Jonathan E. Prousky, ND, MSc, MA, RP
Professor, Chief Naturopathic Medical Officer
Canadian College of Naturopathic Medicine
Originally published in - Prousky J: Helping the distressed clinician by identifying and treating burnout. Townsend Letter, 2022;471:46-56. [Reprinted with permission].
The focus of this article is on clinician burnout, which is characterized by emotional exhaustion, depersonalization, and diminished personal accomplishment. One goal is to provide a comprehensive but brief overview on burnout, including its prevalence among physicians in the United States and Canada, the ways it was impacted by the COVID-19 pandemic, relevant neurobiological and neurochemical associations, its relationship to depression, the known drivers, and diagnosis. Another goal is to review specific evidence-based interventions that have the potential to attenuate chronic stress and burnout, and which may help to prevent burnout altogether. The timing is urgent for such interventions with the goal of enhancing resilience so clinicians recover from burnout and do not become repeat burnout customers.
Neil Young is credited with saying in one of his brilliant songs, “It’s better to burn out than to fade away.” Though this may have some value and cachet in the world of rock and roll, no clinician would tell anyone that their burnout was preferable over a content and productive working experience. Burnout is a brutal psychological syndrome consisting of exhaustion, cynicism, and workplace inefficiency.1 A state of exhaustion is characterized by feeling “overextended and depleted of one’s emotional and physical resources.”1 Cynicism refers to a “negative, callous, or excessively detached response to various aspects of the job.”1 Inefficiency refers to feeling incompetent at work with an associated “lack of achievement.”1 An updated definition of burnout has since renamed the three aforementioned key dimensions to that of emotional exhaustion, depersonalization, and diminished personal accomplishment.2
Burnout is clinically significant due to all sorts of cascading effects that result from this uncomfortable, detached, and exhausted way of existing within one’s place of work. Clinicians care deeply about their work, usually involving some combination of training and/or mentoring students and fellow professionals, and most importantly, caring for patients. When burnout makes its presence known, it becomes nearly impossible to shoulder the burdens of work and patient needs amidst the unending demands of trying to stay afloat while the ship (i.e., the burned-out clinician) is literally sinking in despair, or has unfortunately sunk.
This paper reviews the prevalence, neurobiology, neurochemistry, drivers (i.e., causes and consequences), diagnosis, and management of burnout as it relates to clinicians. Most of the burnout research cited in this paper comes from studies on physicians, and/or is related to the work that physicians do.
Evidence suggests that burnout is a unique clinician experience resulting from multiple influences (i.e., both positive and negative) that emanate from the workplace environment.3 The highest rates of physician burnout in the United States (US) happen among emergency medicine physicians, with the lowest rates among physicians working in preventive medicine/occupational medicine.3 The overall mean rate of burnout (i.e., as assessed by having 1 symptom of burnout, such as scoring high in emotional exhaustion or scoring high in depersonalization as per the 2-item Maslach Burnout Inventory/MBI) among US physicians increased from 45.5% in 2011 to 54.4% in 2014.4 During this same timeframe, physician satisfaction from work-life balance (WLB) declined from 48.5% in 2011 to 40.9% in 2014.4 By contrast, when burnout was assessed among employed nonphysicians in the US, the overall mean rate of burnout was 28.6% in 2011 and 28.4% in 2014.4 WLB among employed nonphysicians in the US increased from 55.1% in 2011 to 61.3% in 2014.4
More recent data has shown that the mean rate of burnout among physicians (i.e., as assessed by having 1 symptom of burnout as noted earlier) decreased to 38.2% in 2020 compared to 43.9% in 2017, 54.4% in 2014, and 45.5% in 2011.5 This same 2020 dataset showed that satisfaction from work-life integration (WLI), a reengineered term to describe WLB, was 46.1% among physicians.5 Based on these results, the mean burnout rate among US physicians compared to the general population is markedly greater, while WLI (or WLB) is markedly lower.
In Canada, the 2018 Canadian Medical Association National Physician Health Survey aggregated data from different residency and medical specialties, and different settings (e.g., hospital, private office/clinic, academic, and administrative/corporate office).6 The same burnout assessment tool that was used to generate the US data - i.e., the 2-item MBI - was also used to assess burnout among Canadian physicians. Unlike the US data, the Canadian data did not compare physician burnout and WLI to that of employed nonphysicians in Canada. In aggregate, the overall physician burnout rate was 30%. The report showed that some 32% of females and 27% of males met criteria for burnout. Some 38% of residents met criteria for burnout compared to 29% of physicians. The highest rates of burnout were among physicians in family medicine/general practice (32%) with the lowest rates among physicians holding administrative positions (19%). Similarly, physicians working in administrative/corporate offices had the lowest rates of burnout (25%) whereas physicians working in their own private offices/clinics had the highest rates of burnout (31%).
How did the global COVID-19 pandemic influence burnout? The US data showed that in the first 6-9 months of the pandemic, the mean rate of physician burnout decreased and WLI increased compared to prior years.5 However, physicians working in specific areas in the US that were considered geographic hot spots early in the pandemic experienced acute stress, as documented in several studies (see p.502 for the specific studies).5 The drivers of increased occupational stress during this time period included high case volumes, working outside of one’s specialty, providing care without adequate personal protective equipment, and managing patients before effective COVID-19 treatments had been established.5 In fact, symptoms of burnout - namely, emotional exhaustion and depersonalization - among US physicians working in medical specialties directly impacted by the pandemic “did not improve…even as these measures of burnout improved for physicians as a whole.”5 In fact, the rates of burnout during the first year of the pandemic increased over time among 4 of 5 US frontline medical specialties, with the most significant increases among hospitalists and primary care respondents.7
In Canada, preliminary data from a survey by the Canadian Medical Association conducted in November 2021 showed that “more than half of physicians and medical learners (53%) have experienced high levels of burnout.”8 This same report noted that almost half of Canadian physicians (46%) have considered reducing their workload in next 2 years, which would have considerable impacts upon the Canadian healthcare system already burdened by access to care issues. The same survey found that 59% of physicians indicated that their mental health worsened since the pandemic began due to increased workload, lack of WLI, abrupt policy/process changes, as well as other issues. The survey also noted that some 47% of physicians reported reduced levels of social wellbeing. In the Province of Ontario, a survey of physicians identified that 29% reported high levels of burnout prior to the pandemic.9 By March 2021, the same survey noted that the rate of burnout increased to 34.6% among Ontario physicians.
For some physicians, the pandemic increased a sense of meaning and purpose in their work, which had attenuating effects on burnout during the early stages of the pandemic.5 For other physicians, however, the pandemic not only caused burnout but also moral injury.10 Moral injury happens when events directly infringe upon physicians’ moral convictions, such as what happens when managing medical decisions in the unfamiliar territory of a pandemic. Physicians had to tell loved ones that they couldn’t attend the bedside of their dying relatives, or attend the births of new relatives. Physicians had to endure considerable stress by determining who would receive life-saving treatments, and for non-COVID patients having life threatening conditions in whom care would be delayed. These types of moral injuries added considerable psychological burdens to physicians who never could have imagined facing these types of restrictions and complications in the delivery of the medical care they provide.
There seems to be a strong association between chronic stress and burnout. Chronic stress broadly refers to “ongoing demands that threaten to exceed the resources of an individual in areas of life such as family, marriage, parenting, work, health, housing, and finances.”11 Chronic stress invokes a “pathological state that is caused by prolonged activation of the normal acute physiological stress response, which can wreak havoc on immune, metabolic, and cardiovascular systems.”12 Burnout invokes a similar pathological state characterized by dysregulation of the hypothalamic-pituitary-adrenal (HPA) axis and immune function.13
Burnout appears to be a consequence of chronic stress that is particular to one’s job and work environment. It has been described as a “cumulative stress reaction to ongoing occupational stressors.”1
When experiencing burnout, a person’s allostatic systems would be persistently activated. Allostasis, coined by Sterling and Eyer,14 refers to biological adjustments that allow an individual to adapt to particular challenges that happen over the lifespan. Adapting to such challenges demands the synchronous though non-linear activation of many different physiological processes, such as neural, neuroendocrine, and neuroendocrine-immune mechanisms.15 Allostasis begins in the brain and happens or is instigated by how an individual perceives and interprets any given situation. Allostasis is about adaptation, but the physiological adaptations may not ensure survival because they can become deleterious over time and cause irreversible damage. Thus, each person’s allostatic responses are unique and depend on (1) “how the individual perceives and interprets the situation” and (2) “the condition of the body itself.”15
Allostatic load (AL) represents a state when the aforementioned physiological adaptations become deleterious, as happens from repeated allostatic responses during stressful situations.16 This happens when an allostatic system fails to habituate to the recurrence of the same stressor, fails to shut off following overwhelming stress, and/or whose response is deficient resulting in heightened activation of other, normal counter-regulatory systems.15,17 If, for example, an individual perceives something as chronically stressful and has not been able to habituate to it (i.e., they lack adequate proactive planning skills and psychological buffers), then they won’t be able to shut-off their allostatic response, and this would likely accelerate brain aging and organ pathology. Additionally, if the condition of the body itself is further challenged by poor physical health and an unstable physiology, the individual would be more vulnerable to AL, which sets in motion problems such as hypertension, obesity, diabetes, atherosclerosis, inflammation, exaggerated autoimmune responses, neuronal atrophy and even neuronal death.15 Allostatic overload (AO) is accordingly an extension of AL, and describes the result of allostatic responses that lead to irreversible damage to body organ systems (i.e., pathology), and/or mental disorders (i.e., diagnosable mental illness). Burnout ought to be considered a specific and persistent work-related state associated with AL and AO, with accompanying psychological distress signals of emotional exhaustion, depersonalization, and diminished personal accomplishment.
Neurobiologically, when an individual is faced with uncertainty arising from chronic stress, and has not been able to habituate to a persistent stressor (or a persistent set of chronic stresses), specific areas of the prefrontal cortex (PFC) activate the anterior cingulate cortex (ACC).18 The ACC “assesses the degree of uncertainty about whether future outcomes are uncertain.”18The PFC “governs high-order reasoning, social cognition, and complex decision-making, including the integration, conceptualization, and critical evaluation of information.”19 In situations when the degree of uncertainty about future events cannot be reconciled, the PFC activation of the ACC triggers the amygdala - to initiate a stress response - followed by the release of norepinephrine.18 This leads to a hypervigilant state, and the simultaneous activation of the sympathetic nervous system, leading to increased glucose for energy utilization due to the high metabolic demands of the brain. There is also activation of the HPA axis leading to increased cortisol output, which plays a vital role in synaptic plasticity and learning after stress.18 The released cortisol passes through the blood-brain-barrier and binds to glucocorticoid receptors in the amygdala, hippocampus, and PFC.18
Activation of these neurobiological processes leads to feelings of threat and loss of control, concomitant with damaging alterations to brain architecture within the amygdala, hippocampus, and PFC. Specifically, the neuronal dendrites housed within the hippocampus and PFC shrink, become shorter and less branched, and these changes result in diminished synaptic output.17 These changes further compromise an individual’s “capabilities for nuanced cognitive function, memory and self-regulation.”17 The same type of chronic stress causes an expansion of dendrites and increased synaptic input to an area of the amygdala known as the basolateral amygdala, which results in heightened anxiety, aggressiveness, and vigilance.17 These changes result in more bottom-up control via the ACC-amygdala complex, and compromised emotional regulation since the PFC is unable to exert effective top-down control.18
The neurobiology of burnout appears to be similar to the neurobiology of chronic stress. In an article on physician distress and burnout, Amsten and Shanafelt review various neurobiological implications.19 They described how significant fatigue (e.g., from sleep deprivation) and uncontrollable stress severely impacts the functionality of the PFC, with a corresponding weakening of its higher-order functions, rendering the PFC to go “offline.” They noted an association between sleep deprivation and “impairments in PFC metabolic and physiologic activity correlating with cognitive deficits.” They also noted that uncontrollable stress leads to high levels of norepinephrine and dopamine being released within the brain, which diminishes PFC function, subsequently impairing cognitive functions, and causing the PFC synaptic connections to atrophy. On the other hand, the high levels of norepinephrine and dopamine that get released in situations of uncontrollable stress, strengthen (i.e., expand) the synaptic connections of the amygdala, striatum, and brain stem, which undermines emotional regulation. When there is effective top-down control, the PFC can sustain important work-related tasks by inhibiting the stress response and maintaining an optimal neurochemical environment. However, when the PFC goes offline, as suspected in burnout, the ensuing PFC dysfunction results in poor top-down control, and an increased probability for medical error and/or unprofessional (i.e., more disinhibited) behavior (Table 1).
A comprehensive 2019 narrative review was unable to show consistent HPA axis findings - i.e., involving measures such as the cortisol awakening response, morning cortisol, diurnal cortisol variation, daytime/evening cortisol, 24-hour urinary free cortisol, and others - among individuals with burnout (for further information, refer to “Table 1 Summary of HPA axis findings in clinical and non-clinical burnout,” p. R151).20 Testing that assesses how individuals with burnout respond to acute stress might eventually yield more convincing and stable neurochemical differences as opposed to measuring “resting state hormonal levels.”20 Similarly, studies that evaluated burnout and immune function were unable to find reproducible and discernable patterns of clinically relevant immune system changes among individuals with burnout.20
A couple of studies are worth discussing, however, since the results suggest that some neurochemical findings could be predictive of burnout, and relate to clinically significant burnout symptoms. In terms of burnout prediction, a study assessed hair cortisol changes among 372 adult healthcare workers from Quebec, Canada.21 The participants completed questionnaires, such as the 2-item MBI, and other questionnaires that evaluated anxiety, depression, and PTSD symptoms. The participants were also sent validated instructions to self-collect hair samples that were used to measure cortisol. From the 6 cm samples of hair that was collected, it was possible to calculate the relative changes in cortisol 3 months before the COVID-19 pandemic and 3 months after the COVID-19 pandemic started. From a sample of 367 healthcare workers, 50.4% had symptoms of burnout, as defined by emotional exhaustion or depersonalization. The hair cortisol increased at the start of the pandemic (i.e., with a median relative change of 29%; p<0.0001), with a 2.6 times more odds of burnout (p=0.002), and resulting in 59.6% of healthcare workers having burnout at that time. There were no associations between changes in hair cortisol and symptoms of PTSD, anxiety, and depression. The results of this study showed that changes in hair cortisol was predictive of burnout at 3 months among healthcare workers after the onset of the COVID-19 pandemic. With respect to clinical applicability, the results support the use of hair cortisol as a possible screening tool to identify clinicians at high-risk for burnout when confronted with a significant stressor (or set of stressors).
Here is an important study that evaluated the relationship between brain-derived neurotrophic factor (BDNF) and burnout.22 BDNF is an essential neurotrophic factor found in the human brain and participates in a myriad of functions, such as neuronal growth and proliferation, synaptic neurotransmission, and neuroplasticity.23 The study compared 37 participants with burnout and 35 healthy controls.22 Many different samples of various analytes (i.e., serum cortisol, serum BDNF, and others) were taken to assess HPA axis function between the different groups. Statistically significant differences were found between the means level of serum BDNF in both groups (p=0.005), such that the burnout group had a lower mean serum BDNF level (88.66±18.15 pg/ml) than the healthy controls (102.18±20.92 pg/ml). The results showed associations between burnout and emotional exhaustion (p=0.05), depersonalization (p=0.005), and depression p=0.025). In fact, depression (odd’s ratio: 0.722; p<0.001) was the most salient factor in distinguishing the burnout participants from the healthy controls. Serum BDNF levels correlated negatively with emotional exhaustion (r=-.268; p=0.026) and depersonalization (r=-.333; p=0.005), and correlated positively with personal accomplishment (r=.293; p=0.015). Based on these results, it was proposed that stress may downregulate the production of BDNF within the hippocampus and reduce the amount produced, which would increase the vulnerability for neuronal damage to take place along with increased clinical symptoms of burnout. In other words, the “low BDNF levels in burnout might be related to the concentration-memory problems and mood symptoms often observed in burnout syndrome.”22
Could hair cortisol measurements and serum BDNF levels be used to assess and/or predict burnout? I believe there is some utility in measuring these neurochemical biomarkers when managing patients at risk, or patients suspected of having burnout. If, for example, a patient presents with significant occupational stress, but they appear to be coping sufficiently, measuring hair cortisol and BDNF could serve as helpful baseline metrics, which then could be remeasured over time as treatment is instituted to encourage allostasis, and to safeguard against burnout. Alternatively, if a patient presents with burnout, capturing measurements of hair cortisol and serum BDNF could be helpful in determining how biologically impacted the patient is. In this situation, the hair cortisol would likely be increased and the serum BDNF likely decreased, which would be modifiable and potentially reversible with effective treatment.
Is Burnout a Specific Form of Depression?
In an editorial by Meier, the similarities between burnout and depression were described.24 He noted that measures of burnout correlate strongly with measures of depression. An earlier cited study in this paper showed associations between burnout and emotional exhaustion, depersonalization, and depression.22 Given the fact that there is strong overlap between burnout and depression, some researchers have put forth the argument that burnout should be redefined as occupational depression.25 One of the main reasons for this consideration comes from data on a construct known as negative affect (NA). NA “refers to emotions experienced as unpleasant or aversive.”24 Thus, “burnout is NA situated in the workplace, experienced as a result of chronic stress.”24 In a review of studies spanning several decades, correlations between NA and burnout/depression yielded an overall effect size of 0.492, and among people that were older and had worked for longer periods of time, the effect size increased to 0.535.26
There are also neurobiological and neurochemical similarities between major depressive disorder (MDD) and burnout. In MDD, chronic stress leads to impairment in PFC function, an overactivated amygdala resulting in more fear-based or bottom-up control, and a concomitant downgrading of hippocampal functioning.27 Some of the common features of MDD, such as neurocognitive impairment, withdrawing from aversive environments, and anhedonia are linked to these brain circuit issues.27Likewise, in burnout there is PFC dysfunction, neurocognitive impairment, and strengthened fear-based synaptic connections.19 In MDD, chronic stress results in low levels of serum BDNF due to hippocampal alterations.23 In burnout, chronic stress also leads to low levels of serum BDNF presumed to result from hippocampal alterations.22 Though burnout will continue to be studied as a separate psychological syndrome from depression, it seems very likely and probable that it could be understood as a unique form of depression that happens or is situated within the work environment (Table 2).
The Drivers and Consequences of Burnout
The symptoms of burnout mentioned earlier - i.e., emotional exhaustion, depersonalization, and diminished personal accomplishment - result mostly from the clinician’s chronically stressful work environment, and less from the personal characteristics of the individual clinician.28 In the subsections below, the more salient drivers and consequences of burnout are described, and have also been summarized in Table 3.