Abstract
Objective: To study prevalence of and factors contributing to burnout, career satisfaction, and well-being in US neurology residents and fellows.
Methods: A total of 938 US American Academy of Neurology member neurology residents and fellows were surveyed using standardized measures of burnout, career satisfaction, and well-being from January 19 to March 21, 2016.
Results: Response rate was 37.7% (354/938); about 2/3 of responders were residents and 1/3 were fellows. Median age of participants was 32 years and 51.1% were female. Seventy-three percent of residents and 55% of fellows had at least one symptom of burnout, the difference largely related to higher scores for depersonalization among residents. For residents, greater satisfaction with work–life balance, meaning in work, and older age were associated with lower risk of burnout; for fellows, greater satisfaction with work–life balance and effective support staff were associated with lower risk of burnout. Trainees experiencing burnout were less likely to report career satisfaction. Career satisfaction was more likely among those reporting meaning in work and more likely for those working in the Midwest compared with the Northeast region.
Conclusions: Burnout is common in neurology residents and fellows. Lack of work–life balance and lack of meaning in work were associated with reduced career satisfaction and increased risk of burnout. These results should inform approaches to reduce burnout and promote career satisfaction and well-being in US neurology trainees.
GLOSSARY
- AAN=
- American Academy of Neurology;
- ACGME=
- Accreditation Council for Graduate Medical Education;
- MBI-HSS=
- Maslach Burnout Inventory–Human Services Survey;
- QOL=
- quality of life
Burnout in physicians is a threat to personal well-being, harms patient care, and may reduce the future physician workforce over time.1,–,5 Physicians experiencing burnout are more likely to have work–home conflicts and health problems, including depression and suicide.6,–,12 The 3 dimensions of physician burnout are emotional exhaustion, feelings of cynicism and detachment (depersonalization), and the sense of ineffectiveness at work (low personal accomplishment).13
Over 50% of physicians in the United States meet criteria for burnout and the rate of burnout is increasing.14 Practicing neurologists have one of the highest burnout rates and one of the lowest levels of satisfaction with work–life balance among all medical specialties.8,14,15
Among US physicians, the rate of burnout appears to be even higher in residents and fellows than among those in practice.16 Resident distress has been linked to occupational body fluid exposures, motor vehicle accidents, medical errors, and poor quality of life.17,–,20 The Accreditation Council for Graduate Medical Education (ACGME) and the American Medical Association are developing guidelines and wellness programs to help training programs recognize and mitigate burnout and trainee distress.
Understanding the threats burnout poses to access to and quality of care for patients with neurologic disorders,15,21,22 the American Academy of Neurology (AAN) formed a task force to study burnout, career satisfaction, and work–life balance; identify associated drivers; and develop resources to improve burnout and career satisfaction. We report the results of a national survey of neurology residents and fellows assessing burnout, career satisfaction, and well-being.
METHODS
Study population.
Details of the survey population and study methods were published previously.23 The aggregate study population (n = 9,561) of interest included neurologists (n = 7,738) and neurology trainees (1,823 residents and fellows) who were current members of the AAN and had a primary address in the United States.
Assuming a 20% response rate and targeting a margin of error no more than ±3.0%, a sample of 5,065 members was selected from the eligible population. The sample included 4,127 neurologists in practice and 938 trainees (residents and fellows). Results for practicing neurologists were previously reported.23 The present analysis focuses on the 938 neurology trainees.
Individuals in the sample were mailed a paper survey on January 19, 2016, sent an e-mail with a link to an online survey, and faxed (if available) a link to an online survey. Nonrespondents received up to 2 additional mailings and faxes and up to 5 additional e-mail reminders. All communications and data collection were conducted on behalf of the AAN by Anderson, Niebuhr & Associates, Inc., who provided the AAN with deidentified (anonymized) data from those who completed their survey by March 21, 2016. Participants were eligible for a drawing for 1 of 20 $500 Visa gift cards.
Standard protocol approvals, registrations, and patient consents.
Consent to participate in the study was implied by completing the survey. After the deidentified data were provided to the AAN, the study was reviewed and granted exempt status by the University of Pittsburgh Institutional Review Board.
Study measures.
The survey consisted of 57 questions covering personal and professional characteristics with standardized instruments to measure burnout and career satisfaction (available at Neurology.org).8,13,14
Burnout was measured using the 22-item Maslach Burnout Inventory–Human Services Survey (MBI-HSS). The MBI-HSS has 3 subscales to evaluate each domain of burnout: emotional exhaustion, depersonalization, and personal accomplishment.13 Using the standard scoring criteria for health care workers, and in keeping with previous studies and convention,8,14,24 we considered neurology trainees with high scores on the emotional exhaustion (≥27) or depersonalization (≥10) subscales as having at least one manifestation of professional burnout. Career satisfaction was assessed using 2 questions from previous physician surveys regarding career and specialty choice.8,14 Meaning in work and job satisfaction were assessed using questions from the Empowerment at Work14,25 and Physician Job Satisfaction26 scales. Two questions were asked about the amount of time spent on clerical tasks directly and indirectly related to patient care.27
Statistical analysis.
Standard descriptive statistics were used to characterize responding neurology trainees. With 354 trainee responses to the survey, the percentage estimates were accurate to 4.7% with 95% confidence. Associations between variables were evaluated using the χ2 test for categorical variables and the Kruskal-Wallis test for continuous variables. All tests were 2-sided with type I error rates (α) of 0.05. Multivariable analyses to identify demographic and professional characteristics associated with the dependent outcomes were performed using binary logistic regression. All analyses were performed using IBM (Chicago, IL) SPSS Statistics version 23.
Modeling for multivariable analysis was different for trainees and practicing neurologists to more accurately reflect differences in work settings for the 2 groups. Work–life balance and year of training were included only for trainees. Autonomy was excluded in the multivariable analysis for trainees because it was less likely to be a significant independent variable.
RESULTS
Response rate and representativeness.
Of the 938 neurology trainees surveyed, 354 responded (response rate 37.7%), with 23 respondents reporting a primary focus in child neurology. Responders did not differ significantly from nonrespondents (all p > 0.05) (table e-1) with respect to age, sex, region, or stage of training (resident vs fellow). Analysis of early responders (the 49.7% who responded within the first 29 days) vs late responders (the 50.3% who responded over the next 31 days) found no significant differences (all p > 0.05) with respect to burnout, sex, age, region, or type of trainee (resident vs fellow). Collectively, there was evidence that the sample was generally representative of US neurology trainees with respect to demographic characteristics and level of burnout.
Personal characteristics.
The median age of all participants was 32 years; 51.1% were female. Responders were well-represented across geographic regions. Roughly two-thirds of responders were residents and one-third were fellows (table 1). Residents and fellows were statistically similar in terms of sex and geographic distribution; however, fellows were older than residents (p < 0.001).
Characteristics of residents and fellows
Professional characteristics.
Responders reported working a mean of 64.6 hours per week, with approximately three-quarters (76.9%; mean 49.7 h/wk) (table 2) of the time spent on clinical care. Residents worked more hours per week on average than fellows (67.5 vs 59.1, p < 0.001), spent a larger proportion of their time in direct patient care (82.5% vs 67.3%, p < 0.001), and devoted less time to research (4.1% vs 19.3%, p < 0.001). Residents spent more nights per week on call than fellows (mean 1.39 vs 1.18, p = 0.003), cared for more inpatients on hospital days (median 10 vs 1, p < 0.001), and spent more weekends per year rounding in the hospital (median 23 vs 4, p < 0.001), but cared for fewer outpatients per week (median 6 vs 10, p = 0.006).
Professional characteristics of residents and fellows
Career satisfaction and burnout.
Table 3 describes burnout, career satisfaction, and well-being among participating trainees. Over one-half of trainees had high emotional exhaustion (181 of 332; 54.5%) and high depersonalization (178 of 332; 53.6%). A total of 31.9% (106 of 332) had a low personal accomplishment score. Overall, 67.2% (223 of 332) of trainees had at least one symptom of burnout (high emotional exhaustion or high depersonalization). Residents had a higher burnout rate than fellows (73.5% vs 55.0%, p = 0.001), due to higher scores in the depersonalization domain (p = 0.001). Personal accomplishment scores were worse (lower) for residents than for fellows (p < 0.001). Demographic and practice characteristics associated with burnout on bivariate analysis are shown in tables e-2 and e-3. Men and women had similar burnout rates.
Burnout and career satisfaction for residents and fellows
With respect to career satisfaction, 56.9% (189 of 332) of trainees reported they would choose to become a physician again while 75.9% (252 of 332) indicated they would choose to become a neurologist again. Career satisfaction scores in both these domains were similar between residents and fellows (p > 0.05).
The median overall quality of life (QOL) score of trainees was 6 on a 0–10 scale, while the median score on the fatigue scale was 5 out of 10. Scores less than 6 on the QOL scale are considered clinically meaningful impairments in QOL,28 and 37.7% of respondents reported scores less than 6. Fewer than 1 in 4 trainees (23.0%; 76 of 331) indicated their work schedule left enough time for personal/family life. There were no differences (p > 0.05) between residents and fellows on QOL and fatigue; however, residents scored lower on work–life balance (p = 0.025).
One third of trainees (33.5%) reported significant autonomy in determining how they do their job. Most trainees (81.0%) reported their work was meaningful, and 66.5% were satisfied with their job. There were no differences between residents and fellows for measures of autonomy, meaning, and job satisfaction (all p > 0.05).
A minority of trainees indicated the amount of time spent on clerical tasks, both directly (28.0%: 92 of 328) and indirectly (15.5%: 51 of 328) related to patient care, was reasonable. A majority of trainees (62.2%: 199 of 320) indicated that they had too little support staff to assist them with their work. There were no differences between residents and fellows on satisfaction with clerical tasks and sufficient amount of support staff (all p > 0.05).
Factors associated with burnout.
Multivariable analyses were conducted to identify factors associated with burnout (table 4). In addition to a model for all trainees, separate models were developed for residents and fellows because of substantial differences in personal and professional characteristics between these groups. On pooled analysis of all trainees, greater satisfaction with work–life balance (p = 0.005), meaning in work (p = 0.010), and having effective support staff (p = 0.017) were associated with lower burnout risk.
Factors associated with trainee burnout using multivariable analyses
Risk profiles differed by trainee status. For residents, greater satisfaction with work–life balance (p = 0.020) and meaning in work (p = 0.015) as well as older age (p = 0.009) were associated with lower burnout risk. For fellows, greater satisfaction with work–life balance (p = 0.035) and having effective support staff (p = 0.009) were associated with lower burnout risk.
Factors associated with career satisfaction.
Multivariable analyses were also conducted to identify factors associated with career satisfaction (table 5). Trainees with burnout were less likely to have career satisfaction (p = 0.013). Trainees who reported their work was meaningful were more likely to be satisfied with their career choice (p < 0.001). Trainees living in the Midwest were more likely to report career satisfaction than trainees in the Northeast (p = 0.008). Personal and professional characteristics were not significantly different (all p > 0.05) between the Midwest and Northeast regions (table e-4).
Factors associated with trainee career satisfaction using multivariable analyses
DISCUSSION
We assessed burnout, career satisfaction, and well-being among US neurology residents and fellows using standardized instruments.23 About 73% of residents and 55% of fellows had at least one symptom of burnout. There were too few child neurology trainee respondents to allow for separate analysis of this group.
The 73% burnout rate for neurology residents is similar to rates previously reported in internal medicine residents.16,23,29 In a 2001 survey using methodology similar to ours, 76% of internal medicine residents met criteria for burnout.16 When the survey was repeated in 2004, 1 year after implementation of ACGME work hour restrictions, 68% met criteria for burnout and 80% reported career satisfaction, up from 66%.29 Although the difference in burnout between the 2 studies (2001 and 2004) was reported as statistically significant, the persistent high burnout rate does not indicate a striking effect from work hour limitations. Likewise, in our study, high burnout among residents was evident years after the implementation of duty hour restrictions. For fellows, however, the significantly fewer work hours compared with residents may have contributed to their lower burnout rate. A study of 150 residents across 13 specialties identified pessimism as a major factor in burnout, among other factors, but poor support staff and too many work demands were not correlated with burnout.30 Lack of support was not specifically associated with burnout for residents in our study, but it was for fellows. Our survey did not explore specific aspects of support, such as outpatient vs inpatient differences, that might explain this finding.
A Dutch study of residents in different specialties found emotional demands at work and work–home interference were main contributors to burnout, similar to our findings.31 While not directly comparable with US training program characteristics, a study of neurology residents in Greece identified an 18% burnout rate and linked burnout to reduced professional development but not stage of training or work hours.32 Likewise, our study did not identify work hours as an independent factor in burnout.
We found that career satisfaction was higher in the Midwest region of the United States compared to the Northeast (p = 0.008). The professional and personal characteristics we measured were not different between these regions (table e-4), but our survey did not ask about possibly relevant factors such as size of metropolitan area, program size, competitiveness, or collegial milieu.
Residents experienced significantly more burnout, mainly attributable to increased depersonalization. Factors different for residents' work included more weekends and overall hours worked, more nights on call, more direct patient care, and more inpatients. The number and intensity of patient experiences for residents compared with fellows, especially when concentrated in the early years of training, may contribute to the disparity in burnout and work–life balance seen between residents and fellows in our study, and may contribute to increased depersonalization. Better work–life balance and meaning in work correlated with older age among residents, perhaps due to less prolonged exposure to intense patient care, fewer days worked, and more supervisory duties. However, age did not significantly correlate with other personal or professional characteristics measured in the survey.
Neurology trainee burnout rate was higher than the previously reported burnout rate among practicing neurologists.23 In that analysis, the burnout rate among neurologists was about 56% in academic practice and 63% in clinical practice. Hours worked, nights on call, number of outpatients seen, and amount of clerical work were independent factors associated with burnout for practicing neurologists but not for trainees. However, themes were similar for trainees and neurologists, including work–life balance, the relationship of both meaning in work and need for support with career satisfaction, and the relationship between career satisfaction and burnout.
In our study, neurology fellows had professional characteristics and a burnout rate similar to neurologists in academic practice.23 In a study of oncology fellows, their burnout rate of 43% was similar to that among practicing oncologists.33
Our results provide insights into factors associated with increased burnout in neurology trainees. As with practicing neurologists, there may be additional determinants not assessed in our survey. One could ask whether there are personality traits of neurology trainees or higher emotional or intellectual demands placed on trainees by patients with severe neurologic disease that predispose to increased burnout.15,21,22 In addition, whereas neurology residents have similar rates of burnout compared with trainees in other specialties, the burnout rate among practicing neurologists is higher than in many other specialties.16,19,23,29,34 Fewer trainees would choose a career in medicine again than would choose neurology again. This may reflect dissatisfaction with the practice of medicine in general rather than neurology in particular. Additional studies may provide a deeper understanding of these findings, including analysis of the free text comments many responders entered in our survey.
Our study had notable strengths. The survey response rate for trainees was almost 38% (212 residents and 111 fellows). These numbers are similar to other published studies16,–,20,29,30 on US trainees, and we obtained statistically significant findings with our respondent population. Neurology trainees and practicing neurologists completed the same survey, allowing direct comparisons between these groups.
Our study had limitations. Although we found no significant differences between early responders compared to late responders, we cannot exclude the possibility of response bias. Because the survey was cross-sectional, we were unable to determine causality or potential direction of effect for the associations observed. Modeling for multivariable analysis differed in some respects for neurology trainees and neurologists in practice.
Intervening to prevent and mitigate burnout among neurology trainees is critically needed to maintain the health and productivity of our workforce. The ACGME proposed that resident well-being be monitored and addressed in training programs. Short versions of the Maslach Burnout Inventory35 or the more easily administered Mayo Physician Well-Being Index36 may enable program directors to monitor trainees over time for signs of distress and intervene when needed.
Elements of burnout and depression can overlap in affected trainees, particularly emotional exhaustion. However, burnout and depression are not synonymous but rather are complexly interrelated.31,37 The combination of depression and burnout in a trainee is especially worrisome,16,31 possibly increasing the risk of suicide.
Many have blamed prolonged work hours for eroding trainees' learning milieu and quality of direct patient interactions.29 However, it is now clear that despite mandated and implemented work hour restrictions, the rate of trainee burnout is still too high. Further evidence that regulating work hours is not sufficient is provided by a study of European residents that reported burnout despite a 48-hour work restriction, considerably lower than the work hour restrictions for US residents.38
A recent meta-analysis investigating the effectiveness of interventions to reduce burnout suggests that a combination of strategies focused at the individual, structural, and organizational levels may be required to produce meaningful reductions of burnout rates.39 Programs devoted to improving personal resilience and stress reduction can promote physician well-being. Trainee burnout can also be lessened by increasing engagement and providing increased opportunities for job development and increased autonomy,32,40 although additional analysis found that autonomy was not a significant factor for predicting burnout in our results.
Finding meaning in work appears to be a major factor in both career satisfaction and burnout. Our findings that burnout correlates with low satisfaction with work–life balance and meaning in work suggest that organizational changes are necessary. Examples include redesigning neurology training curricula to focus on more meaningful internal medicine experiences in the preliminary year of training (for adult neurology residents), emphasizing professional development, teaching coping skills, and reducing clerical burdens. Alternating less intense with more intense clinical rotations in earlier years of training could decrease depersonalization. Training programs should consider providing additional support staff to enable trainees to focus on learning and practicing the intellectual and procedural aspects of neurology during their work hours, and not be distracted by less meaningful tasks. Longitudinal studies in the future may help determine if implemented changes improve career satisfaction and promote engagement during training and over the course of a career in neurology.
AUTHOR CONTRIBUTIONS
Kerry H. Levin: design of the study, interpretation of the data, drafting and revising the manuscript for intellectual content. Tait D. Shanafelt: design of the study, interpretation of the data, revising the manuscript for intellectual content. Christopher M. Keran: design of the study, analysis and interpretation of the data, drafting and revising the manuscript for intellectual content. Neil A. Busis: design of the study, interpretation of the data, drafting and revising the manuscript for intellectual content. Laura A. Foster: analysis and interpretation of data, revising the manuscript for intellectual content. Jennifer Rose V. Molano: analysis and interpretation of data, revising the manuscript for intellectual content. Cormac A. O'Donovan: analysis and interpretation of data, revising the manuscript for intellectual content. Jeffrey B. Ratliff: analysis and interpretation of data, revising the manuscript for intellectual content. Heidi B. Schwarz: design of the study, analysis of the data, and revising the manuscript for intellectual content. Jeff A. Sloan: design of the study, interpretation of the data. Terrence L. Cascino: conceptualization of the study, commenting on manuscript for intellectual content.
STUDY FUNDING
Study funded by the American Academy of Neurology.
DISCLOSURE
K. Levin receives honoraria from the American Academy of Neurology and American Board of Psychiatry and Neurology and royalties from UpToDate and Elsevier. T. Shanafelt reports no disclosures relevant to the manuscript. C. Keran is an employee of the American Academy of Neurology. N. Busis receives honoraria for speaking at American Academy of Neurology courses and for serving as Alternate CPT Advisor for American Academy of Neurology, and is a member of the American Academy of Neurology Board of Directors. L. Foster reports no disclosures relevant to the manuscript. J. Molano serves on the Neurology Now and NEJM Journal Watch Neurology Editorial Advisory Boards. Cormac A. O'Donovan and Jeffrey Ratliff report no disclosures relevant to the manuscript. Heidi Schwarz receives honoraria for speaking at the American Academy of Neurology courses, and is member of the American Academy of Neurology Board of Directors and Chair of the Practice Committee. J. Sloan reports no disclosures relevant to the manuscript. T. Cascino serves as President of the American Academy of Neurology. Go to Neurology.org for full disclosures.
ACKNOWLEDGMENT
Additional members of the AAN resident burnout committee: Blair Ford, Joseph Kass, Sarah Mulukutla, Thomas Shoemaker, Thomas Vidic.
Footnotes
Go to Neurology.org for full disclosures. Funding information and disclosures deemed relevant by the authors, if any, are provided at the end of the article.
Supplemental data at Neurology.org
Editorial, page 421
- Received December 19, 2016.
- Accepted in final form April 3, 2017.
- © 2017 American Academy of Neurology
REFERENCES
Letters: Rapid online correspondence
- Author response to Dr. Sethi
- Burnout among residents and fellows: It is not all just work
REQUIREMENTS
If you are uploading a letter concerning an article:
You must have updated your disclosures within six months: http://submit.neurology.org
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.
