Implementing fatigue optimizing scheduling to reduce resident error

Brady T. Evans, BS
Harvard Medical School, Boston, MA

Frank McCormick, MD
Harvard Combined Orthopaedic Residency Program, Massachusetts General Hospital, Boston, MA

John Kadzielski, MD
Harvard Combined Orthopaedic Residency Program, Massachusetts General Hospital, Boston, MA

Christopher P. Landrigan, MD, MPH
Division of Sleep Medicine, Brigham and Women’s Hospital, and Division of General Pediatrics, Children’s Hospital Boston, Boston, MA

James Herndon, MD, MBA
Chairman Emeritus, Partners Department of Orthopedics, Boston, MA

Harry Rubash, MD
Department of Orthopedic Surgery, Massachusetts General Hospital, Boston, MA


Abstract:

Objective: Medical error is a major cause of death in hospitals worldwide, and fatigue may be a significant contributor. The purpose of this investigation is to critically assess fatigue levels in orthopedic residency programs using the Sleep, Activity, Fatigue, and Task Effectiveness (SAFTE) model and Fatigue Avoidance Scheduling Tool (FAST) via computer simulation. We will then seek to identify specific times of day when residents are most affected by fatigue, implement countermeasures, and use FAST to reduce impairment secondary to fatigue.
Methods: Work schedules from three rotations (day shifts, trauma coverage, and night-float) were assessed by FAST to identify specific time intervals in which residents were potentially most impaired by fatigue. Impairment periods were defined as an effectiveness score of less than 70, which is correlated to a Blood Alcohol Content (BAC) of 0.08, considered legal intoxication. Fatigue countermeasures were then applied to the predicted impairment time interval to determine the plasticity of predicted impairment.
Results: Residents on day rotations had zero periods of predicted fatigue. Residents on trauma coverage had significant impairment only during the early morning hours of overnight call. However, residents on night-float demonstrated significant fatigue impairment while providing coverage. Importantly, “Fatigue Optimization Scheduling” countermeasures for the trauma coverage and night-float rotation resulted in zero time-periods of predicted impairment.
Conclusion: The FAST enables us to identify, quantify, and then modify schedules with high levels of resident fatigue and the potential for error. We have shown there is potentially a significant amount of fatigue impairment in residents at specific time intervals based on idealized schedules. Importantly, impairment due to fatigue can be minimized if properly addressed.


Introduction:

Over the past decade, medical error has become a major focus of research and an impetus for change in hospital systems and residency programs around the country. The Institute of Medicine’s “To Err Is Human: Building a Safer Health System” ranked medical error as the 6th to 8th leading cause of death in the United States(1). Additional studies have documented similarly extensive morbidity and mortality of medical error worldwide (2-4). While it is clear that medical error is a significant problem(5), efforts to reduce medical error have been largely unsuccessful(6).

Considering fatigue as a contributing factor to medical error, the ACGME limited resident work hours to 80 per week in 2003. The Institute of Medicine followed in 2008 with “Resident Duty Hours: Enhancing Sleep Supervision and Safety” which recommended further limitations of resident work hours to better protect patients and health care workers against fatigue-related errors.7 Since the initial limitations on resident work hours were instituted, medical error has been linked to resident fatigue in a series of studies.8,9 Residents working shifts exceeding 16 hours have been found to make more errors in the care of their patients10-13 and to experience more injuries at work and while driving home from work.14,15 Studies in other industries, including aviation and the railroad industry, have likewise shown that fatigue alone increases the risk of error up to four times more than drugs or alcohol.16,17 Despite the accumulation of recent data however, debate persists about the importance of fatigue on error rates in surgery or medicine. Furthermore, the effectiveness of limitations in work hours in reducing error remains controversial.18

In addition to number of hours worked, circadian rhythm disruption, sleep duration and quality, and extended shifts all contribute to fatigue and predispose individuals to impairment.19 For this reason, work hour restrictions alone may be limited in their ability to reduce resident fatigue and in turn decrease error.

The Fatigue Avoidance Scheduling Tool (FAST), which utilizes the Sleep, Activity, Fatigue, and Task Effectiveness (SAFTE) model, can potentially be used to monitor and to reduce resident fatigue. The FAST identifies specific time periods in which individuals are likely to be impaired due to fatigue.20

In this study, we analyzed a series of schedules typically used in orthopedic residency rotations using the FAST. Our objective was to determine which schedules most predispose residents to fatigue, and further identify the specific time intervals that are most problematic. We then sought to test the effects of possible schedule changes predicted to mitigate fatigue-related impairment and potentially reduce errors related to resident fatigue in a novel fashion.


Methods:

Work schedules for three different rotations were analyzed using a computer model of the FAST system. These schedules included day shift (6 a.m. to 8 p.m. during the week), trauma coverage (4 a.m. to 6 p.m. with Friday night overnight call), and night-float (6 p.m. to 8 a.m., six days per week). These schedules represent a range of overnight hours worked, shift length, and schedule variation, and represent many of the most common schedules that residents actually work. For most workdays in day shift and trauma call, residents work 14-hour days, leaving 10 hours for sleep and other activities. We assumed residents would use 1 hour for commuting (30 minutes each way), 1 hour for reading, and 1 hour for fitness, eating, personal hygiene, and family, and would sleep 7 hours per night. Sleep was assumed to be high quality. On every day off, residents were assumed to sleep 8 hours during the night, with a daytime nap of 1 hour.

These simulated schedules were analyzed using the SAFTE model, which has been independently validated against other mathematical models of fatigue and performance in controlled assessments and has predicted the effects of fatigue on performance with a reliability of 0.94.21 The performance effectiveness score is calculated from the SAFTE model and represents composite human performance on a number of cognitive tasks. It is scaled from 0 (100% impaired due to fatigue) to 100 (fully awake and unimpaired by fatigue).22 This effectiveness score also correlates with the Blood Alcohol Content (BAC) that would result in similar levels of impairment. Significant impairment was defined as an effectiveness score less than 70, which is equivalent to a BAC of 0.08.9,23,24 Importantly, these scores are correlated to increased risk for error and previous studies have shown that the SAFTE model is a validated tool for measuring performance readiness and fatigue risk.25,26 The SAFTE model is currently used by multiple federal organizations, including the United States Department of Defense.

The FAST is a computer application of the SAFTE model that allows a continuous record of predicted fatigue during periods of measurement based on scheduling alone. The FAST output is in the form of an easy to read chart and can be used for schedule management and root-cause error analysis.

In our study, when a schedule resulted in significant fatigue, targeted fatigue countermeasures were applied and the revised schedules were reanalyzed using the FAST program to assess the effectiveness of the countermeasures. Specific countermeasures included: replacing 24-hour overnight call shifts with two 12-hour shifts and a 4 a.m. 30-minute nap for those on night-float.


Results:

For the optimized day shift schedule (6 a.m. to 8 p.m., six days a week), there were no periods of significant predicted impairment due to fatigue (Figure 1). In this case, no fatigue optimization was performed.

Figure 1. FAST output for the day shift schedule. Thick red lines indicate periods where residents are awake and thin black lines indicate periods of sleep. For the day shift schedule, there are no periods where residents are predicted to be impaired by fatigue, which would be seen above by intervals below 70% effectiveness.

For residents on the trauma rotations, the FAST program predicted significant levels of impairment due to fatigue during the early morning hours of overnight call (2 to 6 a.m.) at the end of a 24-hour shift (Figure 2). Changing the 24-hour shift at the end of the week to two 12-hour shifts on Friday and Saturday resulted in zero periods where residents were predicted to have significant levels of fatigue.

Figure 2. FAST output for orthopedic trauma rotation schedule. Thick red lines indicate periods where residents are awake and thin black lines indicate periods of sleep. Residents are predicted to be significantly impaired only during the early morning hours of weekly overnight call, which is seen above by the red line below 70% effectiveness on each Friday.

On the night-float rotation, residents were predicted to have significant levels of impairment due to fatigue during hours of coverage nearly every night from 2 to 6 a.m. Additionally, the level of impairment due to fatigue worsened significantly as the week progressed (Figure 3). By the second week of the rotation, residents on night-float dropped below 60% effectiveness on all 5 days of the workweek and below 50% effectiveness on one of those days. This correlates with the impairment induced by a BAC of >0.11 and indicates a very high risk for serious error. For this group, fatigue optimization was accomplished by removing the 24-hour shift at the beginning of the week and allowing residents to take a 30-minute nap in the middle of their shift. These interventions resulted in zero periods of significant impairment due to fatigue (Figure 4).

Figure 3. FAST output for night-float schedule. Thick red lines indicate periods where residents are awake and thin black lines indicate periods of sleep. Residents are predicted to be significantly impaired (red lines below 70% effectiveness) by fatigue while providing coverage during each night of call. Notably, predicted impairment worsens following the first week of overnight call.

Figure 4. FAST output for night-float schedule with fatigue optimization countermeasures. Thick red lines indicate periods where residents are awake and thin black lines indicate periods of sleep. Fatigue optimization was accomplished by removing the 24-hour shift at the beginning of the week and allowing residents to take a 30-minute nap in the middle of their shift. These interventions resulted in zero predicted periods of significant impairment due to fatigue (no significant time periods below 70% effectiveness).


Discussion:

The FAST tool can predict resident fatigue across a wide range of typical schedules and can identify specific times of day that result in the highest risk of error. Potential interventions, e.g., scheduling changes and naps during extended shifts, were tested and found to be effective potential countermeasures for decreasing the risk of error due to fatigue in this model.

Resident physician fatigue, even under current work hour restrictions, is likely a major contributor to medical error. Under the current model of graduate medical education, surgical and medical residencies require residents to work extended shifts with overnight call responsibilities. Schedules during which residents provide strictly overnight coverage during the week and cover extended shifts on the weekends may be generating unsafe levels of fatigue.27 If poorly designed, schedule revisions designed to reduce work hours and limit medical error may in fact increase resident fatigue. Well-designed schedules that are built on principles of sleep and circadian science, however, can be designed with the aid of scheduling tools such as SAFTE and FAST, and these may lead to safer care.

In this study, residents on night-float rotations were at a particularly high risk of error due to fatigue, in part as a result of the circadian rhythm disruption inherent to this schedule. When overnight shifts are preceded by extended work shifts, or occur at the end of a string of consecutive overnight shifts, the risks of impairment due to fatigue may be particularly high, as chronic and acute sleep deprivation can greatly increase the impairment due to circadian misalignment alone.28 Furthermore, risk to patients may be exacerbated on night-float rotations as these residents may have the least direct attending supervision. If resident supervision improves with the forthcoming 2011 ACGME regulations, this anticipated risk may be mitigated.27
Many studies have examined the relationship between fatigue and error and, while not all of these studies have shown that fatigue leads to error, most have shown that fatigue is a major factor in the risk for medical errors.29-31 Importantly, in our study, “Fatigue Optimization Scheduling” countermeasures resulted in zero time periods of predicted resident fatigue, without decreasing the total number of hours worked by each resident. With further analysis and schedule optimization, we may be able to reduce resident fatigue and the inherent risk for error without increasing global work hour restrictions or cost of care.

However, there are several limitations to this study. First and foremost, the study was limited to assessing the predicted effects of a handful of simulated schedules that were 100% compliant with the 2011 ACGME work limits. As real residents were not studied, this analysis uses only idealized sleep schedules with perfect sleep efficiency and a very regular sleep schedule. This method likely overestimates the amount of sleep residents get and assumes very high quality sleep each night. Additionally, we also did not account for the possibility of sleeping at work while not involved in patient care. It is possible that real residents actually experience higher levels of fatigue than predicted by this idealized schedule, but, as our countermeasures show, even short naps can significantly reduce fatigue during a given shift.

Implementation of the FAST program is potentially a very useful strategy for measuring and reducing resident fatigue. However, in order to make effective changes to resident schedules, it is necessary to analyze the schedules of working residents rather than idealized situations. It would be possible to do this using daily logs, but actigraphy may be a more efficient method for measuring the sleep-wake cycles of individual residents. Actigraphy is commonly used for sleep studies and utilizes a simple wrist device that measure wrist movements in order to quantify the level of alertness of an individual.32,33 This technology could be used to collect data on individual residents that could then be analyzed using the FAST program. As the FAST output would be specific to each resident, countermeasures could be individualized. Further research is needed utilizing this technology in both resident and attending surgeons.


Conclusions:

It is possible to predict resident fatigue based on their work and sleep schedules and develop specific interventions to prevent it. Resident fatigue results from a combination of work hours, sleep deprivation, and circadian rhythm disruption. Using the FAST model provides a quick way to identify schedules associated with significant predicted fatigue. Management of sleep debt and circadian rhythm disruption may allow reduction in resident fatigue and minimize risk to the patient. By recognizing the causes of fatigue we can implement measures to mitigate further risk and provide relatively simple solutions to a difficult and controversial problem.


References:

1. Kohn LT, Corrigan JM, Donaldson MS. To Err Is Human: Building a Safer Health System. Washington D.C.: Institute of Medicine; 1999.
2. Baker GR, Norton PG, Flintoft V, et al. The Canadian Adverse Events Study: the incidence of adverse events among hospital patients in Canada. Cmaj 2004;170:1678-86.
3. Davis P, Lay-Yee R, Briant R, Ali W, Scott A, Schug S. Adverse events in New Zealand public hospitals I: occurrence and impact. N Z Med J 2002;115:U271.
4. Vincent C, Neale G, Woloshynowych M. Adverse events in British hospitals: preliminary retrospective record review. Bmj 2001;322:517-9.
5. Donaldson L, Philip P. Patient safety: a global priority. Bull World Health Organ 2004;82:892.
6. Landrigan CP, Parry GJ, Bones CB, Hackbarth AD, Goldmann DA, Sharek PJ. Temporal trends in rates of patient harm resulting from medical care. N Engl J Med 2010;363:2124-34.
7. Johns M, Bagian J, Bhattacharya J, Bisognano M, Carayon P, Cohen J. Resident Duty Hours: Enhancing Sleep, Supervision, and Safety. Washington DC: Institute of Medicine; 2008.
8. Barger LK, Ayas NT, Cade BE, et al. Impact of extended-duration shifts on medical errors, adverse events, and attentional failures. PLoS Med 2006;3:e487.
9. Arnedt JT, Owens J, Crouch M, Stahl J, Carskadon MA. Neurobehavioral performance of residents after heavy night call vs after alcohol ingestion. Jama 2005;294:1025-33.
10. Landrigan CP, Rothschild JM, Cronin JW, et al. Effect of reducing interns’ work hours on serious medical errors in intensive care units. N Engl J Med 2004;351:1838-48.
11. Lockley SW, Cronin JW, Evans EE, et al. Effect of reducing interns’ weekly work hours on sleep and attentional failures. N Engl J Med 2004;351:1829-37.
12. Levine AC, Adusumilli J, Landrigan CP. Effects of reducing or eliminating resident work shifts over 16 hours: a systematic review. Sleep 2010;33:1043-53.
13. Reed DA, Fletcher KE, Arora VM. Systematic review: association of shift length, protected sleep time, and night float with patient care, residents’ health, and education. Ann Intern Med 2010;153:829-42.
14. Barger LK, Cade BE, Ayas NT, et al. Extended work shifts and the risk of motor vehicle crashes among interns. N Engl J Med 2005;352:125-34.
15. Ayas NT, Barger LK, Cade BE, et al. Extended work duration and the risk of self-reported percutaneous injuries in interns. Jama 2006;296:1055-62.
16. Davenport NL, J. Fatigue Assessments in Mishaps: N.A.
17. NTSB. Safety Study: A Review of Flight Crew Involved, Major Accidents of U.S. Air Carriers, 1978-1990. In: Board NTS, ed. Washington D.C.; 1994.
18. Baldwin K, Namdari S, Donegan D, Kamath AF, Mehta S. Early effects of resident work-hour restrictions on patient safety: a systematic review and plea for improved studies. The Journal of bone and joint surgery American volume 2011;93:e5.
19. Mallis MM, Mejdal S, Nguyen TT, Dinges DF. Summary of the key features of seven biomathematical models of human fatigue and performance. Aviat Space Environ Med 2004;75:A4-14.
20. Hursh SR, DP; Johnson, ML; Thorne, DR; Belenky, G; Balkin, TJ; Miller, JC; Eddy, DR; Storm, WF. The DOD Sleep, Acitivity, Fatigue, and Task Effectiveness Model. Aviat Space Environ Med 2004;75:A44-A53.
21. Van Dongen H. Comparison of mathematical model predictions to experimental data of fatigue and performance. Aviat Space Environ Med 2004;75:A15-36.
22. Hursh S, Redmond D, Johnson M, et al. The DOD Sleep, Acitivity, Fatigue, and Task Effectiveness Model. Aviat Space Environ Med 2004;75:A44-A53.
23. Arnedt JT, Wilde GJ, Munt PW, MacLean AW. How do prolonged wakefulness and alcohol compare in the decrements they produce on a simulated driving task? Accid Anal Prev 2001;33:337-44.
24. Dawson D, Reid K. Fatigue, alcohol and performance impairment. Nature 1997;388:235.
25. Hursh SR, TG; Kaye, AS; Fanzone, JF. Validation and Calibration of a Fatigue Assessment Tool for Railroad Work Schedules, Final Report. In: Transportation USDo, ed. Washington DC; 2008:71.
26. Hursh SR, Redmond DP, Johnson ML, et al. Fatigue models for applied research in warfighting. Aviat Space Environ Med 2004;75:A44-53; discussion A4-60.
27. Nasca TJ, Day SH, Amis ES, Jr. The new recommendations on duty hours from the ACGME Task Force. N Engl J Med 2010;363:e3.
28. Cohen DA, Wang W, Wyatt JK, et al. Uncovering residual effects of chronic sleep loss on human performance. Sci Transl Med 2010;2:14ra3.
29. Privette AR, Shackford SR, Osler T, Ratliff J, Sartorelli K, Hebert JC. Implementation of resident work hour restrictions is associated with a reduction in mortality and provider-related complications on the surgical service: a concurrent analysis of 14,610 patients. Ann Surg 2009;250:316-21.
30. Rothschild JM, Keohane CA, Rogers S, et al. Risks of complications by attending physicians after performing nighttime procedures. Jama 2009;302:1565-72.
31. Yaghoubian A, Kaji AH, Putnam B, de Virgilio C. Trauma surgery performed by “sleep deprived” residents: are outcomes affected? J Surg Educ 2010;67:449-51.
32. Littner M, Kushida CA, Anderson WM, et al. Practice parameters for the role of actigraphy in the study of sleep and circadian rhythms: an update for 2002. Sleep 2003;26:337-41.
33. Sadeh A, Acebo C. The role of actigraphy in sleep medicine. Sleep Med Rev 2002;6:113-24.

About these ads

About Arun Shanbhag

Tinkerer, Thinker, Engineer, Scientist, Photographer, Marathoner, Indian, Writer, Editor, Publisher, Citizen Journalist, Manager, Bostonian, American

Posted on October 23, 2011, in Clinical, Residency and tagged . Bookmark the permalink. Comments Off.

Comments are closed.