Fatigue and Error
Matthew B. Weinger, MD discusses the relationship between fatigue and lab error, as well as steps managers can take to reduce the effects of fatigue in the lab.
What is the relationship between fatigue and lab error, and what can lab leaders do to minimize problems caused by fatigue? This interview with Dr. Matthew B. Weinger addresses the relationship between fatigue and error. Dr. Weinger is the director of the center for perioperative research in quality, the Norman Ty Smith chair in patient safety and medical simulation, and professor of anesthesiology, biomedical informatics and medical education at the Vanderbilt University School of Medicine. He also practices at the Middle Tennessee VA Healthcare System. Michael Astion, MD, PhD conducted this interview.
Dr. Weinger teaches and conducts research in patient safety,
human factors, and clinical decision making.
Q: What is fatigue?
A: Fatigue is a global term encompassing the effects of acute sleep loss, chronic sleep loss, and physical and mental exhaustion. Obviously, these factors interact. In addition, fatigue is exacerbated by personal factors like emotional stresses, as well as other negative work factors, such as high work volume and cumbersome processes.
Q: What errors are particularly prone to fatigue?
A: Fatigue is particularly dangerous in situations where a rare, but very salient signal has to be detected; multitasking and prioritization are key elements of work; there is a time gap between when information becomes available and when it has to be used; and creative thought is required. More than one of these circumstances can exist concurrently. Table 1 lists some examples from clinical labs.
Lab Work That Is Particularly Prone to Fatigue-associated Errors
Rare, but salient signal needs detecting
Unusual error flag in automated system
Detecting a low-abundance, pathogenic organism on a microscope slide
Responsibility for operating multiple instruments on the same shift
Operating instruments while handling phone calls
Time gap between when information appears and when it is used
Critical value calls requiring a physician call back.
Work requiring creative thought
Managing a crisis caused by a physical threat like a spill or power failure
Troubleshooting uncommon errors
Manually validating an infrequently encountered result
Q: What are some of the multitasking/prioritization errors we might expect from a fatigued worker?
A: A couple of patterns are observed. The first is called load shedding. Here, the worker has two or three priorities to take care of. The highest priority task is addressed, but the secondary and tertiary tasks, though they have to be completed, are neglected, delayed or performed less diligently. For example, the worker’s main priority might be a high-volume, random-access analyzer used for stat chemistry tests, the secondary priority might be a batch analyzer running concurrently, and the tertiary priority might be manual, kit-based testing that can be performed anytime during the shift.
Sleep Deprivation and Impairment
“Most studies of recurrent partial sleep deprivation have suggested that sleeping only 5 to 6 hours a night can lead to impairment. These decrements in performance accumulate with continued partial sleep deprivation as may be seen in individuals with chronic insomnia (defined as difficulty sleeping on a frequent basis) or in physicians working regularly recurring call or night shifts. In the early morning hours, after nearly 24 hours without sleep (e.g., at the end of a difficult night on call), psychomotor performance can be impaired to an extent equivalent to or greater than is currently acceptable for alcohol intoxication” (from reference 2).
Q: Please describe the second pattern.
A: In the second pattern, workers choose the first priority correctly, but problems develop when addressing lower priority tasks. After moving to the lower priority task, they tend to get stuck there, often on relatively trivial issues, and forget to return to the high priority.
Q: Can you give us some insight regarding why fatigued workers are at higher risk for failing to follow through on critical value calls?
A: Critical values are often handled by the same lab personnel performing testing. Therefore, a critical value requires multitasking and prioritization, tasks that are more susceptible to fatigue. In addition, for some critical values, there is a time delay between when information is available to the technologist and when it is communicated to the care provider. This time gap is the period the technologist may wait for a call back after paging the physician. During this gap, the technologist will return to work on other priority tasks. Sometimes the call back never comes and the technologist forgets about the call. In addition, the call back represents an interruption of the intervening task and an opportunity for fatigue-induced omission errors.
Q: Are there certain tasks that are more resistant to fatigue?
A: Pattern recognition tasks are fairly resistant to fatigue. Essentially, automatic responses are well preserved, as long as they are activated. For example, “textbook” diagnoses and guideline-based therapeutic decisions are relatively well preserved in sleep-deprived doctors.
Q: Are there times when pattern recognition tasks will fall victim to fatigue?
A: Yes, when there are lots of patterns to look at or when there is an uncommon pattern requiring creative thinking. Creative thought suffers in a fatigued worker.
Q: What is the relationship between fatigue and age?
A: In general, fatigue adversely affects people above age 60 more than it does younger people; however, this is mitigated somewhat by experience. For pattern recognition tasks, more experienced workers have more deeply ingrained heuristics or mental schema that will be less susceptible to fatigue.
Q: What is the relationship between fatigue and morale?
A: Chronic fatigue tends to have a negative effect on morale and mood.
Q: What is your view on automation as a solution to fatigue?
A: Automation is an important solution to the problem of fatigue provided the automation is well designed and reliable. Automation tends to produce fewer errors; but when errors do occur, the results can be more disastrous especially if the worker has been “out of the loop” and must figure out what has gone wrong. Therefore, it is important that automated systems provide informative monitors and alarms regarding the state of the system and its likelihood for failure.
Q: Obviously, we want to have well rested workers and work processes that are less susceptible to fatigue-induced errors. But, in cases where lab workers are fatigued, are there temporary interventions besides coffee that can help a person stay alert and decrease the probability of error?
A: On a temporary basis, bright lights, physical activity, conversation, and avoiding junk food are all helpful. The use of pharmacologic interventions is more controversial and has not been well-studied in the healthcare setting. Current research has been focused on the drug Provigil (modafinil). The claims for this drug are that it reversibly blocks fatigue, is not habit forming, and avoids some of the over-arousal associated with caffeine.
Q: What are some of the current controversies related to fatigue and medical errors?
A: One important issue is whether our approach to reducing work hours is causing more problems with transitions of care, for example, handoff errors. Another topic of debate is whether residents receive sufficient exposure to clinical medicine now that work hours have been reduced.
Q: Can you elaborate on the handoff errors?
A: If physicians and other healthcare workers work less, especially avoiding very long shifts, they will be more alert, and this could decrease many kinds of errors. There is, however, a tradeoff. With shorter and fewer shifts, there are more transitions in care where data and system status information must be handed off from the off-going to the on-coming worker. This means there are more opportunities for communication failures, which are one of the most common types of medical errors. Essentially, by shortening work hours, we are trading off fatigue-related errors for handoff-related errors. It is likely that the best approach to this is to develop strategies to improve handoffs, since it is unlikely that there will be a return to the days where clinicians work greater than 80 hours per week.
Q: What are some approaches to minimizing fatigue without creating too many handoffs?
A: One of the best approaches is to use innovative scheduling of shifts. For example, a complete shift changeover at 8a.m. is a significant opportunity for lost information due to multiple handoffs. A better approach would be to use staggered shifts in which a few people come to work or leave work at 4a.m., a few at 6a.m., etc. Staggered shifts lead to better transitions of care because there is always a cohort of workers who are immersed in the system status and are aware of current and ongoing issues. Staggered shifts also allow the option of using the well-rested workers for the most difficult tasks.
Q: Many reference labs run a majority of their tests on the night shift so that the results are available to physicians in the morning. Is the night shift particularly prone to errors?
A: There are insufficient data specific to healthcare; however, studies of shiftwork in other professions consistently find that night-shift workers are at higher risk for making errors. Night-shift workers are also more prone to injuries, such as needlestick injuries, and have poorer occupational health.
Q: From the perspective of reducing errors, is it better for workers to consistently work the night shift, or is it better to work the night shift occasionally as part of a rotation of many workers?
A: Workers who consistently work the night shift are less error prone than those working nights intermittently. The ability for a night worker to optimize performance is based on the ability to establish a new circadian rhythm. This adaptation varies from individual to individual but nonetheless requires the ability to sleep consistently during the day. This is not an easy task for most people, especially given the demands and expectations of most people’s family and personal lives.
Q: In the clinical lab, workers are sometimes asked to work a second shift due to labor shortages. What are the vulnerabilities in working two shifts?
A: In general, for workers who normally work a day shift or an evening shift, for example 4 p.m. to midnight, a double shift involving days or evenings is not as bad as one involving the midnight shift. When the midnight shift is involved, the effect of sleep deprivation comes strongly into play and the risk of errors increases significantly as the second shift progresses.
Q: Besides optimizing the amount of staffing, what can be done to reduce the risk associated with two shifts?
A: I think you can do a lot with innovative scheduling. For example, try to put the least amount of work where workers are most vulnerable, which is at the end of the second shift. Staggered shifts can also be helpful here. For example, to cover for a missing night shift worker, you can have one evening shift employee stay for 4 hours extra and another come in 4 hours early. Although this is not ideal, the effects of sleep deprivation and fatigue may be less for both workers. The worker who stays 4 hours more will be tired but not as tired as if he or she stayed for the full 8-hour shift. The worker who comes in early will not get as much sleep as they normally do, but they should get enough sleep, and this lowers their risk of making an error.
Q: What do you think about breaks and naps?
A: Breaks are a good idea for reducing fatigue, and for many people a short nap during a break is helpful. Naps during a work break are a complicated issue. For most people naps of either less than 15 minutes or greater than 2 hours are helpful and will reduce fatigue-related errors. Naps between 20 minutes and 2 hours can be problematic because of sleep inertia, which refers to the significant impairment occurring immediately upon waking from a nap. This disorientation is worse for intermediate length naps. Another problem with naps is the variability between people regarding an optimal nap length. Some people sleep for 15 minutes and feel great and can get right back into a work task, while others are significantly impaired and need time to acclimate before they are again working optimally.
Q: Do you have some practical advice regarding napping?
A: Napping is not a good idea if you are the only person working on the shift. But if you have people who cross cover, and you are obviously sleepy, a nap of less than 15 minutes during a required work break may be a good idea.
Q: You are an anesthesiologist. What are some interventions that are being used in your field regarding reducing fatigue?
A: Where I work, we are decreasing shift lengths. For example, a faculty shift can not exceed 16 hours. In addition, residents are under work-hour restrictions and cannot work more than 24 hours at one time, nor can they work more than 80 hours per week. Anesthesiologists are also openly talking about fatigue risks and are monitoring ourselves and our colleagues for fatigue. I am hopeful that other physicians and healthcare workers will come to realize that it is not unprofessional to admit that they are tired, and they will take appropriate actions to minimize the effect of their fatigue on patient care.
Cao CG, Weinger MB, Slagle J, Zhou C, Ou J, Gillin S, Sheh B, Mazzei W. Differences in day and night shift clinical performance in anesthesiology. Hum Factors 2008;50:276–90.
Weinger MB, Ancoli-Israel S. Sleep Deprivation and Clinical Performance. JAMA 2002;287:955–957.
Breaks, even naps, can help reduce fatigue.