In the course of studying and writing about the problem of fatigue in health care personnel, our group and others have suggested a multitude of possible countermeasures. In this Special Issue, a world expert on managing fatigue concerns in a variety of high-hazard industries discusses a number of possible approaches. Many countermeasures seem simple to implement at little cost. Others would require substantial reorganization of the clinical work environment and the clinical workforce. These are thought to have substantial costs associated with them, and many of them have not been studied very thoroughly for efficacy (do they work in the lab?), effectiveness (do they work in day-to-day practice?), let alone for their costs. In this article I will raise a number of questions concerning various costs of countermeasures. Few of the questions can be answered at this point, and none can be answered definitively. Attempting to answer them will be an appropriate component of a long-term strategy to come to grips with the demands of a “24/7” health care industry.
First, the growing literature on the adverse effects of fatigue on clinician performance implies that there are substantial, sometimes hidden, costs to individuals and society already imposed by sub-optimal or inefficient patient care and outright errors caused or exacerbated by clinician fatigue. Such costs are difficult to compute. Recent data show unequivocally that error rates were higher by trainees working highly fatiguing schedules vs. those working less fatiguing schedules.1 Extrapolating the excess errors in the more fatigued group from these studies to the wider clinician population is appropriate, but imprecise. On the one hand, outright errors are only the tip of the iceberg. What about orders that are written a little late, just enough to push discharge from an ICU out another day? What about the clinic or OR schedule that gets behind because people are working just a little slower. On the other hand, not all the costs apparently due to fatigue-induced errors can necessarily be recouped using any of the existing countermeasures. Thus, in the absence of very complex and long-term trials that capture direct and indirect costs due to excess fatigue of clinicians, we cannot be sure of how much money could be saved by limiting clinician sleepiness.
The current situation imposes other societal costs related to the health care workers themselves. There is growing evidence that fatigued trainees are at risk for traffic accidents, particularly when driving home after long and/or overnight duty periods.2 There is also strong evidence among truck drivers and others that occupational-related illnesses are related to sleep disturbances due to night work. The cost of such accidents, illnesses, and lost productivity is not typically ascribed to fatigue of personnel in the health care system but is a “real” cost that must also be considered.
Malpractice litigation has surprisingly not become a driving force regarding the safety and economic ramifications of fatigue. Sleep deprivation of clinicians involved in litigation has only rarely been a significant issue even though high levels of fatigue are widespread throughout health care. Over many decades, a rise in such litigation has been expected each time the issue is highlighted in the media (e.g., after the Libby Zion case) or with increasing unionization of residents, but so far this has not happened. If it should ever happen, it will certainly amplify the actual costs of fatigue-induced errors or negative outcomes.
Finally, one of the most uniform findings of fatigue studies of health care personnel is that tired clinicians have negatively altered mood, which surely cannot contribute to caring and humane interactions with patients.3 It is hard to put a cost on callous or insensitive patient care, but the public often seems to place a higher premium on personal attention and caring than on technical excellence. Thus, there are clear and major opportunities to reduce costs and improve care by ameliorating sleepy clinicians. Measuring these costs, or the extent to which they are mitigated by countermeasures, will be very difficult. It is hard to measure the benefits of the “accident that did not occur” (especially when the accidents are uncommon), or gratified patients who otherwise would have been upset with their care. The signals of “safety” are inherently weaker than the signals of “production.”4,5
Having said that, let us look at the diverse countermeasures that have been suggested, as well as some of the direct and indirect costs they would impose on the health care system. Although these costs have never been measured, and cannot be detailed exhaustively in this article, many of them can be articulated for consideration in any future comprehensive financial modeling of potential fatigue intervention policies.
A seemingly easy countermeasure is to educate clinicians about the science of sleep and the ways in which they can act to protect their health and improve their performance. Educational modalities (videos, lectures, and seminar series) require moderate investment to produce and disseminate. For clinicians to attend such training may pull them out of revenue-producing work, and has to be traded off against other sorts of training that they are required to take. And while the costs may be only moderate, the efficacy and effectiveness remain to be tested.
One typical target of education is to exhort clinicians to make sleep a higher priority in their lives. On the one hand, requiring health care professionals to come to work well-rested is no different from forcing them to come to work sober and drug-free. On the other hand, to some degree this policy shifts the cost “burden” onto individual clinicians who are being asked to give up an hour or two of their other activities–whether moonlighting or spending time with their families–in favor of sleep. While they will also recoup some direct benefits to their health, mood, and possibly traffic safety, in part they are being asked to bear the costs of the health care system’s 24/7 demands, rather than having the system bear those costs collectively by hiring more staff.
Fitness for Duty Tests/Standards
The logical conclusion of a variety of countermeasures that attempt to change the pre-shift physiological state of clinicians would be to establish metrics and standards for “fitness for duty.” In this case, clinicians who are impaired would not be allowed to work that day or that shift. The costs for implementing such standards could be high. Replacement clinicians would need to be immediately available to substitute for those who are not fit for duty. Measures would have to be taken to ensure that the system is not abused by those wanting extra time off. And, while some causes of sleep deprivation can be avoided by proper sleep hygiene, others (such as the sick child at home) cannot reasonably be prevented even by the most dedicated clinician. In addition, the health care system already copes with absences due to illness, so adding “fatigue impairment” to other more common illnesses might be a tractable marginal cost.
Work Schedule Changes and Work Hours Limitations
A common countermeasure being imposed around the world, at least for trainees, is limiting total work hours and on duty-periods. The European Union is currently at a 56-hour weekly limit, with other limits on shift duration and rotation. Eighteen months ago, the ACGME imposed regulations on all residency programs in the U.S. that required substantial changes in work schedules. The imposed limit of 80 hours per week is far more lenient than European limitations. Recent studies published in the New England Journal of Medicine demonstrate that altering work schedules can decrease error rates.1 But, the altered schedules require hiring more net clinical employees, or shifting work from trainees to others. Both are relatively expensive solutions. Calculating the costs of changing the schedules is not easy. There are also indirect costs to the system. Medical educators are struggling with how to provide the same training in the fewer hours available. Some solutions to this problem will themselves have costs.
Some studies in other industries (and some in health care) show a benefit to at-work napping during long or overnight duty periods.6 Even short naps of 45 minutes have been shown to be beneficial in some settings, improving alertness and performance.7 Naps are common for trainees working 24-hour (or longer) shifts, but they are available haphazardly, and during many shifts little if any sleep can be accrued. Scheduled guaranteed naps might be useful for such personnel or even for night shift workers during only 8 or 12 hour shifts. What would it take to provide a guaranteed nap for such workers? In some cases it might be feasible for existing personnel to cross cover sufficiently to allow scheduled naps. In other cases it might be necessary to add one or more individuals on the night shift to provide rotating coverage for the napping individuals. Modeling the costs of on-shift napping will be difficult and will depend heavily on the exact staffing and work demands of each clinical setting. Union rules may play a role in what can and cannot be done with existing personnel versus with additional clinicians.
Naps taken after the shift is completed might help with the safety of the drive home and might also reduce accumulated sleep debt. If they are voluntary, the only costs are to provide the facilities for departing workers to nap. Experience shows that most workers would prefer to depart tired in order to get home earlier. Those with very long drives home, or those likely to face heavy traffic, may be the most likely to choose to nap rather than depart immediately. Thus, unless the preferences of workers are changed, or the nap is a mandatory use of unpaid time, the post-shift nap is unlikely to be a robust strategy for the workforce as a whole, even though its costs are modest.
Many clinicians already use a drug—caffeine—on a regular basis to promote alertness during the night. But caffeine is an imperfect drµg and is generally not used very wisely. The new alertness enhancing drug modafinil (approved for the use in patients with narcolepsy, daytime sleep complaints in patients with obstructive sleep apnea, and shift work sleep disorder) may provide an alternative. Whether shift workers should use modafinil on a regular basis is one question (see Dr. Lydic’s and Dr. Smith’s articles in this issue). But if they should, or if they do, who will pay the costs of the drug? Who will pay the costs of any side effects generated by taking the drug? And what might be the costs of litigation that might arise should medical errors be made that either a patient’s attorney, or the attorney of the clinician, ascribe to the use of the drug?
The application of higher levels of ambient light at the right time periods has been shown in laboratory studies to reset circadian rhythms and could be used to aid shift workers.8 Special facilities might be needed to allow workers to obtain light therapy at the right time. The timing of light therapy is very critical to its success—thus, as in the case of napping, relief personnel might be required to allow others to obtain light therapy.
In conclusion, one thing is certain. The emerging research indicates strongly that, both before and after the ACGME rule changes for residents, our health care industry and our professions have ignored substantial costs to patients and society due to fatigue-induced errors, poor performance, and automobile accidents by sleepy clinicians. After years of clamoring for change in the work and duty hours for physicians in training, significant but limited changes have been imposed. Further changes seem warranted, for this group as well as for experienced health care personnel. We must acknowledge that these reforms carry a cost. For the strongest interventions, the costs are likely to be high. Modeling the direct and indirect costs is not a simple matter, and must take into account a large variety of factors that may differ strongly in different work settings, and at different institutions. They may also differ based on the health care discipline involved and the administrative structures of the different components of the workforce. Still, cost and policy modeling is a mature art, if seemingly arcane to most clinicians. The time is right for some serious research on the organizational, logistical, and cost implications of the various proposed countermeasures to best inform policy choices. We call on our colleagues in health services research to become interested in the economic and policy implications of the fatigue issue, and help us find the best solutions to protect our patients, and improve our own health in the most expeditious and cost-effective manner possible.
Dr. Gaba is a Professor in the Department of Anesthesia and the Associate Dean for Immersive and Simulation-Based Learning at Stanford University, Stanford, CA.
- Landrigan CP, et al. Effect of reducing interns’ work hours on serious medical errors in intensive care units. N Engl J Med 2004;351:1838-48.
- Barger LK, et al. Extended work shifts and the risk of motor vehicle crashes among interns. N Engl J Med 2005;352:125-34.
- Veasey S, et al. Sleep loss and fatigue in residency training: a reappraisal. JAMA 2002;288:1116-24.
- Gaba DM. Structural and organizational issues in patient safety: a comparison of health care to other high-hazard industries. California Management Review 2000;43:83-102.
- Reason J. Human Error. Cambridge: Cambridge University Press. 1990:1-302.
- Dinges DF, Broughton RJ. Sleep and Alertness: Chronobiological, Behavioral, and Medical Aspects of Napping. New York: Raven Press, 1989.
- Rosekind MR, et al. Crew Factors in Flight Operations IX: Effects of Planned Cockpit Rest on Crew Performance and Alertness in Long-Haul Operations. NASA Technical Memorandum #108839 Moffett Field, CA: NASA Ames Research Center, 1994.
- Czeisler CA, Dijk DJ. Use of bright light to treat maladaptation to night shift work and circadian rhythm sleep disorders. J Sleep Res 1995;4:70-3.