Burnout is an occupational syndrome characterized by emotional exhaustion, detachment, poor self-worth, and low personal accomplishment [1]. It is common among medical professionals [2,3] and has been associated with worse patient care, increased medical errors, poor physician mental health, and training attrition [4-7]. A particularly vulnerable group of physicians is surgical trainees, as identified in several high-impact publications [2,8,9]. General surgery residents undergo an extended period of training (5 or more years) and are subjected to long work hours and chronic sleep deprivation. Indeed, these are among the reasons most often cited by trainees seriously considering leaving general surgery residency [10]. However, efforts to restrict duty hours have not clearly demonstrated benefit [11], and it has become clear that burnout in general surgery training is a complex phenomenon. Other factors such as discrimination and sexual harassment experienced during training may certainly contribute to both attrition and burnout [8].

The COVID-19 pandemic further exacerbated many of these preexisting issues while introducing new stressors to the training environment [12,13]. General surgery residents were called upon to assist in overwhelmed intensive care units (ICUs) during successive COVID-19 waves. Aside from missed surgical experience and disruption of surgical rotations, the stress of working with patients with COVID-19 was a contributing factor to physician burnout [14-18]. The negative effects of the pandemic have been reflected in the physical health, safety, and mental well-being of residents [19-25]. While there are emerging data on the impact of COVID-19 on resident wellness [12,13], there are minimal data on understanding residents’ experiences of the pandemic and their relationship with the culture of general surgery training more broadly. Given the larger systemic issues within training programs beyond work hours and redeployment, it is important to generate resident-specific data [26].

The aim of this study was to understand general surgery residents’ perceptions of burnout and the interaction with the COVID-19 pandemic experience through qualitative analysis of a wellness survey conducted at a large university-based training program.

This cross-sectional qualitative study of free-text responses among general surgery residents surveyed about burnout and wellness initiatives during the COVID-19 pandemic revealed substantial concerns regarding surgical culture, availability of free time, and the perception of wellness initiatives. Residents identified interconnected issues that resulted in little free time to focus on their well-being, and these problems were exacerbated by mandatory modules and educational sessions that were not perceived as helpful and further impeded time for rest.

Surgery has long been recognized as having a rigid culture with heavy workload, long hours, high call demands, and complex patients [40]. There has been increasing recognition of these issues, but solutions are unclear. A major focus of research and policy has been on duty hour restrictions, which have been widely used in the United States and Europe but less so in Canada. These initiatives were spurred by concerns about patient safety and resident wellness over 20 years ago [41]. With time, evidence has accumulated suggesting that duty hour restrictions do not have a strong impact on resident wellness in surgery [11,42]. Instead, as the results of our study indicate, there are broader problems within surgical culture that contribute to burnout. In 2018, Hu et al [8] conducted a cross-sectional survey of the American general surgery resident population (7409 trainees), with 32% reporting gender-based discrimination, 17% reporting racial-based discrimination, and 30% reporting verbal or physical abuse during training. While our participants did not raise concerns about mistreatment or abuse directly, they pointed to systemic issues such as feeling undervalued and unsupported. Evidence has pointed to deep-rooted cultural issues within the surgical environment that contribute to bullying and burnout. These include hierarchical power structures and problematic individual surgeons with narcissistic tendencies that support these dynamics [43,44]. In a comprehensive examination of this issue, Munro and Phillips [45] quote evidence that surgical trainees are 3 times more likely to experience bullying compared to other trainees. Reasons again include the essential power imbalance that comes from a hierarchical work environment, the public nature of feedback in the operating room, and emulation of bullying behavior itself as a consequence of prolonged training years.

Our study demonstrated that, during the COVID-19 pandemic, some of the barriers to wellness were exacerbated. Although the survey was conducted during the pandemic and residents had experienced at least 2 substantial waves with increased pressures on hospitals and redeployments, participants did not directly relate these events to burnout. Rather, they raised concerns surrounding the cancellation of vacation time and identified issues with receiving entitled lieu days. Literature from early in the pandemic has shown a substantial impact on residents’ educational opportunities and well-being [12-14]. It is possible that our participants were affected by recall bias or they intentionally chose to raise broader issues with respect to burnout rather than focusing on the pandemic specifically. Furthermore, junior residents would have had limited to no prepandemic experience with residency training, thereby not having a comparative experience outside of the COVID-19 pandemic years. Finally, some stressors that are cited in this early pandemic literature may have faded, including poor access to personal protective equipment and fear of infection. In addition to vaccination providing robust protection from serious complications of COVID-19, our resident population was relatively young and healthy and could be expected to experience mild illness. This is an important finding as programs may need to shift away from wellness initiatives directed at pandemic stressors back to preexisting systemic issues in surgery.

There is literature showing that both individual and structural wellness interventions can reduce burnout among physicians [26]. Most of these take the form of mindfulness or stress reduction sessions and small group discussions. Our results suggest that these interventions would be poorly perceived by surgical residents, who felt that these further placed time demands on busy residents and may worsen burnout. Randomized studies of structural and organizational interventions are less common [11,46-48], but these were more widely supported by our participants and may have more success in this population. In a landmark study published in The New England Journal of Medicine, Bilimoria et al [11] randomized 117 general surgery residency programs in the United States to follow strict duty hour restrictions or a more flexible policy that waived rules on shift length and time between shifts. Residents working under the flexible policy were less likely to report patient safety concerns or negative effects on professionalism and education but noted a greater impact on their personal activities. Overall, the trial questioned the evidence base behind restrictive duty hour policies. In a similar trial conducted in Canada, ICU residents were randomized to overnight shifts between 12 and 24 hours in length [48]. Somatic symptoms among residents were more common in the 24-hour shift group, but rates of burnout were similar. There was a tendency toward more preventable errors in patient care occurring in the 12-hour shift group. This literature highlights tension between patient safety and resident wellness and, similar to our results, suggests that institutional policy may be less effective than broader cultural change and sense of belonging at improving burnout [49]. Further research as to the optimal ways to address burnout among trainees is critically needed.

This study has several strengths. Our response rate was high, indicating engagement with the residency program. We used formal qualitative analysis to explore the free-text responses to our survey rather than simple quantitative methods [36,50]. Our assessment of burnout was complemented by formal measures of burnout (MBI-HSS scores), and we have previously described these results [30]. Although data resulting from in-depth interviews may have been richer, given the limitations of conducting research during the pandemic, we were still able to secure a detailed set of responses from a relatively large group of general surgery residents who had experienced considerable disruption in training due to COVID-19. Our study is an important contribution to our understanding of barriers to improving wellness among general surgery residents.

This study has limitations. While the survey was distributed to 82 residents in the program, 51 (62%) completed it, and not all participants chose to offer free-text responses. Thus, our study cannot be considered a random sample of residents in the program. Those with particularly negative or strong opinions or experiences may have been more likely to use the free-text responses and have their views represented in this study. However, we found remarkable consistency in the responses, and the issues raised were not unexpected. As discussed previously, survey responses do not constitute as detailed a data source as interviews [51], although qualitative analyses of surveys are common. This approach is supported by the literature and the methodology of other similar studies, with free-text answers shown to be a rich source of data [52-55]. Inductive analysis is data-driven and an efficient method [37]. Our methodology maps closely to the recommendations of Thomas [56], including a close reading of text, creation of categories and themes, overlapping coding and revisiting uncoded text, and continuing revision and refinement of the category system. We also conducted consistency checks, such as independent parallel coding, checking on the clarity of categories, and research team discussions [56]. However, we still plan to conduct in-depth telephone interviews to further explore the findings generated by this study. We did not provide participants with formal definitions for terms such as “wellness” or “wellness education.” Thus, there may have been heterogeneity in how participants understood these core concepts. These data, particularly with respect to COVID-19, may not be generalizable to other programs. At the time the survey was completed, Toronto had greater experience with COVID-19 in health care settings than other Canadian cities, and the residents likely had more disruption to their training compared to colleagues in less dense centers. Conversely, they were likely less disrupted compared to American programs in similarly large cities—Canada had far less COVID-19–related morbidity on a per capita basis compared to the United States [57]. The residency program at the University of Toronto is also highly academic, with a larger number of research residents and unique work environments that may further limit generalizability to other institutions. A final limitation of our analysis is that we did not conduct member checking, where the conclusions of the study are fed back to the respondents to ensure validity [58].

This qualitative analysis of survey responses among general surgery residents working during the COVID-19 pandemic led to the recognition of 3 important themes related to wellness: the friction between the culture of surgery and wellness prioritization, the impact of the pandemic on working hours, and the difficulties with wellness interventions. These results represent the resident experience at a unique time in health care and may inform future efforts to design interventions to decrease burnout.