Emergency Physicians' Fatigue Study

Date: March 28, 2023 Reference: Fowler et al. Objective assessment of sleep and fatigue risk in emergency medicine physicians. AEM March 2023 Guest Skeptic: Dr. Justin Morgenstern is an emergency physician and the creator of the #FOAMed project called www.First10EM.com Case: You arrive at 7am to relieve your colleague after a night shift. You find her at the desk, asleep with her face on the keyboard, patient documentation half finished. As she tries to wipe the drool out of the keyboard, you ask how she has been sleeping recently. She confides that she is worried that fatigue might be impacting her care of patients. Background: This might come as a surprise to some listeners, but emergency physicians are frequently tired. Realistically, this episode may not require a background section. All emergency physicians are intimately aware of the impacts of shift work and the resultant poor sleep. However, as we struggle to cope with our constant exhaustion, we may lose track of the many detrimental effects of poor sleep. Our sleep schedules impact our overall health. Shift work is associated with increased rates of cancer, cardiovascular disease, and accidents. (Knutsson 2003) I imagine many of us know of a doctor who has been in a car accident when driving home from work. However, for most clinicians, it isn’t the personal risk that bothers us. The big concern is that fatigue impacts the care we provide for our patients. Even moderate levels of fatigue can impact performance similarly to being intoxicated with alcohol. (Dawson 1997) Industrial studies indicate errors increase by as much as 30-50% on night shifts. (Akerstedt 2010) Evidence is somewhat limited in medicine. We have mostly studied residents and how badly we treat them, while ignoring the plight of staff physicians. But there are numerous studies tying fatigue to clinical errors, impaired cognition, reduced empathy, and increased interpersonal conflict. Clinical Question: What is the percentage of time that emergency physicians spend in a fatigued state? Reference: Fowler et al. Objective assessment of sleep and fatigue risk in emergency medicine physicians. AEM March 2023 Population: A convenience sample of emergency physicians from a single academic emergency department. Intervention: Sleep periods were recorded with actigraphy, using a commercially available device that measures wrist movement. Comparison: None Outcome: A “Readiscore” fatigue score was measured before and during clinical shifts. This score consists of 3 factors: sleep quality, sleep duration, and sleep efficiency (total sleep time divided by total time in bed). This is an SGEMHOP and we are pleased to have two of the author on the show. Lauren Fowler PhD Dr. Lauren Fowler is a Professor of Neuroscience at Wake Forest University School of Medicine who teaches medical students not to be afraid of the Neuroscience module. Much of her work focuses on physiological variables related to circadian desynchronization and how fatigue affects healthcare worker cognition, empathy, burnout, and perceptions. She is also studying sleep, fatigue and cognition in cancer survivors (breast and prostate), with the aim of improving cancer survivor sleep to minimize detrimental effects of sleep loss on cognition, pain perception, and quality of life. Dr. Emily Hirsh Our other guest is Dr. Emily Hirsh. She is an Associate Professor of Emergency Medicine at the University of South Carolina School of Medicine Greenville. She also serves as the Director for Faculty Wellbeing in the Department of Emergency Medicine. She survived severe burnout, left academic medicine for a while, completed a two-year fellowship in Integrative Medicine, and then realized that she wanted to help transform the health care system into one that truly cares about the people that work within it, so that they can live their unique, noble desire to care for patients and be able to do so sustainably and happily for many, many years.  Authors’ Conclusions: “Fatigue is an issue for many EPs. The present study addressed the percentage of time EPs are in a fatigued state when on shift over an extended duration of time. More research is needed to examine system-level interventions for reducing fatigue in EPs.” Quality Checklist for Observational Study: Did the study address a clearly focused issue? Yes Did the authors use an appropriate method to answer their question? Unsure Was the cohort recruited in an acceptable way? No Was the exposure accurately measured to minimize bias? NA Was the outcome accurately measured to minimize bias? Yes Have the authors identified all-important confounding factors? No Was the follow up of subjects complete enough? Yes How precise are the results/is the estimate of risk? Unsure Do you believe the results? Yes Can the results be applied to the local population? Yes Do the results of this study fit with other available evidence? Yes Funding? Prisma Health Sciences Center seed grant Results: Of 131 emergency physicians in the department of emergency medicine, 17 volunteered to participate in the study. Nine (53%) were female and two (12%) were full time nocturnists. The collected data for a total of 392 shifts, or 23 shifts per participants over a 2-month period. Key Result: Emergency physicians were fatigued for almost one-quarter of their time on shift. Primary Outcome: Readiscore fatigue score (sleep quality, duration and efficiency) Sleep Quality: Averaged 7.7/10 (SD 1.84), which they state is indicative of poor sleep quality. Sleep Duration: Mean number of hours slept per night was 6.77 (SD 1.84) Sleep Efficiency: Mean 87 (SD = 9) Participants spent 725 h (23.52%) on shifts with fatigue scores indicative of significant impairment Shift start time only accounted for 1% of the variance among Readiscores, although they did note a trend to lower Readiscores with shifts that started both earlier and later in the day. Shift type (day, evening, night) was significantly associated with fatigue score, where night shifts were associated with higher fatigue scores. Listen to the podcast to hear Emily and Lauren answer our five nerdy questions. 1. Blinding: Participants were blinded to their sleep data during one of the two months of the study, but it is not clear if they were blinded to the hypothesis or intent of the study. 2. Hawthorne Effect: Participants knew they were in a study focused on sleep. They only had the sleep tracker available during one of the two months, but it is possible that simply being observed may have changed their usual sleep habits. 3. Selection Bias: Only 17 of 131 available physicians volunteered to participate, which creates a high risk for selection bias. Is there any data on how these participants compare to the individual who decided not to participate? They were also from one academic center which makes it difficult to extrapolate to other practice environments like non-academic urban centers, community EDs, pediatric EDs, rural ED or critical access hospitals. We do not know much about those who decided to participate. Were they known to have good or poor sleep prior to the study? Did they use any pharmacologic or non-pharmacologic sleep aids? How much caffeine did they consume on average? These and other baseline characteristics would have been very helpful to understand who volunteered for this study. 4. Lack of a Comparison: In isolation, this sleep scores are very difficult to interpret. This is not a standard measure. The result 7.7 out of 10 sounds pretty good to me, but you state in the results that it is indicative of poor sleep quality. Comparing to other medical specialties, other healthcare professions (RNs, PAs, NPs, etc), other professions (aviation, transportation, military), other shift workers (security, factory workers, etc) or even EPs on days when they weren’t working or on vacation would provide valuable context to the clinical meaning of this score. 5. The Readiscore and Clinically Significant Numbers: We are all familiar with research scores in which relatively large numerical differences can nevertheless remain undetected by patients. How well does the Readiscore correlate with subjective feelings of fatigue, or objection function on testing? What is the minimal difference on a Readiscore that would be noticeable to a practicing physician? For example, do you think the small variance between Readiscores by shift start time has any real-world impact? Does a statistical difference in the readiscore contribute to clinically important medical errors? Comment on Authors’ Conclusion Compared to SGEM Conclusion: We agree with the authors conclusions. A friendly amendment would be to modify the last sentence to say; “More research is needed to examine system-level interventions for reducing fatigue in EPs and if any reduction would result in an important benefit to patients.” SGEM Bottom Line: Fatigue is an issue for many emergency physicians, and perhaps an issue many of us still need to take more seriously, for our own sakes and for our patients. Case Resolution: You listen compassionately to your colleague but realize that 7am after a night shift is not a great time for further discussion, so take handover as quickly as possible to get her home to rest. However, you do set up departmental rounds to discuss strategies to improve sleep quality and manage the difficulties of shift work. Justin has a great First10EM blog post called “Some evidence for working night shifts”. It covers a paper by Wallace and Haber on the Top 10 evidence-based countermeasures for night shift workers.  Dr. Justin Morgenstern Clinical Application: Part of being an emergency clinician is acknowledging the impact that fatigue has on our performance,