Keywords

FormalPara Case Illustration

Jane, a 28-year-old second-year internal medicine resident, hits her alarm clock at 5 AM to go to the hospital. When it goes off, she lets out a sigh and feels very little motivation to get out of bed. Yet, like she has done for almost 2 years, she gets out of bed and drives to work. While at work, she rounds on her patients, snapping at one who asks her for a glass of water, and dismissing a family member who tries to ask her a question. A peer who has noticed that Jane has been a lot more negative lately and has not been acting like herself asks Jane what is wrong. Jane replies, “I just hate everything about being here and I have another full year to go.”

Part I: What Is Burnout?

In 1974, a psychologist named Herbert Freudenberger coined a term to describe job dissatisfaction related to work stress. He named it “burnout” (Freudenberger 1974). Since that time, our understanding of burnout as a syndrome has evolved. Broadly speaking, the World Health Organization defines it as a “state of vital exhaustion” (Fralick and Flegel 2014), and it is commonly understood to be a state of mental and physical exhaustion related to work (Ishak et al. 2009). More specifically, burnout is often broken down into three domains as described by Maslach and Jackson (1981): emotional exhaustion, depersonalization, and reduced personal accomplishment (Maslach and Jackson 1981).

To better understand the meaning of the domains of burnout, Jane’s case will serve as an example. Jane has very little enthusiasm to get out of bed and go to work and describes “hating” her job and feeling trapped (“I have another full year to go”). These are symptoms of emotional exhaustion , often caused by prolonged stress, which can lead to feelings of helplessness, defeat, and loss of enthusiasm at work. Jane is also indifferent toward her patients and negative toward her colleagues and her profession. She is therefore also exhibiting signs of depersonalization . The final domain, reduced personal accomplishment , is represented by detachment from one’s job and withdrawal from responsibilities. In Jane’s case, this happens when she does not want to get out of bed. It also shows itself more subtly in her interactions and enjoyment with day-to-day work (Maslach and Jackson 1981; Schaufeli et al. 1996). Therefore, by definition, Jane is burned out, a constellation of symptoms that could be quantified further using the scale developed by Maslach and Jackson (1981). While burnout or burnout syndrome gives a name to what Jane is suffering from, burnout is not yet a diagnosis and is not included in diagnostic classification systems like the International Statistical Classification of Diseases and Related Health Problems (ICD)-10 or the Diagnostic and Statistical Manual of Mental Disorders (DSM)-5.

Part II: Epidemiology of Burnout

Like Jane , approximately one-half of US physicians will experience burnout throughout their careers, a number that has been increasing from 2011 to 2014 (Shanafelt et al. 2015b). It is present in all stages of training, affecting 28–45% of medical students and 27–75% of residents (Ishak et al. 2009), and varies across specialties. In fact, the highest rates are seen in physicians that are considered to be on the “front lines” of care, or those who practice family medicine , internal medicine, and emergency medicine (Shanafelt et al. 2012). It is clear that burnout is a significant issue in physicians and physicians-in-training.

To better understand the uniqueness of burnout to physician-hood, studies have compared physician populations with matched population controls. The data suggest that medical students, residents and fellows, and early-career physicians are all statistically more likely to be burned out than their matched samples (Dyrbye et al. 2014). Physicians were also more likely than working adults in the US to be dissatisfied with their work-life balance (Shanafelt et al. 2012). In fact, being a physician or physician-in-training is a risk factor for burnout. While those with higher levels of education and professional degrees were found to have lower rates of burnout compared to high school graduates, this was not the case for MD or DO degrees. Instead, simply having a degree in medicine increases burnout (Shanafelt et al. 2012) suggesting that an MD or DO degree is not protective like other professional degrees but instead is a risk.

Burnout also varies across physicians’ careers , starting even before medical school. One study noted higher rates of burnout in premedical students than in other college students (Fang et al. 2012), yet in a different study, matriculating medical students had lower burnout rates than age-similar college graduates (Brazeau et al. 2014). It is uncertain from this data whether higher risk students choose to enter medical school in the first place or if medical school training itself leads to burnout. What is certain, however, is the high frequency of burnout in this group; in one study of 4287 medical students, 49.6% reported burnout (Dyrbye et al. 2008). Rates of burnout seem to peak after medical school and in residency. The data suggest that overall burnout and higher feelings of depersonalization and fatigue were the most prevalent in residents and fellows, as compared to early-career (EC) physicians and medical students (Dyrbye et al. 2014). Rates also vary largely by specialty and year of training, with highest rates in OB-GYN residents and in first-year residents in one study (Martini et al. 2004). When controlling for age, sex, relationship status, and career stage, being a resident or fellow was associated with increased odds of burnout (Dyrbye et al. 2014).

Looking beyond training and into physician-hood, burnout continues to vary over the course of one’s career, peaking in the middle. Studies have also shown a variation by gender, with one study noting higher rates of burnout in women (38.5% vs. 28%) (Trockel et al. 2016). While early-career physicians had the highest rates of depersonalization, physicians had the most emotional exhaustion and the most burnout in the middle of their careers. This was seen across genders, specialties, and practice types (Dyrbye et al. 2013). In fact, physicians in the middle of their careers were most likely to consider leaving medicine in the next 24 months for reasons other than retirement (12.5%), when compared to early- (4.8%) and late- (5.2%) career physicians, respectively (Dyrbye et al. 2013). Ultimately, burnout is a significant factor in all stages of a physician’s career and is becoming a much more prevalent and significant problem over time.

Part III: Causes of Burnout

With a clear understanding of the breadth of the problem, it is important to consider the causes or drivers of burnout in physicians. Some typical causes include high workload, lack of efficiency, lack of autonomy, challenges with work-life integration, and lack of meaning in work (Shanafelt et al. 2016c). These drivers can then be influenced by the individual physician, the team at work, the organization (the hospital, the university), and national factors (like insurance or reimbursement) (Shanafelt et al. 2016c). Given the high rates (over 50% of physicians), burnout is much more likely to be rooted in global causes like the work environment and the care delivery system, rather than personal factors or personal characteristics (Shanafelt et al. 2012), though both can contribute.

Analyzing the physician’s day-to-day environment , one can identify many causes for increased stress. Studies have found that long hours (Martin 2002), job demands (Alarcon 2011), heavy work load (Shanafelt et al. 2003), sleep impairment (Trockel et al. 2016), and inefficiency due to excessive administrative work (Shanafelt et al. 2003) increase rates of burnout. In fact, physicians who used the electronic health record (EHR) were often less satisfied with the amount of time they spent on administrative work and were subsequently more likely to be burned out (Shanafelt et al. 2016a). Significant workload, with long hours and frequent clerical work, can also contribute to a loss of a sense of meaning from work, which has also been associated with higher rates of burnout (Shanafelt et al. 2003). In one study, the time a physician spent on their “most meaningful activity” was the largest predictor of burnout, or, more specifically, those who spent less than 20% of their time working on the activity they found the most meaningful had significantly higher rates of burnout (Shanafelt et al. 2009b). As most physicians reported that they derived the most meaning from patient care (68%) (Shanafelt et al. 2009b) and EHRs can often distract from clinical tasks (33% of work clinical and 49% clerical) (Sinsky et al. 2016), it is perhaps not surprising that EHRs may themselves be contributing to the rise in physician burnout (see Chap. 6).

Loss of autonomy often created by the constraints of the team and institution is also associated with high rates of burnout (Shanafelt et al. 2003). Studies have shown that constraining organizational structure, low quality and safety standards (Lee et al. 2013) and low organizational commitment (Alarcon 2011) lead to higher emotional exhaustion.

For example, rules around the scheduling of vacation , sick leave, call, and clinic start and end times can limit flexibility, and the rigid application of practice guidelines can limit clinical decision-making (Shanafelt et al. 2016c). Additionally, compensation determined entirely on billing (i.e., number of direct clinical hours) (Shanafelt et al. 2009a) and insurance regulations around note writing, prescriptions, and referrals all contribute to loss of autonomy by physicians (Shanafelt et al. 2016c). Given the effect of the organization on one’s autonomy, it is perhaps not surprising that the reverse would also be true; a culture of wellness— which included things like perceived appreciation, personal alignment with the organization, and peer supportiveness—could increase professional fulfillment and thus help prevent burnout (Trockel et al. 2016) (see Chap. 2).

An additional cause of burnout is difficulty integrating one’s personal and professional lives (Shanafelt et al. 2003), or work-life balance. This was found to be a much more significant contributor to burnout in the US than in Europe (Lee et al. 2013). Work-life balance can be affected on the individual level by one’s spouse or children, at the level of the team due to coverage and leave, and at the level of the organization-given rules around part-time work and licensing requirements (Shanafelt et al. 2016c). In a population of surgeons, having children and working more than 60 hours a week was associated with higher levels of burnout (Shanafelt et al. 2009a). Additionally, family stress and being unmarried were associated with higher burnout in residents (Martini et al. 2004), and a positive life event, like a marriage, was associated with lower burnout in medical students (Dyrbye et al. 2009). It is important to note that causes of burnout may differ in different stages of training and career. Medical students might have fewer opportunities for meaning in their roles and thus might have a low sense of personal efficacy that will lead to burnout. Their learning climate, including being on the wards or a rotation with call, can also contribute to burnout (Dyrbye et al. 2009). Residents, however, may be burdened by new work responsibilities and the new day-to-day stress of life as a physician (Dyrbye et al. 2014). All years are affected by burnout; no stage of training or career is immune.

Part IV: Outcomes of Burnout

Why do we even care about burnout?

High rates of burnout can lead to significant consequences for physicians both personally and professionally. This, in turn, can lead to poor health outcomes for patients and systemic problems for a hospital system.

First, let’s examine how burnout affects the individual physician . Studies have shown that burnout is associated with worsened mental health (Asai et al. 2007), problematic alcohol use (Shanafelt et al. 2003; Oreskovich et al. 2012), tobacco use (Soler et al. 2008), psychotropic medication use (Soler et al. 2008), and stress-related health problems (Martini et al. 2004). Burnout is also associated with suicidal ideation (Shanafelt et al. 2011). In fact, in medical school, burnout has a dose-response relationship with suicidal ideation, i.e., the higher the burnout, the more likely one is to be suicidal (Dyrbye et al. 2008). Yet, as burnout has also been associated with higher stigma and lower likelihood of seeking mental health care (Dyrbye et al. 2015), burnout becomes a serious threat to physicians’ livelihood, mental health, and life.

Burnout can lead to low job satisfaction , which can have significant effects on the medical workforce. Studies have shown that higher stress can increase physician’s intention to withdraw from practice (Williams et al. 2010), and increased burnout is related to early retirement (Leenders and Henkens 2010). Higher burnout is also associated with intention to leave a current position (Dewa et al. 2014b), intention to quit a specialty (Dewa et al. 2014b), and intention to leave academic medicine (Shanafelt et al. 2009b). One study showed that physicians with lower job satisfaction and higher levels of emotional exhaustion were more likely to reduce their full-time equivalent (FTE) over the following 24 months in a dose-response relationship (Shanafelt et al. 2016c). More specifically, when controlling for age, sex, and specialty, for every 1-point increase in emotional exhaustion and 1-point decrease in satisfaction, there was a 43% and 34% higher likelihood of reducing FTE, respectively (Shanafelt et al. 2016c). In another hospital, 21% of physicians who reported burnout in 2013 no longer worked at that hospital by 2015. The number of physicians with burnout who had left the hospital was more than two times the number of those without burnout who had left the hospital (Trockel et al. 2016). In medical students, one study showed that those with burnout were three times more likely to have serious thoughts of dropping out of medical school (Dyrbye et al. 2010c), with burnout always preceding thoughts of dropping out (Dyrbye et al. 2010c). Interestingly, when students “recovered” from burnout, their thoughts of wanting to drop out returned to baseline frequency (Dyrbye et al. 2010c).

As all domains of burnout have also been associated with more use of sick leave (Soler et al. 2008), it is clear burnout is a significant cost to productivity (Dewa et al. 2014b). Decrease in staff and productivity also carries a high price tag. One study found the cost of burnout-related departure of staff alone to be between US $22 and US $88 million (Trockel et al. 2016). Another study estimated the cost of burnout for the Canadian health-care system by surveying all physicians in Canada (more than 60,000 participated) and assessing the net difference between retirement and cutbacks in services between those physicians who were very dissatisfied and those who were not. Their results, which they emphasized were likely a conservative measurement, found the cost to be about 213 million Canadian dollars ($185.2 million due to early retirement and $27.9 million due to reduced clinical hours) or approximately US $159 million (Dewa et al. 2014a). Ultimately, burnout significantly affects a hospital system.

Besides decreasing the workforce, burnout can also affect patient care. Studies have shown that increased burnout is associated with decreased quality of care (Klein et al. 2010), decreased amount of time devoted to patient care (Shanafelt et al. 2016c), and self-reported suboptimal patient care in residents (Shanafelt et al. 2002). Burnout is also associated with more medical errors, with higher self-report of errors in residents who are burned out (West et al. 2006). In medical students, burnout was associated with self-reported unprofessional conduct and less altruistic professional goals and values (Dyrbye et al. 2010a). Perhaps because of the difference in care provided or difference in empathy from the provider, job satisfaction and burnout can also affect patient adherence to recommended therapy (Dimatteo et al. 1993) and the degree of trust and confidence that patients have in their physician (Haas et al. 2000). Having higher satisfaction with one’s own job can lead to higher satisfaction ratings by patients (Haas et al. 2000). Those who are burned out are not at all satisfied.

Part V: Prevention/Intervention Strategies

As burnout can be caused by the individual, the team, the organization, or the system, it is perhaps not surprising that strategies at both individual and structural or organizational levels are warranted to reduce burnout (West et al. 2016). A systematic review indicated that both individual- and organizational-level strategies produced meaningful reductions in burnout (West et al. 2016). We will briefly examine different interventions to both prevent and decrease burnout in physicians (Shanafelt et al. 2017).

Individual-Level Interventions

Several interventions at the individual level have been published in the literature. One such group of interventions focuses on enabling physicians to recognize their own needs and ask for help. Evidence suggests that providers who seek help or use coping strategies have lower emotional exhaustion (Ito and Brotheridge 2003). This might be easier said than done, however, given high levels of stigma in burned-out physicians. Only one-third of burned-out medical students seek help (Dyrbye et al. 2015). As motivation to get help is limited, it might be necessary instead to provide physicians with the knowledge and skills to better help themselves. One might hypothesize that simply educating medical students on self-care and burnout would be an effective intervention to change health habits; however, one such intervention study yielded no effects on depression, alcohol use, or stress (Ball and Bax 2002). Instead, it seems most effective to teach skills to physicians.

There are many different types of skills that have helped decrease burnout in physicians. Teaching mindfulness and enhancing self-awareness are the most studied. For medical students, a ten-session mindfulness meditation course improved mood as compared to controls (Rosenzweig et al. 2003). Another study found that an all-day stress management workshop for residents decreased emotional exhaustion, with effects still present 6 weeks after the intervention (McCue and Sachs 1991). For physicians at other levels of training or practice, mindfulness has also been proven effective. In one study, physicians were trained in four key mindfulness practices and were subsequently found to have significantly decreased burnout and enhanced mental well-being (Goodman and Schorline 2012). Another study included a 19-session, biweekly discussion group consisting of mindfulness and self-reflection for physicians. Following these sessions, physicians found work more meaningful and experienced reduced depersonalization; effects were sustained 12 months later (West et al. 2014). An additional program involved an intensive, yearlong (52-hour) curriculum on narrative medicine and mindfulness meditation. Participating physicians had large and sustained improvements in burnout, mood, and empathy (Krasner et al. 2009). Exercise was also trialed as an intervention. Twelve physicians were found to have reduced emotional exhaustion after completing 12 one-hour aerobic sessions (Gerber et al. 2013). With minimal time investment in learning new skills, individual physicians experienced significant gain and improvement in burnout. For more information on mindfulness training, see Chap. 3.

In addition to skills building, reflection groups also seem to be effective interventions for physicians. In one study, Balint sessions were found to be preventative against stress and burnout (Benson and Magraith 2005). At the Mayo Clinic, providing physicians with 1 hour of protected time to meet over food and discuss physician-hood improved meaning and reduced burnout (West et al. 2015). At Stanford, the Balance in Life Program allows surgical residents to practice team dynamics and receive leadership training with a process group led by a clinical psychologist (Joseph 2017). Self-development groups have been effective at reducing burnout (Williams et al. 2015). Though more outcome measurements are needed to determine the best type of group to address burnout, groups are a promising intervention.

There are several other avenues at the individual level that might be effective in burnout reduction but have not yet been studied. Brief web-based cognitive behavioral interventions were found to be effective in reducing suicidal ideation (Guille et al. 2015); however, no such brief intervention has looked at the effect of skills learning on burnout. It is possible that in-person courses on mindfulness might be translated into online modules for clinicians with less time to participate (see Chap. 3). It is also possible that knowledge-based interventions that were ineffective previously might be more effective as online resources. In-person groups might translate to confidential online forums or virtual support groups where many rely on the support of others virtually. Facebook groups like the Physician Moms Group, for example, could perhaps function to reduce burnout through validation or a decrease in social isolation. Though unstudied, peer advocates (Robledo-Gil et al. 2018), who are most often sought out for mental health and relationship issues, might also help decrease burnout for trainees. Ultimately, however, at the individual level, promoting personal resilience is often preferred. While skills teaching and in-person groups have been proven effective at reducing burnout, there are many more avenues of promising interventions that warrant further investigation and interest.

Organizational-Level Interventions

While individual interventions are important, organizational interventions are key to burnout reduction. In one hospital, over 60% of physicians rated potential organizational strategies, including leadership development and collaborative practice improvement, as helpful or extremely helpful (Trockel et al. 2016). In fact, a combination of both personal and organizational interventions was found to have much longer-lasting effects (12 months or more) (Awa et al. 2010) than either alone.

Simply acknowledging that burnout is a problem, measuring burnout as a performance measure, and coming up with ways to intervene are key to overall reduction in burnout in physicians (Shanafelt and Noseworthy 2017). One primary care clinic prioritized wellbeing to the same extent as quality of care. They analyzed the factors affecting wellbeing and implemented plans for improvement. Following the implementation, physicians were found to have less emotional exhaustion (Dunn et al. 2007). At Mayo Clinic, different work units focused on specific issues that were leading to burnout and developed interventions to improve them. Each unit, no matter what the specific intervention, had a significant reduction in burnout (Swensen et al. 2016). Even at the level of the American Medical Association, the STEPS Forward Program has online modules aimed to educate physicians on how to streamline workflow, boost efficiency, and reduce administrative burden (Joseph 2017) with hopes of lowering burnout. The governing bodies of graduate medical education have also called attention to burnout by holding symposia on wellbeing and collecting educational resources for training programs to use to respond to resident suicide and encourage wellness (Joseph 2017).

The organization is also responsible for creating the learning and working environment. Studies have shown a strong dose-response relationship between the level of support (from family, friends, peers, and the medical school) and satisfaction with the environment. Higher satisfaction, in turn, was found to be protective against burnout (Dyrbye et al. 2010b). One factor that contributes to a supportive environment for trainees is the leadership and role modeling of faculty. An environment where student education was perceived as a priority for faculty was associated with recovery from burnout (Dyrbye et al. 2010b). Additionally, a focus on leadership development can lower burnout, as an increased score in leadership is related to decreased burnout (Shanafelt and Noseworthy 2017; Shanafelt et al. 2015a).

The working environment may also be improved upon by other organizational interventions. Duty hour restrictions created by the Accreditation Council for Graduate Medical Education (ACGME) have been effective in reducing burnout (Williams et al. 2015; Shanafelt et al. 2016b; Shanafelt et al. 2017). Residents benefit from access to a refrigerator full of energizing food options (see Chap. 14) and a home base area where they can relax and build a sense of community and belonging (Joseph 2017). Regulatory reform can help decrease clerical burden (Shanafelt et al. 2016a). Changing incentive and reward structure (i.e., not having salary simply based on patients and face-to-face time) has also been shown to reduce burnout (Shanafelt and Noseworthy 2017). Strategies like aligning values with culture and pass/fail grading (Shanafelt and Noseworthy 2017) have also been effective.

Hypothesized organizational-level solutions to improve burnout are wide reaching. Workspaces and offices should be optimized to see whether there is a design that may ease burnout (e.g., printers in rooms, e-card readers, windows). Organizations often provide vouchers for free rides home if residents feel unsafe to drive, and this, along with call room design, meal vouchers, and availability and utility of nap pods, should be studied. Additionally, though the electronic medical record contributes to physician burnout (Shanafelt et al. 2016a), more information is needed to formulate targeted potential interventions to ease physician workflow. Solutions like hiring scribes and administrative assistants for clerical work, such as prescriptions, should be further studied, particularly their cost-effectiveness considering burnout cost reduction. Given the importance of leadership and role modeling, it is important to look at the workplace culture as an avenue for intervention (see Chap. 2). Physicians should be encouraged to admit their own limitations, and adjustments should be made to value balance over hours worked. Implementation of culture change is perhaps one of the hardest interventions, but promoting faculty who prioritize prevention and mitigation of burnout in medical trainees might be a start (Dyrbye et al. 2010b).

The evidence suggests that burnout is a significant problem in physicians, increasing in frequency, and prevalent at all stages of training. It is associated with negative personal and professional outcomes (to the hospital and to the patient) and needs to be sufficiently and urgently addressed. The American Medical Association now has a fourth aim, in addition to lower cost, enhanced quality, and increased access: professional satisfaction (Joseph 2017). Prioritizing physician wellness and job satisfaction at the organizational level is a key first step to improving burnout. However, other interventions are needed at the individual, team, and organizational levels to effectively address, prevent, and decrease the burnout epidemic in our physicians.