Association of Self-Reported Burnout and Protective Factors in Single Institution Resident Physicians Open Access

Work-related burnout in residents is prevalent and associated with increased depression levels, medical errors, and patient safety problems.^1–5  Graduate medical education programs have implemented a variety of approaches to address burnout.⁶  Yet a 2017 systematic review⁷  found that the only intervention to consistently improve the domains of emotional exhaustion and overall burnout scores was changing resident work hour limits; all other interventions produced mixed results.

Other studies^8–12  offer some evidence to suggest there may be protective factors for burnout in residents: grit, resiliency, and social support. Grit is a personality trait; it describes an individual's perseverance toward reaching a long-term goal⁹  and measures one's ability to maintain sustained effort for an extended period of time.¹³  Resiliency is a skill that helps individuals “bounce back” from stressful events, regain composure, and persevere. Both are useful in overcoming challenges but may function differently as protective factors against burnout. Another factor, high levels of social support from friends, family, and colleagues,¹⁴  may provide a layer of safety and comfort as well as means for coping with difficulties.⁸ 

Psychological flexibility is another potentially protective construct. Psychological flexibility has not been studied in residents, but it has been evaluated in other workplace settings.¹⁵  It is a measure of an individual's ability to focus on the current situation and take action to achieve their goals, depending on the available opportunities. It has been considered an important process in helping individuals reframe their thinking before they feel overwhelmed, possibly mitigating against burnout in residents.¹⁶  The purpose of this study was to use mixed methods to explore the association of these 4 potentially protective factors for burnout—grit, resiliency, social support, and psychological flexibility—across a wide variety of residency programs at 1 institution to gain insights for future program interventions.

Participants included residents and fellows who were training in 1 of the 20 programs at the Medical College of Georgia at Augusta University. The study employed an explanatory sequential design,¹⁷  in which quantitative data are followed by qualitative data for interpretive purposes. An online survey was created with questions from previously used scales. Burnout was measured with the Maslach Burnout Inventory–Human Services Survey,²  which is a shortened 1-item version. Scores range from 1 to 5, with higher scores indicating a lower level of burnout. The 1-item measure was selected over the full 22-item inventory to reduce the total number of survey questions, given that the main survey focus was on the 4 potentially mitigating psychological factors.

Grit was measured using the 8-item Short Grit Scale.^18,19  Responses ranged from 1 (not like me at all) to 5 (very much like me). The total Short Grit Scale score is calculated as an average of the 8 items. The maximum score on this scale is 5 (extremely gritty), and the lowest score is 1 (not at all gritty). Resiliency was measured using the 10-item Connor-Davidson Resilience Scale.^20,21  Responses are on a 5-point scale, ranging from 0 (not true at all) to 4 (true nearly all the time). The 10 items are summed to yield a total score ranging from 0 to 40, with higher scores reflecting greater resilience. Social support was measured using the 12-item Interpersonal Support Evaluation List-12 questionnaire.²²  For the purposes of this study, reverse-scored items were reworded so that all item responses ranged from 1 (definitely false) to 4 (definitely true). The 12 items are summed to yield a total score of 12 to 48, where a higher score indicates higher social support. Psychological flexibility was measured using the 7-item Work-Related Acceptance and Action Questionnaire.²³  The items used a scale ranging from 1 (never true) to 7 (always true), in which higher scores indicate a greater level of psychological flexibility. The 7 items are summed to yield a total score of 7 to 49. The survey also included demographic information (gender, age, race, ethnicity, relationship status, specialty, residency year). These 5 survey instruments were combined into a single online survey and sent to all 527 residents in the 2017–2018 academic year.

All statistical analyses were performed using SAS version 9.4 (SAS Institute Inc, Cary, NC), and statistical significance was assessed using an alpha level of .05. Descriptive statistics (frequencies and percentages or means and standard deviations) were calculated. The Spearman rho correlation coefficients were calculated to determine the association of burnout with grit, resiliency, social support, and psychological flexibility scores. Due to the ordinal nature of the dependent variable (burnout), ordinal logistic regression was used to determine the association of burnout with grit, resiliency, social support, psychological flexibility, and demographic variables. This model estimates a proportional odds ratio (OR) for each predictor (protective factors) when shifting to the next burnout category (1–5). Each independent variable was first examined in a bivariate model on burnout.

Additionally, 13 residents volunteered to participate in individual semistructured interviews that explored their experiences with stress and burnout (table 1). They volunteered to participate in the interview by including their names and contact information at the bottom of the survey. The research team chose an educational researcher (T.R.W.) not involved with the residency programs to conduct the interviews. The team thought the interviewer's lack of prior involvement would allow her to appear impartial and nonthreatening when probing and asking follow-up questions.

The interviewer took a pragmatic approach²⁴  in the process of data collection and analysis, which accepts philosophically that there are singular and multiple realities that can best be measured and observed using a combination of mixed-methods research. In this case, the survey examined the strength and interrelationship among the constructs, while the interviews examined the constructs as applied in “the real world.”²⁴  The interview protocol was tested among the researchers, piloted on residency directors, and revised several times until all questions were clear and prompted participants' recollection and articulation of their stress and burnout experiences (interview questions provided as online supplemental material).

Each face-to-face or telephone interview was audio-recorded and lasted 45 to 60 minutes. Through ongoing memos and discussions, the research team iteratively analyzed the interviews after each trainee was interviewed. The research team believed saturation had been reached by the 13th interview when the data began to repeat itself. The interviews were then transcribed and analyzed initially by a research assistant and a second time by the interviewer using direct content analysis.²⁵  The research assistant initially conducted a deductive analysis to identify where trainees implicitly or explicitly discussed grit, social support, and psychological flexibility as protective factors for burnout. The research assistant, an undergraduate student interested in educational research as part of his preparation for medical school, was unaffiliated with graduate medical education.

Following each of the first 3 interviews, the interviewer checked the accuracy of the initial coding, engaged in discussions to resolve coding discrepancies, and amended the code book for subsequent coding. The interviewer then conducted an inductive analysis to understand how and when residents applied these factors to mitigate stress and burnout. The inductive analysis was then discussed with the research assistant to ensure accuracy and trustworthiness of the interpretation. Resiliency as a protective factor was not explored because it was not included in the final linear regression model.

The study was approved by the Augusta University Institutional Review Board.

Of the 527 residents and fellows training at Medical College of Georgia, 268 responded (51%). Responses with missing scores for burnout were removed from the data set (n = 4). The average age of participants was 30.6 years (SD 4.1); 54% (142 of 265) were male; 67% (178 of 266) were married or had a long-term partner; and 63% (163 of 258) were non-Hispanic white. The results indicated residents were experiencing occasional burnout. The average score for burnout was 3.6 (SD 0.9) with 148 (56%) answering, “Occasionally I am under stress,” and 59 (22%) answering, “I am definitely burning out.” Relationship status was associated with burnout (ie, “married or long-term partner” respondents were less likely to report highest level of burnout, while “other or divorced” were more likely), but none of the other demographic characteristics, year in residency, or specialty were associated with different burnout levels. Results of Kruskal-Wallis tests showed that modest-to-trivial differences existed in all 4 protective factors by burnout level (table 2). All protective factors were associated with decreased burnout, according to the Spearman rho correlations (provided as online supplemental material).

Although all 4 factors and residents' relationship status were associated with burnout in each of the bivariate models, resiliency and relationship status were no longer significant in a single model with all 4 factors (table 3). Associations are reported as proportional odds ratios and related 95% confidence intervals (CIs) and P values. Higher psychological flexibility scores were associated with lower burnout score (5, no burnout; OR 1.12; 95% CI 1.08–1.17). Higher grit scores were associated with lower burnout score (5, no burnout; OR 1.65; 95% CI 1.06–2.57). Similarly, higher social support scores were associated with lower burnout score (5, no burnout; OR 1.07; 95% CI 1.03–1.11).

Resident interviews indicated all 13 residents were attempting to balance the competing demands of training within the context of personal and professional goals. They found this balance and the effort required to learn how to prioritize tasks and activities challenging. In describing grit, residents referred to it as an asset to their performance, yet also detrimental to their overall health and well-being. Grit fueled determination and perseverance, which allowed residents to push through challenging times, but this was seen as coming at a high price, especially in how much residents sacrificed their personal goals throughout training. For example, a neurology resident described it this way:

Grit as both a proactive factor and contributor to burnout was a surprising finding given that the literature conceptualizes it only as a protective factor.

Residents' social networks included family, friends, social media, mental health therapists, and faith communities. However, despite the multiple networks available, residents preferred their immediate peer group for support. About half (6) of the 13 residents perceived their peers as having a more nuanced understanding of the work environment and therefore the best network for working through their stressful experiences. Talking to other residents yielded a sense of satisfaction in processing negative feelings because of a shared understanding of the tasks, activities, and contexts that compose residency. Residents who did not report having this network discussed the lack of avenues to process daily emotional burdens. However, those who described using their peer network found it helpful, as this pediatric resident explained:

The shared understanding among residents helps them take shortcuts in explaining daily events and processing their unarticulated negative feelings. However, some residents expressed conflict in approaching their colleagues with their stressful experiences because sharing negative experiences can affect the whole team. Residents explained that, although it helps to process negative feelings with colleagues, it also contributes to an environment that potentially jeopardizes the team's function.

Nine residents discussed employing psychological flexibility to deal with negative issues contributing to feelings of burnout. The majority of the residents talked about psychological flexibility in terms of disengaging with the stressful moment and trying to establish another perspective. This idea of “taking a step back” was employed at multiple levels as residents performed their daily activities. Some used this strategy to help frame their development as a resident; others reflected on the stressful situations and compared them to the stress experienced in previous years. Overall, this strategy helped residents gain perspective in the stressful moment and better engage with the situation. This process was described by a pediatric emergency medicine resident:

Temporarily disengaging from the stressful situation and taking on a new perspective helped residents manage moments that felt insurmountable in their clinical practice.

In this study of the associations between 4 protective factors and residents' burnout levels, our results support that grit, resiliency, social support, and psychological flexibility are important variables related to burnout level. Although all 4 factors appear related to burnout, in this study of residents in 20 different specialties, resiliency and relationship status were no longer associated with residents' burnout levels when all 4 were entered into a model (figure). In addition, grit had the largest association with burnout, which confirms previous research.^8,11,12 

The study findings also suggest that residency programs should consider developing structured opportunities for residents to establish meaningful connections with each other in an effort to mitigate stress and burnout.²⁶  Residents appear to benefit from a strong sense of community in which they can feel psychologically safe to talk with one another about their challenges.²⁷  These opportunities may provide a form of emotional catharsis and a means of emotionally processing events. Residency programs that have created these structured opportunities report long-standing effects on residents' wellness and engagement²⁸  and may serve as models for others interested in creating programs at their institutions.

Additionally, many programs focus on resiliency in their wellness programs, but our findings show that resiliency was no longer significant when the other constructs were added to the model. Whereas resiliency assists individuals with recovery after a setback,²⁹  grit assists individuals in focusing on long-term goals, which is useful in environments requiring constant high achievement for success. Therefore, residency programs may want to emphasize the importance of having grit as an essential characteristic of residency success but equip residents with the tools they need to psychologically disengage when needed.

Our study findings are limited by our use of a single institution, which does not necessarily generalize to other settings. In addition, the response rate of 51% does not allow us to conclude that the findings are representative of the full population of residents. The survey questions were modified from those used in other settings and were not tested for validity in this particular population for this purpose, which also may affect our findings. The use of a single item for the definition of burnout may have affected the associations we found. Finally, the selection of 13 residents may have influenced our interpretation of how these psychological constructs interact with burnout in indeterminable ways.

The results indicate this line of inquiry is potentially worth pursuing. We suggest that any future research investigate the relationship between grit and burnout in a multisite study, potentially using the 22-item burnout scale rather than the 1-item scale. In future studies researchers may wish to examine whether interventions targeting grit and psychological flexibility may reduce resident burnout.

In this study of 20 different residency programs, we found grit, resiliency, social support, and psychological flexibility to be associated with burnout in residents, but when placed in an ordinal logical regression model, resiliency and relationship status were explained by the 3 other factors. Grit alone had the largest association with burnout. Resident interviews highlighted the importance of peer support and psychological flexibility in dealing with the stresses of residency.