleep schedule or work schedule.238,239 Use of these models is commonplace before, during and after duty periods in commercial and military aviation,240,241 where shift scheduling is typically tightly managed. However, they are deemed to be of more limited utility in settings where shift scheduling is on demand or otherwise less strictly under operational control.242

Tools for monitoring sleep have become widely available in the consumer market over the last decade. Frequently integrated with physical activity and health monitoring systems, they usually consist of wearable sensors coupled with a smartphone-based software application.243 Such tools do not usually meet professionally accepted criteria for sleep assessment,244 but have been found helpful in allowing individuals to monitor and potentially improve their sleep245 or potentially seek medical evaluation. Tools for monitoring sleep may also serve to determine a person’s sleep schedule for use with a biomathematical model of fatigue in order to predict impending safety risks.246 A significant challenge with wearable technologies is user acceptability; people often do not tolerate wearing additional devices, especially if they are uncomfortable or obtrusive, interfere with work, or do not also fulfill other useful functions.247

Tools for detecting and warning about fatigue, commonly referred to as fatigue (or drowsiness) detection and warning systems, are manifold and diverse. Technologies for fatigue detection aim generally to measure fatigue continuously from unobtrusively observable signals that correlate with fatigue, such as various ocular measures, sleepiness-related variables derived from the electroencephalogram (EEG), indices of heart rate variability, changes in speech and voice, or facial expressions. A diversity of modalities for providing fatigue warnings have been implemented, including visual, auditory, and haptic alerts as well as warning messages transmitted to third parties for possible intervention. In cars and trucks, fatigue and error detection systems may activate driver assist technologies to help avoid accidents.69

In contrast to fatigue detection and warning systems, tests for fitness-for-duty (or readiness-to-perform) typically provide snapshot measures of fatigue and often require an individual’s active involvement such as performing a reaction time task248 or taking a balance test.249 While fatigue testing for fitness- for-duty provides an instant, objective estimate of a person’s level of fatigue, such testing does not track subsequent changes in fatigue over time unless the test is conducted repeatedly (although such changes could conceivably be predicted using a biomathematical model). Most fitness-for-duty tests are not specific for fatigue; other sources of impairment (e.g. distraction, alcohol intoxication) produce similar test results. For fitness-for- duty tests that rely on user response to assess fatigue, variations in motivation and effort may also influence test results.

Fatigue prediction, detection and warning technologies that have been shown to be both sensitive and specific to relevant levels of fatigue are rare. However, both of these accuracy attributes are important. Technologies need to be sensitive enough to detect or predict levels of fatigue that compromise performance and safety, as failures to identify these could have critical, even fatal consequences. At the same time, technologies need to be specific enough not to inadvertently trigger alerts when fatigue levels are low, as too many false alarms will quickly desensitize the user. Many of the currently available technologies are proprietary, and the extent to which they have been subjected to independent research is often limited. For most of these technologies, publicly accessible data on accuracy, reliability, and validity are limited or unavailable.

Collection of objective data in operational settings may have important implications for privacy and could carry liability for employers and employees. For example, following an accident in which fatigue is suspected to be a main contributor, sleep monitoring data may be subpoenaed to determine the amount of sleep obtained prior to the accident. This potential issue notwithstanding, when used as part of a comprehensive fatigue risk management system, fatigue prediction, detection and warning technologies may contribute to a worthwhile, data- driven process of continuous improvements in risk mitigation (see Appendix D9).

D9. Risk Mitigation Tools, Policies, and Practices

Even though shift duration and its attendant fatigue may be the source of risks to performance, safety, and health, countermeasures do not necessarily have to address the source of these risks to mitigate the risks themselves.70 There are many possibilities for “fatigue proofing” that mitigate or avoid risks through other means,250 such as use of task checklists and standard operating procedures (e.g. for end-of-shift hand-offs),5 team work strategies,29 extra staffing,144 scheduled rest breaks,81 providing safe transportation options after work shifts,251 engineering solutions (e.g. safety valves in factories, rumble strips on the road, automation in the operation of machines),5 and implementation of safety policies, procedures, and management systems23 (see also Appendix B). Furthermore, an educated workforce is a key component of risk mitigation. Fatigue risk management training should help workers, their managers, and other operational personnel (e.g. scheduling, human resources) understand how to manage work, sleep, and the application of countermeasures in order to maximize performance, safety, and health.122

Fatigue risk management training and other risk mitigation strategies may be incorporated into a fatigue risk management system (FRMS), which is a proactive, data-driven process whereby an organization undertakes a formal risk management approach to reducing the effects of fatigue in the workplace.121,252 An FRMS program involves all stakeholders (workers, management, unions, regulators), includes proactive data collection steps, and adapts to dynamically changing risks so as to be self-improving. The implementation of a FRMS is unique to the needs of a given industry and company, but the elements of FRMS programs are similar across industries (see also Appendix B). A FRMS is often embedded within a larger safety management system (SMS) to allow for integrated risk management. FRMS program managers are typically responsible for developing and implementing policies designed to mitigate fatigue-related risks (e.g. scheduling and rest requirements, and countermeasure use), investigating fatigue-related incidents and accidents, monitoring safety performance indicators, evaluating the impact of work schedules on performance and safety, providing mechanisms for individuals to report fatigue at work, providing recommendations for modifying schedules to improve fatigue-related outcomes, and providing regular training and resources to employees on best practices for managing work, sleep, and the application of countermeasures.121,124