“Going through the motions”: A qualitative exploration of the impact of emergency medicine resident burnout on patient care

Abstract

Objectives

Burnout occurs frequently in emergency medicine (EM) residents and has been shown to have a negative impact on patient care. The specific effects of burnout on patient care are less well understood. This study qualitatively explores how burnout may change the way EM residents provide patient care.

Methods

Qualitative data were obtained from a sample of 29 EM residents in four semistructured focus groups across four institutions in the United States in early 2019. Transcripts were coded and organized into major patient care themes.

Results

Residents described many ways in which feelings of burnout negatively impacted patient care. These detrimental effects most often fit into one of four main themes: reduced motivation to care for patients, poor communication with patients, difficult interactions with health care colleagues, and impaired decision making.

Conclusions

According to EM residents, burnout negatively impacts several important aspects of patient care. Resident engagement with clinical care, communication with patients and colleagues, and clinical care may suffer as a result of burnout.

INTRODUCTION

Burnout is an occupational phenomenon characterized by emotional exhaustion, depersonalization, and a sense of low personal accomplishment. ¹ , ² Burnout is common in resident physicians, ³ , ⁴ with those in emergency medicine (EM) reporting rates of up to 76%. ⁵ Organizations such as the Accreditation Council for Graduate Medical Education (ACGME) ⁶ and the National Academy of Medicine have taken steps to underscore the importance of burnout on physicians‐in‐training. Although national organizations have examined burnout through a quantitative lens via survey, ⁷ the qualitative examination of residents' self‐reported experiences with burnout has been relatively limited, with much of the literature focusing on the management of burnout or consisting of focus groups in non‐EM providers. ⁸ , ⁹ , ¹⁰ , ¹¹ , ¹² , ¹³ , ¹⁴ , ¹⁵

There is ample evidence from quantitative research that provider burnout has negative effects on patient safety and outcomes, ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹ but we do not fully understand how or why they occur. Literature that explores how burnout translates into negative behaviors and attitudes among clinicians is limited and findings are found to be inconsistent from study to study, but research suggests that burnout may lead to reduced engagement, a hardened view of patient suffering, and safety issues. ¹³ , ²⁰ , ²¹ , ²² Understanding this aspect of burnout among residents is particularly important because postgraduate training is a critical period during which residents establish the skills and expertise necessary to provide safe and competent medical care. Patient care habits that develop as a result of burnout during training may have lasting effects throughout physicians' careers. To our knowledge, no study has qualitatively examined the relationships between burnout and EM residents' clinical care behaviors and practices. In an attempt to fill this knowledge gap, this study aimed to qualitatively explore the impact of EM resident burnout on their patient care.

METHODS

Study design

We performed a cross‐sectional qualitative focus group study of second‐ to fourth‐year EM residents at four U.S. institutions during early 2019 to examine the impact of burnout on residents' clinical care behaviors and practices. Prior work from the same study examined burnout's impact on EM residents' career plans and educational experiences and has been previously reported. ²³ , ²⁴ Here we focused specifically on burnout's influence on residents' patient care (e.g., burnout's role in communication with patients and colleagues, burnout's contribution to medical errors). Participants' demographic data and burnout scores were generated through an electronic survey while qualitative data were generated through a series of semistructured focus groups. Taking into account team members' assumptions, values, and beliefs, we applied a constructivist perspective and inductive thematic analysis to our qualitative data so that findings and theories would emerge from data rather than preexisting theoretical frameworks. ²⁵ The study was approved by all participating sites' institutional review boards.

Study setting and participants

We enrolled EM residents from a convenience sample of four institutions located in the Northeast, Southeast, Midwest, and West for purposes of geographic diversity. Two of the institutions operated four‐year programs, while the other two operated three‐year programs. All sites were Level I trauma centers with annual emergency department volumes that ranged from 65,000 to 85,000 visits. We recruited participants at each location via email invitations on a first‐response basis and limited each focus group to eight participants to ensure everyone had sufficient opportunities to speak. ²⁶ We excluded postgraduate year 1 (PGY‐1) residents because at the time of study recruitment they had only been in their roles as trainees for less than a year. As such, their perspectives on burnout may still be partly based on their time as medical students. We also excluded residents who participated in the pilot focus group that occurred at one of the four sites, during which study materials and procedures were tested and refined. Focus groups occurred in January and February 2019 and were conducted face‐to‐face at each location in a private room. Each focus group consisted of residents from one residency and lasted 80 to 90 min, of which at least 20 min were specifically dedicated to the subject of burnout and patient care. We provided a $100 Amazon gift card in addition to a list of wellness resources to each participant after completion of the survey and participation in the focus group.

Data collection and processing

In the 2 weeks prior to participating in the focus groups, participants completed a survey consisting of the Maslach Burnout Inventory (MBI, used under license from Mind Garden, Inc.) ²⁷ and demographic information (Appendix S1). As in previous studies, burnout was dichotomized and defined by high scores on the depersonalization (>12) or emotional exhaustion (>26) subscales of the inventory. ²⁸ Quantitative data were collected through an electronic survey administered through Qualtrics.

Qualitative data were generated through a series of in‐person, semistructured focus groups centered on gathering information about participants' experiences with burnout and its role in their patient care. Focus groups rather than interviews were chosen for efficiency and to allow participant interactions that would enrich the data. For each group, the same moderator (DWL) was physically present to promote effective group process through the use of ground rules and a moderator guide (Appendix S2). A second facilitator (CAG, SWN, JJ, TDS) audio‐recorded the session, took field notes, and assisted in maintaining the flow of conversation. Moderators included active faculty at two of the four sites, though none were in formal roles of program leadership (i.e., program director or associate and assistant program director). DWL and TDS are core faculty with expertise in health services research and burnout. CAG, SWN, and JJ are core faculty with experience and leadership roles in undergraduate medical education. To promote psychological safety and encourage frank discussions during focus groups, we informed participants that all conversations were confidential and that no one should share information outside of the study session. In addition, we made clear to participants that they were free to answer or not answer any question that was posed to the group. Finally, we emphasized that only deidentified data from the study would be analyzed and presented.

The moderator guide and group procedures were developed based on group process and focus group principles ²⁶ with attention to the study's purpose. Semistructured questions and probes were developed through a modified nominal group technique ²⁹ using information from published research findings, informed by expert opinion (DWL, TDS), undergraduate and graduate medical educators (CAG, SWN), and the professional experiences of the study team (DWL, TDS, CAG, SWN). We did not provide participants with a specific definition of burnout to not limit participants' exploration and discussion of the subject. Congruent with established focus group methodology, ²⁶ no push toward ensuring consensus was made during each focus group. We reached thematic sufficiency after three focus groups but included data from all four focus groups for analyses.

Analytic methods

We analyzed quantitative data from Qualtrics using SPSS for Windows, v.25.0 (SPSS, Inc.). We used descriptive statistics and confidence intervals to describe the characteristics of the study sample.

We used Dedoose software, v.8.2.14 (SocioCultural Research Consultants), to aid qualitative analysis and data organization. We used the constant comparative approach to conduct an inductive thematic analysis of the study data so that our qualitative analysis would be data‐driven and performed without preconceptions. ³⁰ , ³¹ We also did not analyze our qualitative data with regard to participants' burnout status. We aimed for methodologic rigor and trustworthiness via several strategies. ³² , ³³ We confirmed data authenticity by cross‐checking transcripts and audiotapes; identified negative cases in the data; linked central themes and meanings about burnout across participants and settings to form conclusions; asked clarifying questions during the focus groups to ensure understanding; held regular research team meetings; maintained a master list of codes, themes, and meanings; sought transparency by frequently discussing team members' assumptions, values, and beliefs; and member checked via solicitation of voluntary participant feedback on summaries of analyzed and synthesized data to confirm our interpretations resonated with participants' experiences.

RESULTS

A total of 29 out of 50 eligible EM residents participated in four focus groups that ranged in size from seven to eight participants. Participant characteristics are summarized in Table 1. A total of 62.1% of participants in this study reported high scores on either the depersonalization or emotional exhaustion subscales of the MBI.

TABLE 1.

Characteristics of study participants.

Characteristic
Age (years), median (IQR)	30 (28–31)
Gender identity, n (% of 29)
Male	18 (62.1)
Female	11 (37.9)
Current training level, n (% of 29)
PGY‐2	7 (24.1)
PGY‐3	10 (34.5)
PGY‐4	12 (41.4)
Chief resident role, n (% of 29)	6 (20.7)
Institution, n (% of 29)
Institution 1–Northeast	8 (27.7)
Institution 2–Southeast	7 (24.1)
Institution 3–West	7 (24.1)
Institution 4–Midwest	7 (24.1)
Burnout (dichotomized), n (% of 29)	18 (62.1)
Maslach Burnout Inventory, mean (95% CI)
Emotional exhaustion	25.0 (20.9–29.2)
Depersonalization	15.7 (12.8–18.6)
Personal accomplishment	40.1 (38.2–42.0)

In our qualitative analyses, we noted several major themes in which resident burnout was associated with patient care. These themes were: (1) burnout‐reduced resident engagement and their motivation to care for patients, (2) burnout resulted in poor communication with patients, (3) burnout negatively impacted professionalism with colleagues, and (4) burnout triggered residents to cognitively offload and impaired resident decision making. A summary of these themes is conceptualized in Figure 1.

FIGURE 1.

Concept map summarizing this study's findings on the self‐reported effects of resident burnout on patient care.

Burnout reduced resident engagement and their motivation to care for patients

Residents in our study reported that burnout affected their level of engagement with clinical care. Reduced engagement was described in a variety of ways, ranging from low motivation to evaluate patients to limited concern for patient well‐being. Some residents mentioned a limited desire to complete small patient care tasks. Other residents detailed a decreased psychological capacity to care about patient outcomes, seemingly as a result of the depersonalization and emotional exhaustion.

There's a point that I get to where I don't care … I'm just there. I just want to go home and go to sleep so that one more day will pass. (PGY‐3)

Residents described being focused less on their patients and more on getting through the day. They also reported less motivation to pick up new patients or reevaluate patients when they feel burned out. This desire to minimize time providing bedside patient care was a recurring theme:

If I'm burned out, I'm just like, have the nurse check on them. It's not really how this is supposed to go, but I don't have the emotional capacity to deal with you right now. (PGY‐4)

These feelings were particularly prevalent in specific clinical situations, such as transitions of care, which occur frequently in the workflow of EM clinicians:

I [can] definitely tell the difference [with] how frequently I re‐evaluate patients, especially sign‐outs. [Burnout] affects patient care because if you're feeling it, just getting up out of your chair is tough. (PGY‐3)

Notably, residents reported that decreased clinical engagement made them intellectually less curious and less motivated to think critically about their patients. Patient workups, even for potentially life‐threatening chief complaints, could subsequently be negatively impacted:

Your curiosity is what makes you a good doctor. And so when you [are burned out], you stop becoming a good doctor because you don't care about the description of their chest pain … I think it affects your workup. (PGY‐3)

Burnout resulted in poor communication with patients

In our analyses, residents frequently reported that burnout negatively affected the way they communicated with patients. These changes in communication manifested in decreased ability to express empathy and in unprofessional behavior toward patients. Some residents also described a correlation between their burnout and the quality of their discharge instructions:

I've noticed that my discharge instructions are better if I'm not feeling burned out. I'll put something in the comment section, as opposed to just filling out the form and putting in generic information. (PGY‐4)

Burnout also impaired the communication between EM residents and their patients by making residents less likely to trust patients and to seriously consider important aspects of their clinical history. This subsequently negatively impacted the way these residents worked up their patients:

[When I'm] feeling burned out, I'm more skeptical. I don't necessarily take the patient at their word, and I think that can contribute to workup bias. (PGY‐3)

Residents reported that their ability to communicate empathetically with patients was reduced when they felt burn out. This lack of empathy would at times lead to unprofessional interactions with patients:

The times when I felt the most burnt out, I know I can't give my patients what they deserve in a doctor as far as empathy and kindness. I just can't offer that anymore. (PGY‐3)

I found myself yelling at this guy … and I verbally escalated on a patient, which I don't really ever do. (PGY‐2)

Ultimately, residents consistently described emotional exhaustion and depersonalization, the key characteristics of burnout. Instead of caring for their patients as people, residents viewed them as impersonal medical tasks that needed to be completed and reported a general inability to be present and engaged:

[Feeling burnout led to] losing the emotional capacity to really be present with patients. And just sort of going through the motions. (PGY‐3)

Burnout negatively impacted professionalism with colleagues

In addition to poor communication with patients, residents described that burnout was often associated with difficult interactions or conflict with colleagues. These reports were inclusive of physician colleagues, such as consultants, as well as nursing and other medical staff. In an environment where timely, open, and professional communication with ED staff is essential for high‐quality patient care, residents noted important changes in how they communicated with others when feeling burnout:

I actually yelled at two nurses on a shift, which is very uncharacteristic for me. I feel like it's not just pure exhaustion, but the cumulative effects of [burnout over] the past three to four years … This is what burnout looks like in my life right now. (PGY‐4)

Similar conflicts were reported with regard to consultant interactions. Many residents felt burnout negatively impacted these interactions by changing their willingness to discuss cases with consultants:

[Burnout] affects my ability to work with others, particularly other professionals … My rope is shorter with [consultants] and that probably affects patient care even though I don't recognize it. (PGY‐4)

I don't call consultants. It's just easier. I discharge more people and I don't talk to consultants and that improves my quality of life. (PGY‐4)

Burnout triggered residents to cognitively offload and impaired resident decision making

In addition to their burnout affecting motivation, communication, and collegial interactions, the residents in our study felt that burnout changed their medical decision making and at times led to medical errors. One common theme was that burnout affected residents' abilities to pay attention to detail. For example, residents frequently discussed missing changes in a patient's clinical status or overlooking abnormal findings in a workup:

I think what I noticed in these on‐shift burnout moments is that my attention to detail is incredibly challenged. (PGY‐4)

Burnout [increases my risk of making a mistake] in the sense that I'm more willing to look past abnormalities and gloss over things like unexplained tachycardia. I don't stop and think about what else I could be missing. (PGY‐4)

There will be times when I will have a little bit of downtime, when in theory I could run through labs and recheck some stuff. But I'm just like, I can't. I can't do it right now. (PGY‐4)

Many residents noted that they think less critically about their patients when they were burned out and that this often led to more algorithmic patient care:

I'm less likely to think about patients. And whether you miss things or not, you end up just doing lazy medicine, where you're like, “Well, I'm just going to CT them and then I'm going to admit them.” (PGY‐4)

This perception that burnout led to “lazy” medicine and less critical thinking was prevalent among residents and was viewed as a way for them to cognitively offload when burnout made critical thinking too difficult or burdensome:

When I'm feeling burned out, I recognize I have limited cognitive capacity. And my goal for myself is to unload the cognitive burden. The easiest way for me to do that is to get more tests, so I have to think less about possibilities. (PGY‐3)

In addition to increased testing or algorithmic medicine, the desire to offload additional cognitive or emotional work also led to residents' reduced efforts toward verbal deescalation and quicker decisions to chemically restrain patients. This was in line with other negative changes to patient care and communication described in the preceding themes:

I'm sedating faster … At the end of five shifts in a row, I'm going right to sedation. I'm not even going to [attempt verbal de‐escalation]. (PGY‐4)

Notably, residents recognized that these changes in patient care were not consistent with the standards by which they would like to treat patients:

I think about what I would want done if this were my family member in the ER. And I feel like I often fall short of that care when I'm burned out.

DISCUSSION

The results of this qualitative study reveal that residents feel burnout adversely impacts the way they care for patients. Although not all of the study participants endorsed all of the identified themes, nearly all residents reported some negative change in their clinical practice due to burnout.

First, many residents reported that burnout reduced their engagement with and motivation to provide patient care. Similar decreases with resident engagement and motivation attributed to burnout have been reported with regard to resident attitudes toward graduate medical education. ²⁴ Reduced motivation to engage in patient care often led to less frequent patient reevaluations and a reduced capacity to think critically about patient presentations or results. Particularly concerning were recurrent comments that suggested burnout led to a reduced capacity for residents to care about patient outcomes. These focus groups comments are illustrative of depersonalization, of one of the MBI's main subscales, defined as a tendency to treat others as objects, develop callous emotional responses, and lose the ability to care what happens to others. ³⁴ Many studies have demonstrated the significant negative associations between burnout and quality outcomes. ²⁰ , ²¹ Our study qualitatively illustrated how the relationships between burnout and patient outcomes may manifest at the individual level.

Second, residents experiencing burnout found it more difficult to communicate effectively, compassionately, and professionally with patients. Residents described that they were less likely to trust patient histories, to provide adequate discharge instructions, or to convey empathy toward patients. It is not hard to imagine how these communication challenges could lead to poor provider–patient relationships and low patient satisfaction. ³⁵ , ³⁶ , ³⁷ Interestingly, our data suggested that residents often had insight and were well aware of the connection between their feelings of burnout and a lack of empathy toward patients. Future work could focus on how to take advantage of these clinician moments of insight or clarity and leverage them toward opportunities for burnout interventions.

Third, residents reported that burnout increased their conflicts with colleagues. EM is a team sport, and open and professional communication between all members of the team are key to ensuring safe and high‐quality patient care. Prior work has indicated that emergency physicians in part attributed burnout to the difficult interactions they have with consultants, ³⁸ and there are data that suggest a link between burnout and unprofessional behavior in medical students and physicians of other specialties. ³⁹ , ⁴⁰ Our study suggested that interpersonal conflict with colleagues was both a contributing factor and a manifestation of burnout. Negative interactions with colleagues and burnout may be components of a vicious cycle, in that conflicts with colleagues may lead to burnout, which in turn increases the likelihood of future unprofessional interactions with colleagues. Indeed, one of the MBI's main subscales, emotional exhaustion, is defined in part by difficulty working with others or feelings of stress when having to do so. ³⁴ In this light, unprofessional clinician behavior may be viewed as a symptom of burnout, as opposed to the result of individual bad actors that may prompt punitive action. Institutional mechanisms (e.g., health care incident reporting systems) that health care workers use to report unprofessional conduct by other health care workers may therefore be a useful marker to indicate potential provider burnout and opportunities to address it.

Finally, residents indicated that burnout decreased their attention to detail and caused them to develop potentially unsafe strategies to mitigate cognitive load. Burnout was consistently connected to feelings of cognitive fatigue. Residents described taking mental shortcuts to decrease their cognitive burden due to an inability to think critically and the realization that algorithmic (or “lazy”) medicine may suffice. Residents worried that these changes may lead to medical errors and negative changes in their medical decision making. This is consistent with literature that demonstrated the higher incidence of real and perceived medical errors among physicians with burnout. ⁴¹ , ⁴² Our study's results are also in keeping with previous research that suggested emergency physicians were aware of the effect that burnout had on reducing the quality of care they provided. ⁴³ Indeed, burnout has been found to be associated with higher referral rates and resource use by physicians. ⁴⁴ This suggests that monitoring physician resource use could be an interesting quantitative surrogate indicator or monitor of burnout. Future work would need to confirm the utility of this potential relationship.

Notably, our study did not aim to compare resident perspectives on burnout based on postgraduate year, program, or other demographic characteristics. Future work should aim to quantify and compare burnout experienced by residents of different levels of training and demographic backgrounds.

Overall, this paper describes changes in resident engagement, motivation, communication, professionalism, and clinical care as a result of burnout. These changes are particularly concerning given that residency is a time when physicians learn habits and establish practice patterns that will influence their independent practice for years to come. Burnout's influence on clinician behavior is insidious but real. However, it can be difficult to notice when discharge instructions become short and generic or when CT use and referral rates increase and if these changes are due to burnout or some other reason. Institutions and leaders in graduate medical education should be aware of the myriad ways in which burnout among resident physicians may manifest itself in patient care and consider monitoring burnout via mechanisms other than traditional survey tools and mitigating burnout with system changes and individual support.

LIMITATIONS

There are several limitations to this study. First, our qualitative study included only EM residents and our results may not be generalizable to residents or physicians in other specialties. Second, focus groups were conducted during the winter and this timing may have heightened participants' sense of burnout, though the literature has not demonstrated consistent relationships between burnout and seasonality. Third, the residents who participated in these study groups did so voluntarily, so there is the potential that results are affected by self‐selection bias. Fourth, this study did not intend to demonstrate causality between burnout and changes in patient care. Mentions of causality in our results are indicative of resident perceptions of how burnout impacted their clinical care. Fifth, participants may have been primed to discuss burnout given that they were informed that this was the focus of the study. Sixth, even though none of the moderators were members of participants' program leadership teams, participants may not have been fully forthright in their responses with EM faculty. Finally, the qualitative data collected in this study took place prior to the COVID‐19 pandemic.

CONCLUSIONS

Emergency medicine residents reported that burnout reduced their level of engagement and motivation with clinical work, adversely changed the way they communicated with patients, increased conflict with colleagues, and impaired their medical decision making. These effects are illustrative of important burnout phenomena, including depersonalization and emotional exhaustion. Burnout's influence on patient care can manifest in subtle but important ways.

FUNDING INFORMATION

This work was supported by the Earl P. Charlton Fund Research Award through Tufts University School of Medicine.

CONFLICTS OF INTEREST

The authors declare no potential conflict of interest.

Supporting information

Appendix S1

Appendix S2

Akhavan AR, Strout TD, Germann CA, Nelson SW, Jauregui J, Lu DW. “Going through the motions”: A qualitative exploration of the impact of emergency medicine resident burnout on patient care. AEM Educ Train. 2022;6:e10809. doi: 10.1002/aet2.10809