Evidence-Based Interventions that Promote Resident Wellness from the Council of Emergency Residency Directors

Initiatives for addressing resident wellness are a recent requirement of the Accreditation Council for Graduate Medical Education in response to high rates of resident burnout nationally. We review the literature on wellness and burnout in residency education with a focus on assessment, individual-level interventions, and systemic or organizational interventions.


Parsons et al.
Evidence-Based Interventions that Promote Resident Wellness programs, well-being, stress, burnout, physicians, residents, and health personnel. Articles were prioritized if they focused on EM residents. When EM-specific literature was not available, we included relevant articles pertaining to wellness among other healthcare personnel. Bibliographies of all relevant articles were reviewed for additional studies. The literature search yielded 2931 articles, which were screened by two authors to include any papers addressing the following themes: assessment tools for wellness/burnout; individual interventions to treat/prevent burnout; and organizational interventions to treat/prevent burnout. After screening, 112 articles were deemed directly relevant for inclusion. We provide level and grade of evidence for each statement according to the Oxford Centre for Evidence-Based Medicine criteria (Tables 1 and 2). When supporting data were not available, recommendations were made based upon the authors' combined experience and expert opinions. Prior to submission, the manuscript was reviewed by the CORD Best Practices Subcommittee for additional comments and identification of missed references. It was additionally posted to the CORD website for two weeks for review from the CORD community.

Assessment of Burnout/Wellness
The first step in any successful intervention is a needs assessment. While large population-level studies have demonstrated a need to improve physician well-being, 2 a targeted needs assessment and problem identification is suggested prior to any specific intervention, according to the second step in Kern's six-step approach to curriculum development. 17 A comprehensive needs assessment first requires identification of the population of interest (e.g., EM residents in a single training program) and the specific problem (e.g., burnout). A common pitfall is failure to narrow the scope of the problem that an intervention is designed to address. This can be challenging as there is lack of an agreed-upon definition for "burnout" or "wellness." 18,19 Because burnout is not listed in the Diagnostic and Statistical Manual of Mental Disorders-V, 20 it is common for the term "burnout" to be incorrectly used to refer to anything ranging from depressive symptoms to increased work demands. For this reason, surveys that ask participants to self-identify their perceived burnout as a single-item Likert response (e.g., "rate your level of burnout") are not valid measures.
Wellness may be variably defined as "work-life balance," "life satisfaction," or "the absence of burnout," depending upon the context of an intervention. 21 It is not sufficient to state that an intervention is designed to "promote wellness" without first specifying the framework that defines wellness in a particular circumstance. Knowledge of the operating characteristics of commonly used assessment tools can facilitate problem identification. For example, burnout, 22 depression, 23 anxiety, 24 perceived stress, 25 and resilience, 26 each have validated assessment tools for measurement in different demographic groups, such as medical students, physicians, and the health professions in general. Table 3 provides a review of the key tools available to assess these components, including whether it was validated specifically in physician populations.
Other things to consider when selecting an appropriate assessment tool include cost, ease of completion, potential confounding factors, and whether it has been studied in target populations. Cost may be a significant factor for selection of the assessment tool. Although some surveys such as the Maslach Burnout Inventory (MBI-HSS) require payment for use, 27 many tools are available at no cost. Survey length can have a negative impact on the response rate, influencing the tool selected. One must also consider the influence of confounding factors; for example, the MBI-HSS emphasizes the assessment of thoughts and feelings in the workplace and is less likely to be confounded by non-workplace related factors than other tools. 28 Lastly, not all tools have been created with attending or resident physicians in mind. This particular population may face unique stressors and challenges and may not exhibit the same response patterns as the general public.
The MBI-HSS is probably the most widely recognized measure of burnout in physicians. Maslach and colleagues *"Extrapolations" are where data were used in a situation that has potentially clinically important differences than the original study situation.
Evidence-Based Interventions that Promote Resident Wellness Parsons et al.
operationally defined burnout for workers in the helping professions (e.g., healthcare workers, first responders, social workers) as a combination of three domains: EE, DP and lack of PA. 22 The original authors preferred that burnout be reported as a continuous variable (i.e., high, medium, or low) rather than as a dichotomous one (i.e., burnout or no burnout). 22 Since then, multiple researchers have applied their own criteria for burnout using the three subscales, which has resulted in at least 47 distinctly different definitions of burnout using the MBI-HSS alone. 29 Therefore, when considering the assessment of burnout, it is essential to remember the framework used and the test characteristics of the tool selected.
Ongoing assessment and program evaluation are essential to determine the efficacy and successful achievement of goals and objectives. The interval of measurement can vary from weeks to months depending upon the scope and outcome of interest. For example, optimism is considered a relatively stable quality 30 and unlikely to be sufficiently changed by a single fourweek curricular intervention; therefore, an optimism/pessimism assessment tool (e.g., the Revised Life-Orientation Test) would be ill-suited for a pre-/ post-measurement in this context. It is also important to note that survey fatigue can result in lower quality data by introducing bias, including non-response bias. 31 Finally, a distinction should be made between institutionand individual-level assessment. The updated ACGME Common Program Requirements emphasize the need for mechanisms to identify residents at risk of burnout, depression, substance abuse, suicidal ideation, and the potential for violence. 32 They specify the need for program-level assessment, individual-level assessment, and self-screening measures. The ACGME has endorsed the MBI-HSS, Mayo Clinic Well-Being Index, Patient Health Questionnaire-9, and the Professional Quality of Life Scale as potential tools for assessment at the program-level and individual-level. 12 While each of the recommendations and tools discussed above can be used for assessment of either the program or individual residents, extra caution should be taken when assessing specific residents who may be at risk. Evidence exists that trainees at risk of suicide completion are often difficult to identity, 33,34 and overreliance on any one assessment tool should be avoided. Identification of at-risk residents should be a composite evaluation of attention to professional duties, clinical performance, known work-related or personal crises, and concerning changes in behavior or language used by the resident. 35

Individual Interventions
Over the last two decades, there have been increasing calls for high-quality randomized controlled trials of specific burnout interventions. A recent, systematic review summarizing psychosocial interventions for managing physician workplace stress identified over 15,000 studies containing the keywords physician, stress, and burnout. Unfortunately, only 20 were intervention studies, and among these only 12 included pre-and post-intervention assessments. 36 None were deemed high-quality or specific to EM residents. 36 A similar systematic review in GME found only three intervention studies. 21 Several barriers common to medical education research exist, including variable definitions, small sample sizes, the ethics of randomization, difficulty with individual assessments, long-term follow-up, and external validity. Among the higher-quality intervention studies, relaxation training, behavioral interventions, and self-care were demonstrated to be most effective. 21,36 1. Clearly define the purpose and need for an intervention, then choose an assessment tool for gathering data, establishing a baseline, measuring outcomes, and/or monitoring.

Best Practice Recommendations for Assessment
Mindfulness is defined as an awareness of the present situation while accepting thoughts, emotions, and physical sensations. 37 Being mindful can be a valuable tool for combating burnout. One randomized controlled trial focusing on mindfulness-based coping strategies demonstrated enhanced well-being and decreased rates of burnout after the intervention. 38 Another study using a modified mindfulnessbased stress reduction program, consisting of a workshop followed by eight weeks of daily meditation, showed improvement in general health and stress. 39 Similar mindfulness and meditation studies have been reported in the undergraduate and the nursing wellness literature. [40][41][42][43][44] More feasible interventions, such as 10-20 minute mindfulness meditation for 30 days and a one-hour online module focusing on "mindbody skills training," have also been demonstrated to reduce stress and burnout. [45][46][47][48][49][50][51][52] Mindfulness and self-awareness can be enhanced by journaling, narrative medicine, and reflective questioning, potentially reducing burnout rates, 50-51 although other studies have not shown improvement in burnout. 52 Focused training in behavioral skills, such as cognitive reframing, self-compassion, and empathy, can also improve wellness. [56][57][58][59][60][61] A survey of winners of the American Medical Association Foundation's Pride in Professions Award identified self-compassion and self-care as key components for combating burnout. 62 A study of internal medicine residents found that residents who employed strategies of active coping and positive reframing of difficult situations had lower rates of EE and DP. 25 Similar results have been found in the nursing literature. 63 While behavioral skills training has been shown to improve physician

Evidence-Based Interventions that Promote Resident Wellness
Parsons et al.
wellness, stress management training has not been shown to improve burnout rates in physicians. 52,64 Physicians who are able to engage in regular self-care, such as ensuring adequate physical health, sleep, nutrition, and exercise, consistently have lower rates of burnout. [65][66][67][68][69][70] Although scheduling can be a challenge, regular exercise improves physician wellness. 71,72 Incentivized exercise programs, however, have not been shown to improve physician wellness. 52,64 A study of approximately 7000 surgeons showed that those who visited their primary care physician (PCP) in the prior 12 months had lower rates of burnout and a higher quality of life. 71 Unfortunately, a recent survey reported that nearly half of surgical residents reported not being able to visit their PCP regularly and gaining weight during residency. 73 Fortunately, the literature demonstrates that we can teach our trainees how to implement effective behavioral change plans to improve personal behaviors such as exercise, nutrition, sleep, personal hygiene, and emotional health. 74 In addition to physical health, psychological health is also important. A survey of anesthetists and intensivists showed that alcohol dependence, abuse of sedative medications, and overeating were correlated with higher rates of depression and burnout. 75 Another study looking at residents found that only 24% of residents who felt they needed mental health care sought treatment. 76 Residents cited lack of time, concern of confidentiality, and cost as barriers to treatment. Encouragingly, studies looking at physicians who participated in either individual or group counseling with a trained professional showed a lasting reduction in EE for up to three years. 77 Building a strong friend and colleague support network is associated with lower burnout scores. [78][79][80][81] Indeed, Dr. Maslach herself recently proposed that civility and teamwork are fundamental to physician wellness. 79 A survey of 198 physiciansin-training showed that loneliness was significantly associated with both personal and professional burnout. 81 Peer-and facultymentoring programs may help create these needed support networks as they are correlated with lower burnout rates. 82 For residents currently suffering from burnout, the road back to wellness begins with an empathetic discussion with a program faculty member or impartial third party (e.g., designated institutional officer). Confidential meetings need to ensure the resident's health and safety and to develop a personalized wellness plan specific to the type of burnout, circumstantial or existential. Circumstantial burnout originates from acute, selflimited situations. 83 Helpful interventions may include speaking with a professional therapist, developing strategies for mitigating life or workplace difficulties, creating daily time for self-care, and even providing brief time off from clinical duties. 83 Existential burnout originates from a chronic loss of joy from the practice of medicine itself. Helpful interventions may include speaking with a professional therapist, examining the origins of burnout, developing better relationships with patients and colleagues, and even reshaping the resident's professional identity. 83 Consistent follow-up ensures that the resident is recovering and provides opportunities to identify when other interventions are needed.
It is important to note that while individual interventions to improve wellness show some promise, evidence suggests that intervention programs based on the individual are associated with only small benefits on burnout and should be supplemented by the adoption of organizational approaches. 84

Organization-Level Interventions
In addition to individual interventions, institutional, organizational and departmental wellness committees have been widely recommended as a strategy to address physician, staff, and trainee wellness. 40,85,86 Committees should be composed of residents and faculty members, who meet regularly to assess, analyze, and develop systematic initiatives to improve the clinical learning environment for all. 40 Committee members should contribute to the creation of a wellness program or curriculum. Lefebvre discussed the key components of a resident wellness program, which include creating a safe space; having one-on-one meetings with residents; and designing residency events focused on physical, mental, social, intellectual, and community wellness. 82 Wellness programs should combine both passive (e.g., safe places, lectures, website resources) and active (e.g., meetings, workshops, outings, small group activities, service projects, gym access) strategies. 86 Other components could include strategies to help deal with the wide range of issues that may be encountered during residency, such as stress management, behavioral issues, marriage or family problems, financial troubles, substance abuse, disruptive colleagues, or mental health issues. 85,[87][88] In addition to institutional programs, residency curricula have been shown to be beneficial. 45,86,[90][91][92][93][94][95][96][97] Studies recommend a multicomponent wellness curriculum, including resilience, professionalism, emotional wellness, physician suicide, social wellness, financial wellness, team building, and mindfulness. [89][90]98 Occupational wellness components have also While curricula should be program-driven, most studies recommend the inclusion of out-of-hospital components, including retreats, workshops, and social outings. 86,89,93 Studies of curricula that emphasize mindfulness, resilience training, and stress management have demonstrated improved physician wellness and reduction in burnout scores. [101][102][103] While it seems intuitive that mindfulness and resilience, the ability of an individual to effectively cope with and adapt to adverse situations, would have positive effects on the wellness and burnout of EM residents, it is useful to review the literature on systemwide interventions in these areas. One particular study by Krasner, involving a longitudinal curriculum on mindful communication, noted both short-term and sustained improvements in well-being and attitudes associated with patient-centered care. 104 Another study by West involved a ninemonth curriculum, in which physicians met in small groups on a biweekly basis for discussion groups that incorporated elements of mindfulness, reflection, shared experience and small-group learning. This curriculum improved rates of high DP, which was sustained at 12 months, as well as improvements in empowerment and engagement at work. 101 Evaluation of a stress management and resiliency training (SMART) curricula found improved stress, anxiety, and overall quality-of-life scores among both radiology and internal medicine faculty physician participants. 105,106 Another institutionally implemented resiliency curriculum for palliative care and neonatal providers led to improved compassion sensitivity and burnout scores after completion of the program. 107 Critical care fellow participants in a SMART program intervention felt the training provided them with tools to apply during stressful situations, but did not demonstrate improved burnout scores. 108 Similarly, Maher found that a departmentally-instituted educational program designed to improve surgical resident performance during stressful scenarios showed a trend toward improved performance scoring but no difference in anxiety levels. However, 91% of residents rated the stress training as valuable. [108][109] While residents and fellows consistently report subjective benefit from resiliency training, improvement in burnout scores have not been reliably demonstrated. Resiliency training is not the only intervention that has failed to show an improvement in burnout scores. Studies of stress management workshops and training sessions have also demonstrated no difference in physician burnout rates. 52,64 Similarly, a recent study evaluated burnout scores of EM residents before and after implementation of a corporate wellness intervention, "The Happiness Practice." The resident burnout scores did not improve; in fact, 43% of residents stated that this intervention worsened their overall level of burnout. 110 Despite lack of overwhelming evidence that resilience training programs improve burnout scores in residents, there are several studies that demonstrate the importance of the personal trait of resilience in preventing burnout. 99,101,[111][112] One study assessed the role of resilience in the relationship between burnout and health among critical care professionals and found that resilience was a key component in mitigating burnout syndrome. 113 Another study demonstrated that a resiliencebuilding intervention for physicians improved meaning and work engagement while also reducing DP, with sustained results at 12 months. 101 A well-diversified pool of social resources and interests, together with realistic expectations and good selfknowledge, were found to support sustainable coping in a study of 200 physicians from multiple specialties and career stages. 111 A 2015 survey of 616 ED healthcare professionals demonstrated that an individual's coping style may be a predictor of burnout and compassion fatigue. 112 This study found that task-oriented coping is associated with a decreased risk of burnout in contrast to emotion and avoidance coping styles. 112 The cumulative data strongly suggests that resilience is a burnout mitigating factor; therefore, residency programs should consider making resilience training programs available to residents either on an individual or systemwide level.
Resident wellness can also be optimized by evaluating workplace and workflow interventions. [114][115][116] Workflow interventions include electronic health records (EHR) optimization, improving staff-provider communications, and offloading both clinical and non-clinical tasks that could be performed by other members of the medical team. [116][117] EHR efficiency training and the use of scribes or dictation devices has been shown to decrease stress and burnout in attending physicians. [114][115] Additionally, delegating administrative tasks to non-clinical staff has been shown to improve overall wellness. 115,118 Improved workplace conditions, including optimizing workflow, can lead to overall decreased resident physician stress and burnout.
Resident schedules are often a topic of discussion in the medical education community. Since the ACGME duty-hour changes over the past decade, many studies have evaluated the effect of duty-hour restrictions on patient outcomes and resident wellness across specialties. 64,117,[119][120][121] Studies suggest that working >80 hours per week correlates with higher rates of burnout when compared to working <80 hours per week. 120 Another study found that working >60 hours per week was associated with higher rates of burnout and psychological morbidity. 121 According to multiple meta-analyses, the implementation of the ACGME guidelines for duty hours resulted in an increase in resident wellness and PA, as well as a decrease in EE, DP, and burnout. 64,117,119 Additionally, it has been suggested that implementing protected sleep time, an uninterrupted period of sleep during overnight call, better aligns with circadian physiology and can improve fatigue and prevent burnout. 117 This may be of particular importance during offservice rotations and transitions between rotations.

Evidence-Based Interventions that Promote Resident Wellness
Parsons et al.
Scheduling can also affect the ability to access personal medical care. Resident physicians are significantly less likely to have a PCP than their demographically-similar peers outside of medicine. 122 According to a study by Cedfeldt and colleagues, residents in a department with a personal time policy were more likely to find time to fulfil personal needs. 122 Residents who took personal time off had significantly higher proportions of positive experiences and emotions, lower proportions of negative experiences and emotions, higher satisfaction with their career choice, and less perceived stress. 123 Another recent study looked at implementation of a universal well-being assessment for residents, by scheduling each resident for a mental health evaluation based on the residents' schedule with the ability to opt out. 124 The study found that 93% of residents participated in the program, increasing resident utilization of mental health resources. 124 The residents also felt that the scheduling provided convenience, allowing residents to prioritize their mental health and self-care. 124 In addition to total hours worked, many residents also reported that scheduling directly affects their wellness. 122,125 Lack of control over their schedule can make it challenging to find time with family and friends, increasing burnout. 125 EM residents have been shown to appreciate shift work guidelines that focus on the importance of circadian scheduling (i.e., advancing shift times progressively from day to evening to night) but, if given one option, prefer having the ability to request days off and have a full weekend off. 126 A recent study showed that 93% of EM residency programs allow residents to make schedule requests. 127 Programs should consider having "special requests" days each month to allow residents to attend important life events and to ensure residents have protected time off to attend healthcare appointments. Program leadership should work with residents to identify ways to balance increased resident control of scheduling while ensuring appropriate emergency department coverage. Program leadership should engage residents in the reevaluation of current scheduling/staffing models, especially night shift models. Residency programs should recognize that giving residents more direct control of their schedule, schedule requests and sleeping patterns may help improve overall wellness.

LIMITATIONS
It is important to consider several limitations with respect to this article. While we used multiple methods to identify relevant articles, it is possible that some articles may not have been identified by the current review. However, we used an inclusive search strategy, as well as review of the bibliographies of included articles to identify the most relevant literature. We also included several nationally recognized experts on wellness and engaged in pre-publication peer review by the CORD Best Practices subcommittee and the larger CORD community.
Additionally, article selection was based upon relevance to the specific themes selected. The topic of wellness is extensive, and we selected for review only specific components deemed to be most relevant to the clinician educator. Finally, while 4. Consider developing workflow interventions, such as EHR optimization, and providing increased administrative support (Level 3b, Grade C) 5. Schedules should be optimized to allow residents to request personal time off, to avoid excessive work hours, and to ensure appropriate transitions and circadian rhythms (Level 2b, Grade C)

Best Practice Recommendations for Institutional Interventions
EHR, electronic health records.
preference was given to data directly evaluating wellness assessment and interventions in EM residency programs, the data were limited. Therefore, when data specific to EM residency programs were not available, we used data from other medical residencies and fields as a surrogate.

CONCLUSION
This paper provides an evidence-based review of the literature on wellness in residency education. Strategies for identification, as well as individual and system-level interventions that have shown improvement in resident wellness are discussed along with recommendations for best practices. After reading this paper, readers should have a greater understanding of how to engage in wellness assessment and intervention at their home institution.