Resident Well-Being Before and During the COVID-19 Pandemic Open Access

The COVID-19 pandemic has been challenging for health care workers. Early studies demonstrate COVID-19-related PTSD, depression, and anxiety.^1–4  One survey of residents in New York City during the pandemic identified substantial life stressors and high rates of suicide or self-harm ideation, while a multicenter study of internal medicine residents identified anxiety regarding uncertainty and availability of personal protective equipment (PPE).^5,6 

Although physician burnout is highest during residency,^7,8  the impact of the COVID-19 pandemic on resident burnout is unclear. Resilience, the process of adapting well in the face of adversity, may be protective during the pandemic.⁹  Additionally, the nature of the pandemic highlights loneliness as an important measure to consider due to social distancing and quarantine requirements.¹⁰  It is also unclear whether these traditional measures of well-being and their associations have been similar during the pandemic.¹⁰ 

The objective of this study is to examine the early effects of the COVID-19 pandemic on burnout, resilience, and loneliness levels in residents by comparing measures during the COVID-19 pandemic to baseline. By examining resident burnout, resilience, and loneliness during the pandemic, we can consider the impact of this crisis on resident well-being and identify trends and lessons learned to apply to future challenges and resident well-being initiatives.

The University of Chicago conducted a voluntary anonymous resident well-being survey to identify burnout, resilience, and loneliness.^11–13  The survey was administered pre-COVID-19 in spring 2019 (March–May), fall 2019 (October–December), and post-COVID-19 in spring 2020 (March–May) and fall 2020 (October–December) via MedHub software (www.medhub.com). March 2020 marked the initial surge of COVID-19 cases in Chicago and restrictions on elective surgeries. By fall 2020, full hospital services had returned. Based on programs providing in-person care for COVID-19 patients, we defined frontline specialties as internal medicine, pediatrics, family medicine, general surgery (primary residents for trauma surgery services), emergency medicine, and anesthesiology. For analysis, we also grouped specialties according to Accreditation Council for Graduate Medical Education (ACGME) categories: hospital-based, medical, and surgical.¹⁴ 

We assessed resilience via the 10-item Connor-Davidson Resilience Scale (CD-RISC 10, scale 0 to 40, and a 5-point difference being meaningful).¹²  Burnout was assessed using the Maslach Burnout Inventory (MBI).¹¹  Residents with high scores on the depersonalization (DP) or emotional exhaustion (EE; defined as having DP ≥ 10 or EE ≥ 27) subscales were categorized as having symptoms of burnout as defined by previous studies.^11,15,16  Loneliness was measured by the UCLA Loneliness Scale (a 3-item scale from 3 to 9 with higher scores indicating greater loneliness and a 1-point difference being meaningful).¹³ 

Only complete responses to the CD-RISC 10 survey, UCLA Loneliness Scale, and MBI were included; unpaired t tests were used to compare group means within time periods since data were unlinked. Categorical variables were analyzed using chi-square tests. Survey measures during the COVID-19 pandemic (2020) were compared to survey measures from the same time period in 2019. Analyses were performed using Stata 15.0 (StataCorp LLC, College Station, TX). The study was granted exemption status by the University of Chicago Institutional Review Board.

Overall response rates were: 53% (508 of 964) spring 2019, 55% (538 of 982) fall 2019, 51% (498 of 984) spring 2020, and 57% (563 of 985) fall 2020. The overall response rate for frontline residents in spring 2020 was 62% (269 of 434) and 58% (252 of 433) in fall 2020. Respondent demographics were similar regarding gender, postgraduate year (PGY) of training, and ACGME specialty grouping (Table 1).

There was no significant difference in the proportion of residents identified as having symptoms of burnout via the MBI pre/post-COVID-19 (Table 2). The burnout rates via MBI were similar to the previous year in all ACGME specialty groups (Table 2). Frontline residents had a higher rate of burnout compared to non-frontline residents (59% [122 of 207] vs 38% [97 of 252], P < .001) in the spring 2020 survey and in the fall 2020 survey (59% [141 of 239] vs 46% [125 of 270], P = .004). However, the proportion of frontline trainees with burnout remained stable from the previous year (Table 2).

Mean resilience by CD-RISC 10 in respondents ranged from 30.5 to 31.3 and remained constant before and after the pandemic (Table 2). Resilience scores in the frontline group in the spring 2020 survey were similar to non-frontline trainees (30.2±6.4, n = 211 vs 30.7±6.8, n = 260, respectively, P = .41). Resilience scores in the frontline group were also constant before and after the pandemic.

The mean loneliness scale ranged from 4.4 to 4.7 and remained constant before and after the pandemic (Table 2). Loneliness scores in the frontline group were similar to non-frontline residents (4.7±1.8, n = 202 vs 4.5±1.7, n = 254; P = .23). Loneliness scores in the frontline group were also not different before and after the pandemic (Table 2). Surgical specialty loneliness scores in the fall 2020 survey were lower than the hospital-based group (4.3±1.5, n = 124 vs 5.1±1.8, n = 132; P < .001; Table 2).

During the COVID-19 pandemic we did not observe changes in burnout, loneliness, or resilience measures among our residents. Even among frontline residents, whose burnout rates were higher pre-pandemic, burnout, loneliness, and resilience rates remained stable.

The stability of resident well-being measures is surprising, given the disruption to education, clinical care, and life overall. We serve a patient population disproportionately affected by the COVID-19 pandemic,¹⁷  and Chicago at one time led in number of COVID-19 cases in the United States.¹⁸  As the COVID-19 surge approached, planning at our institution sought to reduce stressors and increase support, and so these findings could be explained by our specific institutional response. Reduction in patient care services and reassignment of residents reduced work hours across all groups. The delayed onset of COVID-19 cases in Chicago¹⁸  gave our institution time to maintain, develop, and implement safety and well-being initiatives. PPE supplies were adequate and ventilator stock sufficient. Mental health supports were expanded; communication from all levels of leadership increased. The institution initiated pandemic response training, well-being sessions, care packages, and free meals for frontline residents. In addition, heightened meaning in work, teamwork, and physician appreciation may have preserved well-being and fostered adaptability.⁶  Continued interpersonal interaction due to health care work may have prevented loneliness. Post-traumatic growth—the process of positive psychological change after adversity—may have played a role in maintaining well-being.¹⁹  Many newly instituted centralized well-being efforts will be retained.

The survey was conducted in a single institution, which may limit external validity. The response rate of 54% raises the concern for selection bias due to non-response. The surveys were anonymous and may include different respondents. Studying personal exposure to COVID-19 or to witnessing the effects of COVID-19 on patients could be illuminating. We additionally did not examine respondent use of our COVID-19 support services and association with well-being measures, though this would be an important area for future study. It is possible we measured well-being too soon to capture the full pandemic effects or that surveys were administered too close together to identify differences.²⁰  Well-being can worsen in response to communal trauma after the crisis has ended, and studies during the Severe Acute Respiratory Syndrome epidemic revealed elevated rates of chronic stress 2 years later.^21,22  Continued longitudinal assessment of burnout, resilience, loneliness, and interventions that bolster post-traumatic growth may further elucidate our findings.¹⁹  It will be important to longitudinally analyze well-being effects of the COVID-19 pandemic and apply lessons learned thereafter.

There were stable rates of burnout, loneliness, and resilience in our GME population early in the COVID-19 pandemic.

The authors would like to thank the GME Resilience Committee, Urban Health Initiative, residents, and fellows at The University of Chicago, and the GME office administrators and staff.