Abstract
Objective To evaluate levels of burnout and correlates of burnout among US rheumatology fellows.
Methods US rheumatology fellows were invited to complete an electronic survey in 2019. Burnout was assessed using the Maslach Burnout Inventory. Measures of depression, fatigue, quality of life, and training year were also collected. Open-ended questions about perceived factors to promote resiliency and factors leading to increased burnout were included. Bivariate and multivariate regression analyses were used to examine correlates of burnout. Open-ended responses were analyzed using thematic analysis.
Results The response rate was 18% (105/582 pediatric and adult rheumatology fellows). Over one-third (38.5%) of postgraduate year (PGY) 4 and 16.7% of PGY5/6 fellows reported at least 1 symptom of burnout. Of PGY4 fellows, 12.8% met criteria for depression compared with 2.4% of PGY5/6 fellows. PGY4 fellows reported worse fatigue and poorer quality of life compared with PGY5/6. In multivariable models controlling for training year and gender, older age (> 31 years) was associated with lower odds of burnout. Thematic analysis of open-ended responses identified factors that help reduce burnout: exercise, family/friends, sleep, support at work, and hobbies. Factors contributing to burnout: pager, documentation, long hours, demands of patient care, and presentations and expectations.
Conclusion This national survey of US rheumatology fellows reveals that early trainee level and younger age are associated with worse levels of fatigue, quality of life, and burnout. Although awareness of and strategies to reduce burnout are needed for all fellows, targeted interventions for younger fellows and those in their first year of training may be of highest yield.
Burnout, a psychological state characterized by emotional exhaustion, depersonalization, and perceived lack of effectiveness (Maslach 1981),1 is an increasingly recognized entity within the medical community. Previous studies have demonstrated a higher prevalence of burnout among physicians as compared to nonphysician peers, and, in particular, increased odds of burnout in residents/fellows as compared with population control samples.2 A survey of US physicians from 2011 to 2014 demonstrated worsening burnout and satisfaction with work-life balance with more than half reporting professional burnout.3 The 2019 Medscape National Physician Burnout, Depression, and Suicide Report of more than 15,000 physicians across 29 specialties currently practicing medicine in the US found that burnout among practicing rheumatologists is substantial, albeit on the lower end of the spectrum of all physicians, with 41% self-reporting burnout.4 However, rheumatology trainees were not included. Among practicing rheumatologists,5 lower job satisfaction ratings correlated with one specific component of burnout: emotional exhaustion. Conversely, better job satisfaction correlated with higher personal accomplishment.5
Prior studies have shown that dissatisfied physicians were 2 to 3 times more likely to leave medicine than satisfied physicians.6 This is particularly relevant as physician and trainee retention in rheumatology is critical given the significant projected workforce shortage. In the US, the demand in 2030 is expected to exceed supply by 4133 clinical full-time equivalent (FTE) rheumatology practitioners, up from a baseline estimated shortage of 700 FTE in 2015.7 In this environment, assessment of burnout among rheumatology trainees is a crucial first step in addressing rheumatologist retention.
Gender-based differences in physician burnout are increasingly identified8 and emphasize unique societal, individual, and institutional risk factors for female physicians. One study of primary and specialty nonsurgical physicians (N = 5704) reported the odds of women reporting burnout as 1.6 times that of men.9 The 2019 Medscape survey found that 50% of women physicians self-reported burnout vs 39% of men.4 The demographics of practicing rheumatologists are shifting toward a majority women subspecialty. The 2018 workforce data reported that men comprised 59% of the rheumatology workforce; however, it is anticipated men will constitute only 43% of the workforce in 2030.7 Thus, addressing gender-based differences in burnout is an important factor in rheumatologist retention.
Although prior studies report significant burden of burnout in physicians and in trainees as a whole, to our knowledge, there are no studies to date that have assessed rates of burnout among US trainees in rheumatology fellowship. Fellowship represents a unique period in which physicians early in their careers voluntarily choose to undergo additional training to become subspecialists while possessing sufficient training to be an independent practitioner. This study aimed to investigate burnout levels and identify correlates of burnout among US rheumatology fellows and thus fills an existing gap in the literature.
METHODS
Design. This is a cross-sectional study of US rheumatology fellows conducted anonymously using a REDCap electronic survey tool from January 2019 to February 2019.
Participants. All fellows at Accreditation Council for Graduate Medical Education–accredited adult and pediatric rheumatology programs were invited to participate through email.
Procedure. The American College of Rheumatology (ACR) program directors’ email listserv and ACR Fellows-In-Training Google group were used to disseminate the invitation. Participation was voluntary. Participants were given an incentive for participation in the form of a raffle of eight $25 Amazon.com, Inc. gift cards. We have obtained the required ethical approvals and consent for this manuscript. The study was approved by the institutional review board (IRB) of Oregon Health & Science University (IRB no. 00018734).
Survey measures. Participants provided information on demographics (age, gender, ethnicity, race, marital status), training year, and average hours worked per week, as well as on burnout, depressive symptoms, fatigue, and quality of life.
Primary outcome: burnout. Burnout was measured using the validated Maslach Burnout Inventory (MBI)-Human Services Survey.10 This instrument includes 22 items that measure the 3 components of burnout: emotional exhaustion, depersonalization, and personal accomplishment. We measured burnout using the single-item measures for emotional exhaustion and depersonalization adapted from the full MBI as has previously been used and described.2,3,11-12 A high emotional exhaustion or depersonalization score on the MBI was considered when a respondent marked a frequency of “once a week” or more often and thus were categorized as “burned out.” There were 7 response categories in total ranging from “never” to “every day.”
Secondary outcomes: depression, quality of life, and fatigue. Depression was measured using the 2-item Patient Health Questionnaire, which has demonstrated validity in prior studies.13,14 Quality of life and fatigue measures were obtained using the standardized visual analog scale (0 = worst, 10 = best) as captured in prior studies of physician burnout.2
Additional survey items: institutional wellness programs and satisfaction. Included in the survey were 2 items regarding presence and use of an institutional wellness program. Additionally, 2 items inquired about satisfaction of work-life balance and satisfaction of current professional position, as has been done in prior studies.15
Open-ended questions. Open-ended questions asked respondents to report perceived factors they felt promoted resiliency and reduced burnout as well as factors that worsened or lead to increased levels of burnout. Open-ended questions were worded as follows: “What strategies or factors have you found beneficial to promote resiliency and reduce burnout?” and “What factors have you found that worsen or lead to increased levels of burnout?”
Statistical analysis. We used summary statistics to describe respondent characteristics and used Student t tests (for continuous variables) or chi-square tests (for categorical variables) to compare groups. Bivariate and multivariate logistic regression analyses were used to examine correlates of burnout. Analyses were conducted using STATA statistical software (StataCorp).
Qualitative analysis. Open-ended responses were analyzed using thematic analysis. Two independent reviewers (JM and JLB) evaluated all comments, and common themes were identified. Two independent reviewers (JM and JLB) read all responses and independently identified themes. Reviewers met in person to discuss findings and resolved discrepancies by consensus.
RESULTS
The response rate was 18% (105/582 total pediatric and adult rheumatology fellows) based on data of trainees from 2018 to 2019.16 Survey responders were 64.8% female, 69.5% aged between 31 years and 35 years, 59% White, and 61.9% married (Table 1), which was similar to the overall population of rheumatology fellows in terms of gender and race. A smaller proportion of postgraduate year (PGY) 4 fellows responded to the survey as compared to the overall population (35.2% vs 45.2%).16
Characteristics of US rheumatology fellow study participants.
Primary outcome. Over one-third (38.5%) of PGY4 and 16.7% of PGY5/6 rheumatology fellows were found to have at least 1 symptom of burnout when assessing burnout using the single-item measures for emotional exhaustion and depersonalization adapted from the full MBI (Table 2).
Burnout, depression, and quality of life of US rheumatology fellows by training year.
Secondary outcomes. Of PGY4 respondents, 12.8% met criteria for depression compared with 2.4% in PGY5/6 (P = 0.02; Table 2). In addition, PGY4 fellows reported worse fatigue, with 51.3% reporting high fatigue levels. Also, PGY4 fellows reported poorer quality of life compared with PGY5/6 fellows, reaching statistical significance in overall, mental, and physical quality of life categories.
Correlates of burnout. Older age was associated with decreased odds of burnout in both bivariate and multivariate analyses (Table 3). Additionally, PGY5/6 training year was associated with reduced odds of burnout in the unadjusted model (OR 0.32, 95% CI 0.13-0.80), but this effect was attenuated in adjusted models. There was no significant variation by gender, race, or relationship status.
Unadjusted and adjusted odds of burnout among US rheumatology fellows.
Open-ended responses. Table 4 highlights results of the open-ended survey responses. Participants identified factors that promote resilience and reduce burnout, including exercise, family/friends, sleep, support at work, and hobbies. Factors that contribute to burnout included pager, documentation, presentations/expectations, long hours, and demands of patient care. The following are participant quotes regarding specific factors that contribute to burnout.
Thematic analysis of open-ended responses on factors that improve and worsen burnout.
• Documentation:
“…with the number of outpatient and inpatient notes, I find myself spending hours after coming home finishing up notes and working on presentations nearly every day, as well as spending the weekends catching up on notes.”
• Demands of patient care:
“Hands down, MyChart/patient portal messages—especially after hours—is the biggest contributor to physician burnout, especially because it is an unpaid service.”
“High clinical load without ability to ‘offload’ some tasks from time to time”
“A high inpatient census of severely ill patients”
• Long hours:
“…A busy service when not leaving hospital until 9 pm on many nights…”
• Pager:
“Overnight pagers for nonurgent issues”
“Excessive pages overnight”
• Presentations and expectations:
“Research requirements, scholarly requirements outside of regular days’ work”
“Pressures in research to pursue a particular kind of research (eg, institution can only fund basic/translational research when fellow wants to do clinical research or vice versa), writing numerous grants, writing IRB proposals.”
DISCUSSION
This national survey of US rheumatology fellows identified that among PGY4 fellows, over one-third (38.5%) met criteria for burnout. Additionally, 12.8% screened positive for depression and 51.3% reported high levels of fatigue. Early trainee level and younger age among respondents were associated with worsened levels of fatigue, quality of life, and burnout. A prior study demonstrated that the first year of residency was independently associated with burnout in 8 different medical specialties.17 Further study is needed to identify factors to explain this similar finding among rheumatology fellows. We hypothesize that potential factors include greater clinical workload and difficulty adjusting to the new training program environment and to the field of rheumatology. These findings support the need for efforts to raise awareness and identify strategies to prevent and reduce burnout among rheumatology fellows, particularly those who are younger and early in training. Regarding other fellow populations, our population of PGY4 fellows demonstrated lower rates of burnout than pediatric fellows (53%)18 and neurology fellows (55%)19 but rates of burnout similar to those of pulmonary/critical care fellows (32%)20 and nephrology fellows (30%).21 As our group was predominantly adult rheumatology fellows, it is possible that adult medical subspecialties fellows have lower rates of burnout than other fellow groups. However, further study is needed to clarify and better understand the differences between these groups.
Further, our open-ended survey responses were particularly illuminating, and could be used to guide changes to fellowship programs to reduce burnout. Responses highlighted certain elements as particularly problematic for worsening burnout, such as documentation, long hours, and scholarship, including expectation for research and oral presentations. The identified factors leading to reduced burnout were largely factors outside of medicine, including family/friends, sleep, exercise, and hobbies. Some fellows emphasized that support from colleagues and superiors served as an important factor to reduce burnout.
Retention of rheumatologists is critical in light of the predicted rheumatologist shortage in the US. This effort to support and retain physicians should begin while physicians are in training to become rheumatologists, particularly during their first year of fellowship, to avoid drop-out and to promote lifelong resiliency strategies to prevent burnout. Although awareness of and strategies to reduce burnout are needed for all fellows, targeted interventions for younger fellows and those in their first year of training may be of highest yield.
Although our study is the first, to our knowledge, to evaluate burnout in rheumatology fellows on a national scale, it is not without limitations. Our response rate was suboptimal, with only 18% of rheumatology fellows participating. The low response rate may be explained by the fact that the survey was sent to the program directors, who were asked to disseminate it to their fellows as well as to a voluntary listserv, which not all rheumatology fellows subscribe to. Hence, in all likelihood, we were unable to reach all fellows-in-training with our invitation to participate. In the future, the ACR may consider adding questions regarding burnout to the in-service exam, as a previous study relying on such an approach among surgical residents achieved a 99% response rate.22 Additionally, our survey results may be biased, as those experiencing substantial burnout may have been less likely to participate, and therefore we may have missed capturing those at highest levels of burnout. Conversely, those most affected by burnout may have been more likely to participate as this topic is directly related to their current situation. Also, we are unable to evaluate if nonresponders were similar to the responders. Our study design was not longitudinal; thus, we were not able to measure changes in burnout as fellows advance through training and into their careers. Such a design would be beneficial, especially as burnout levels likely fluctuate through training as rheumatology fellows become more comfortable with the field of rheumatology and their role within it. Finally, we did not assess if fellows remained at their residency-training home institution or in the same geographic location for fellowship, which could potentially affect burnout.
Moreover, our survey was conducted prepandemic, and the effect of coronavirus disease 2019 (COVID-19) on burnout in our population is unknown. A large study of 2440 US physicians showed significantly higher rates of burnout in 2021 in comparison to 2020 (62.8% vs 38.2%; P < 0.001).23 Initial studies have shown mixed data regarding trainee burnout levels, with some studies surprisingly demonstrating similar prevalence of burnout in medical and emergency medicine residents during the pandemic compared to prepandemic levels,24,25 although these studies were completed early in the pandemic. Another study showed high levels of burnout in residents of multiple specialties (65.8% vs 51% prior to pandemic)26 18 months into the pandemic. Additionally, pediatric trainees during the pandemic were found to have increased odds of burnout (emotional exhaustion) vs practicing pediatricians (OR 5.94, 95% CI 1.85-19.02).27 Overall, data are limited thus far, especially in trainee populations. Further longitudinal research is indicated to better identify the effect of the COVID-19 pandemic on burnout levels, at-risk populations, and targets for intervention.
Last, the evaluation of interventions aimed at preventing or reducing burnout in rheumatology and medicine is needed.
In conclusion, burnout among physicians, medical students, and trainees remains a serious, ongoing problem, and significantly worsened throughout the COVID-19 pandemic.27,28 Our study revealed higher levels of burnout and depression prepandemic among first-year rheumatology fellows compared with senior fellows. Targeted interventions toward younger fellows or those in earlier years of training may be of highest yield. Factors identified that may contribute to lowering burnout rates include exercise, spending time with family and friends, adequate sleep, hobbies, and support at work. Learning how to promote wellness in life may be as important as the lifelong learning strategies we teach fellows. Ongoing efforts to address the issue of burnout are necessary across fields of medicine. These efforts are of the utmost importance for rheumatology fellows to promote retention and avert an even higher projected workforce shortage in rheumatology.
Footnotes
The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health or the Department of Veterans Affairs.
The authors declare no conflicts of interest relevant to this article.
- Accepted for publication March 2, 2023.
- Copyright © 2023 by the Journal of Rheumatology
