Physician burnout is an increasingly common condition with far-reaching consequences. Burnout is a syndrome related to chronic workplace stressors, and it is characterized by emotional exhaustion, depersonalization, and a low sense of personal accomplishment. In 1981, Christina Maslach, a psychologist based at the University of California Berkeley, created and standardized the Maslach Burnout Inventory (MBI) to quantify the syndrome. A four-item questionnaire, it assesses the severity and frequency of these three symptoms as well as an optional fourth factor assessing involvement.¹

Emotional exhaustion (EE) refers to the feeling of being drained or fatigued and loss of enthusiasm for one’s work.² EE is the primary quality and the clearest symptom of burnout,³ and a necessary, but not sufficient, criterion for burnout diagnosis.³ The second symptom is depersonalization (DP), which is characterized by cynicism and lack of interest in and empathy for patients. The third is a low sense of personal accomplishment (PA) that manifests as the feeling of ineffectiveness and the sense that one’s work loses its meaningfulness.³

Burnout is the primary reason for diminishing physician wellness. Wellness is a state of being that includes the interdependent aspects of absence of physical illness, good mental health, and positive social relationships. Wellness allows for prosperity in both professional and personal domains as well as a sense of professional fulfillment. The latter is key to physician wellness and consists of the basic elements of job satisfaction, self-esteem, feeling effective at work, and overall happiness.^4–6

Of the many professional roles that physicians assume, patient care and the easing of misery appear to be the most fundamental and satisfying one. Other rewarding aspects of being a physician include scholarly activities (research, teaching, writing, etc) and individual interactions with coworkers and patients.^2,7 A physician, therefore, can feel fulfilled through a number of avenues including educational pursuits, productivity, and innovative personal advancement.²

In this chapter, we will discuss the prevalence of burnout, its consequences and potential causes, and methods to prevent it and enhance physician wellness.

Burnout affects physicians at all levels of their education and career. In a large cohort study (n = 4287), 49.6% [95% confidence interval (CI), 47.5%–51.8%] of students from seven medical schools reported experiencing burnout. Suicidal ideation (SI) within the previous year was reported by 11.2% (CI, 9.9%–12.6%) of the surveyed cohort.⁸ With intervention, burnout improved in 26% of responders. Improvement in burnout was associated with significant reduction in SI.⁸

In a survey of 16,154 internal medicine residents, burnout was reported in 8343 (51.5%). Quality of life (QOL) was rated poor in 2402 (14.8%). Overall burnout was higher among US medical graduated than international ones [58.7% vs. 45.1%; odds ratio (OR), 0.70 (99% CI, 0.63–0.77); P < .001]. Amount of student debt was correlated with higher rates of burnout [61.5% vs. 43.7%; OR, 1.72 (99% CI, 1.49–1.99); P < .001 for debt >$200,000 relative to no debt].⁹

In a survey of 7905 practicing surgeons in the United States, 40% of responding surgeons reported being burned out, 30% screened positive for symptoms of depression, and 28% had a mental QOL score of over half a standard deviation below the population norm. Only 36% of surgeons felt that their work left enough time for personal and family life, and only 51% stated that they recommend their children pursue a career in medicine. Younger age, having children, area of specialization (greater satisfaction among most subspecialists), number of call nights per week, work hours per week, and having compensation determined entirely based on billing were independent factors associated with burnout.¹⁰

Another survey compared 7288 practicing physicians to a probability-based sample of 3442 working US adults. Burnout was more common among physicians (37.9% vs. 27.8%) (P < .001), with 45.8% of physicians reporting at least one symptom on the MBI. The highest rates of burnout among physicians were among those on the frontlines of healthcare access (e.g. emergency medicine, family medicine, and general internal medicine).¹¹

Using the MBI, Shanafelt et al (2015) again surveyed 6880 physicians and other adults in nonhealth related work force in 2011 and 2014. The prevalence of reporting at least one symptom of burnout had increased from 45.5% to 54.4% (n = 3680) (P < .001) over three years. Commensurately, satisfaction with work/life balance declined in physicians between 2011 and 2014 (48.5% vs. 40.9%; P < .001). In contrast, there were only minimal changes in burnout or satisfaction with work/life balance between 2011 and 2014 in probability-based samples of working US adults. After multivariate analysis, physicians remained at an increased risk of burnout (OR, 1.97; 95% CI, 1.80–2.16; P < .001) and were less likely to be satisfied with work/life balance (OR, 0.68; 95% CI, 0.62–0.75; P < .001).¹² Although early career physicians had the lowest job satisfaction and biggest conflicts in balancing work and home life, middle-career physicians experienced higher rates of burnout due to more work hours and overnight calls and low satisfaction with their specialty choice.¹³

Burnout is not a uniquely American problem. Physician surveys from Switzerland, Canada, Greece, and Italy have reported a physician burnout rate ranging from 25% to 60%.⁵ As in the United States, long work hours, more overnight calls, and perceived lack of autonomy in decisions as well as financial concerns are the main correlates of burnout in other countries also.⁵

Box 16-1 lists the prevalences of burnout in various studies.

Burnout can have many consequences, including the following:

• 
  

  Medical errors: In a survey of 380 US medical residents, the three elements of burnout (as measured by the MBI) were all associated with increased odds of perceived medical errors. DP and EE were both associated with higher rates of perceived medical errors, with an OR of 1.09 (95% CI 1.05–1.12) p < 0.001 and OR 1.06 (95% CI 1.04–1.08) p < 0.001, respectively. A sense of PA was conversely correlated with perceived medical errors, with OR of 0.94 (95% CI 0.92–0.97) p < 0.001.¹⁴ Similarly, in a survey of 86 Korean internal medicine residents, EE and DP were associated with higher odds of perceived medical errors, with ORs of 1.07 (95% CI 1.02–1.13) p < 0.005 and 1.11 (95% CI, 1.02–1.21) p < 0.013, respectively.¹⁵ This association between burnout and medical errors is not confined to residents, but can affect practicing physicians as well. In a survey of 7905 surgeons in the United States, 700 (8.9%) reported that they may have made a major medical error in the previous three months.¹⁶ Only 15.1% of surgeons attributed the error to system-level issues, while the rest attributed it to personal factors.¹⁶ Reporting an error during the three months preceding the survey had a large, statistically significant adverse relationship with all three domains of burnout.¹⁶ ORs were 1.048 (95% CI 1.042–1.055) p < 0.0001, 1.109 (95% CI 1.096–1.122) p < 0.0001, and 0.965 (95% CI 0.955–0.975) p < 0.0001, respectively.¹⁶ Each 1 point increase in DP (scale range, 0–33) was associated with an 11% increase in the likelihood of reporting an error, while each 1 point increase in EE (scale range, 0–54) was associated with a 5% increase.¹⁶

  
  

  Another study among providers at multiple primary-care clinics showed that lower burnout was associated with a trend of reduced error rates, with OR 1.44 (95% CI 0.94–2.23) p = 0.09.¹⁷ Similar results were reported from Japan in a survey of 836 practicing physicians that significantly linked burnout to medical errors.¹⁸ Having committed medical errors can also lead to EE, which in turn can lead to more medical errors, thus creating a vicious cycle. In a survey of Finnish general practitioners, having committed a medical error in the previous three months was marginally predictive of EE, with OR 2.4 (95% CI 1.0–5.9) p = 0.057.¹⁹ Finally, a meta-analysis of 15 studies reported an OR range of medical errors with burnout ranging from 1.07 to 5.5.²⁰

  
  
• 
  

  Depression, anxiety, and suicidality: Depression and burnout often coexist and can compound each other’s consequences. A survey of 6351 Austrian physicians from 2010 to 2011, using the Major Depression Inventory (MDI) and Hamburg Burnout Inventory (HBI), reported the following findings.²¹ Burnout was present in 50.7% of participants, and major depression in 10.3%.²¹ The OR for major depression among physicians with burnout ranged from 2.99 (95% CI 2.21–4.06) to 92.78 (95% CI 62.96–136.74), depending on the severity of the burnout.²¹ A US study of 125 OB/GYN residents across 23 residency programs from around the country reported that 89.8% were moderately burned out, and 34.2% were depressed.²² The absence of any of the three elements of the MBI was significantly correlated with lower risk of depression.²²

  
  

  Another survey of 123 pediatric residents in three US programs reported a depression rate of 20% and burnout of 74%.²³ Those who were depressed made 6.2 times more medication errors per resident per month than those who were not depressed.²³ In this particular study, burnout independent of depression did not correlate with more medical errors.²³ The meta-analysis mentioned here documented an OR of medical error with depression, ranging from 2.21 to 3.29.²⁰ A Finnish study of 2555 dentists not only showed higher risk of depression with burnout but also higher risk of burnout among those who are depressed at baseline.²⁴ Of those who had symptoms of depression without burnout at baseline, 63% had burnout 3 years later.²⁴ The adjusted OR of depressive symptoms for burnout was 2.2 (95% CI 1.4–3.4), while that of burnout for depressive symptoms was 2.6 (95% CI 2.0–3.5).²⁴ Out of those with burnout without depression at baseline, 23% had depression 3 years later.²⁴

  
  

  Anxiety was also found to be associated with burnout among hospital physicians in a Taiwanese study. The overall OR of anxiety with burnout was 1.103 (95% CI, 1.020–1.193), with men being at slightly higher risk than women. Anxiety also correlated well with the different elements of burnout in medical residents using the Burnout Symptom Inventory with r (the correlation coefficient) ranging from 0.454 (p < 0.001) for EE to 0.411 (p < 0.001) for dehumanization, and from 0.359 (p < 0.001) for emotional detachment to -0.270 for PA (p < 0.001).²⁵ SI is also independently associated with burnout. In a group of 436 gynecologic oncologists, the OR of SI in the setting of burnout was 4.92 (95% CI 2.59 Burnout Symptom Inventory 9.34) p < .001.²⁶ Similarly, in a group of 2115 Dutch residents, SI was reported in 20.5% of those with burnout vs. 7.6% of those without it; 9χ2 182.9, p < .001.²⁷

  
  
• 
  

  Substance abuse: Abuse of alcohol and other psychoactive substances is also commonly associated with burnout. In a survey of 4402 US medical students, 32.4% met the diagnostic criteria for alcohol abuse/dependence.²⁸ Students who were burned out (p = 0.01), depressed (p = 0.01), or reported low QOL in the mental (P = .03) or emotional (P = .016) domains were significantly more likely to meet the criteria for alcohol abuse/dependence. Of the three MBI domains, EE and DP were strongly associated with alcohol abuse/dependence.²⁸ With multivariate analysis, burnout (OR 1.20; 95% CI 1.05–1.37; P < .01), having an education debt between $50,000 and $100,000 (OR 1.21 vs. <$50,000; CI 1.02–1.44; P < .05) or >$100,000 (OR 1.27 vs. <$50,000; CI 1.08–1.48; P < .01), being unmarried (OR 1.89; CI 1.57–2.27; P < .001), and being younger (for every 5 years, OR 1.15; CI 1.02–1.28; P = .01) were independently associated with increased risk for alcohol abuse/dependence.²⁸

  
  

  A nationwide survey of practicing US physicians showed that 12.9% of male physicians and 21.4% of female physicians met the diagnostic criteria for alcohol abuse or dependence. Alcohol abuse and dependence were significantly associated with burnout and depression as well as lower QOL (all p < 0.001). Other associations included SI (p = .0004), less career satisfaction (p = .0036), and recent medical errors (p = .0011).²⁹ Other independent risk factors for alcohol abuse and dependence included age (OR = .985; p < .0001), work hours (OR = .994; p = .0094), male gender (OR = .597; p < .0001), being partnered (OR 1.989; p = .0003), and having children (OR .745; p = .0049).²⁹ A survey of 118 residents at a Lebanese academic medical center revealed that 27% of respondents met the criteria for burnout. Burnout was significantly associated with depression, SI, and drug abuse. Alcohol abuse increased with illicit substance use but did not independently correlate with burnout.³⁰

  
  
• 
  

  Professionalism: In a survey conducted among 2682 US medical students, burnout was independently associated with unprofessional behavior, such as cheating/dishonest clinical behavior [OR 1.76 (95% CI 1.45–2.13) P < .001], and having less altruistic opinions about physicians’ responsibilities [OR 1.65 (95% CI 1.35–2.01) P < .001].³¹

  
  
• 
  

  Accidents: In a survey of 301 residents conducted at the Mayo Clinic’s Internal Medicine program, 23 (7.6%) reported accidental exposure to body fluids and 168 (56%) reported a motor vehicle accident. Each 1-point decrease in the PA domain of MBI was associated with an 8% increase in the odds of accidental bodily fluid exposure in the subsequent 3 months. The OR for an actual motor vehicle crash or any motor vehicle incident associated with excessive daytime sleepiness (EDS) were 3.05 and 2.51, respectively. Each 1-point increase in the EE or DP domain of MBI was associated with a 3% and 4% increase, respectively, in the odds of reporting any motor vehicle incident.³²

  
  
• 
  

  Sleep complaints: Two nationally representative samples of Hungarian female physicians were surveyed in 2003 and in 2013. Burnout, and particularly lower PA, increased with the workload and were significantly associated with worsening insomnia, fatigue and decreases in the quality and quantity of sleep over the 10-year period, as measured by the Athens Insomnia Scale. This was independent of the prevalence of depression, which was significantly higher at both time points than that of the control group of nonmedical professional women.³³ In Taiwan, a large nationwide survey of 15,150 hospital physicians, matched with 45,450 nonmedical professionals, revealed that burnout increased the odds of insomnia, with OR of 2.028 (95%CI, 1.892–2.175).³⁴ Shift work (SW; as in the emergency room/ER), male sex, and younger age were risk factors for burnout-associated insomnia.³⁴ EDS, as measured by the Epworth Sleepiness Score, was associated with work hours, not hours of daily sleep in a study surveying physicians of a single academic center in 2009.³⁵

  
  
• 
  

  Shift work disorder (SWD): Since SW can lead to significant sleep problems, and this in turn can contribute to burnout, it is important to know what SW disorder is and how to address the sleep issues associated with it. The most consistent definition of SW has been working, at least in part, between 7 p.m. and 6 a.m.³⁶ In the industrialized world, nearly 20% of employed adults work in SW,³⁷ with healthcare workers among the professions with higher prevalence of it.³⁷ When the SW schedule interferes with the endogenous sleep-wake rhythms, the resulting sleep disorder is known as shift work disorder (SWD). Symptoms of SWD are falling asleep at work, inability to sleep during the daytime when they are not working, and struggle with both insomnia and EDS on days off.³⁸

Complications of SWD include increased prevalence of depression and substance abuse, relationship conflicts, decreased productivity, increased accidents, obesity, cardiometabolic problems, and increased risk of certain cancers.³⁸

The goal of SWD treatment is realignment of the sleep-wake and work schedules to the endogenous circadian rhythms. Although not always possible because of work/life balance and other family obligations, the following strategies have been shown to help:

1. 
   

   Bright light (3000–5000 lux) exposure during the first half of the night shift immediately improves alertness and speeds up the alignment over time.

   
   
2. 
   

   Avoiding bright light (using sunglasses) in the morning is also recommended, as this may improve the person’s ability to sleep once at home, but this also may lead to drowsy driving.

   
   
3. 
   

   A dose of 0.5 to 1 mg of melatonin taken before daytime sleep may help realignment and improve sleep quality.

   
   
4. 
   

   Maintaining a sleep-wake schedule on days off that is between the traditional sleep schedule and the workday sleep schedule may also be helpful. This compromise position, however, has only been tested in SW simulations and not in real-life conditions.

   
   
5. 
   

   Wake-promoting agents like modafinil and armodafinil have been shown to help improve alertness modestly during the night shift when administered at the start of it, and hence are indicated by the US Food and Drug Administration (FDA) for SWD. They do not, however, lead to improved work performance.³⁸

Box 16-2 lists possible strategies to decrease burnout and increase physician wellness.