Resident Burnout Trends During COVID-19: A Retrospective Observational Study

Abstract

Objective

This study was designed to assess the effects of the coronavirus disease pandemic on job satisfaction (JS) and burnout among surgical sub-specialty residents.

Design

This is a retrospective, observational, survey-based study. We administered a web-based questionnaire to surgical sub-specialty residents, and results were compared to a prior study in 2016. The questionnaire included elements on demographics, JS, burnout, and self-care habits. Basic statistical analyses were used to compare data from 2020 and 2016.

Setting

This study takes place at Robert Wood Johnson University Hospital, a single mid-sized New Jersey-based academic institution.

Participants

This survey was sent to all obstetrics and gynecology and general surgery residents from every postgraduate year based out of our institution. The survey was sent to a total of 50 residents across the 2 programs. Forty total residents responded to the survey, with a response rate of 80%.

Results

JS was significantly higher in 2020 than 2016 (p < 0.001). There were no differences between postgraduate years for 2020 or 2016 in emotional exhaustion (p = 0.29, p = 0.75), personal accomplishment (p = 0.88, p = 0.26), or depersonalization (p = 0.14, p = 0.59) burnout scores. A total of 0% of residents in 2020 worked fewer than 61 hours per week. Residents in 2020 exercised more (40.0% vs 21.6%), had similar alcohol usage (60%), and had similar diets to residents in 2016. Residents in 2020 were less likely to regret their specialty (7.5% vs 21.6%), consider changing residencies (30.0% vs 37.8%), or consider a career change (15.0% vs 45.9%).

Conclusions

JS scores were significantly higher during the coronavirus disease pandemic. The cancellation of elective surgeries led to a lighter workload for surgical residents. Residents were uncertain of their role during the pandemic, however, new stressors encouraged residents to seek alternative methods for personal wellness.

INTRODUCTION

Burnout is defined by the World Health Organization (WHO) as a syndrome characterized by emotional exhaustion (EE), depersonalization (DP), and low sense of personal accomplishment (PA) resulting from workplace stressors that are not well-managed.¹ This syndrome is increasing among medical professionals, and has negative effects on patient and provider, including medical errors, professionalism violations, suicidal ideation, and decreased patient satisfaction with care.2, 3, 4, 5, 6 The recent Coronavirus disease (COVID-19) pandemic has added additional stressors to the daily lives of healthcare workers, exacerbating any pre-existing burnout syndrome, especially among resident physicians, who are already high-risk for burnout.⁷

A study previously done at Robert Wood Johnson Medical School evaluated the severity of burnout among residents at Robert Wood Johnson Medical School. It used a survey with questions pulled from the Maslach Burnout Inventory and the Wellness Council of America (WELCOA) job satisfaction (JS) survey, validated wellness surveys for healthcare workers.^{8 , 9} This study repeats that prior survey in the midst of the COVID-19 pandemic and compares results to those obtained previously in 2016, with a goal of quantifying the severity of surgical sub-specialty resident burnout during these unprecedented times and identifying patterns with respect to demographics and risk factors such as control over one's schedule, satisfaction with personal life, or year in training.

MATERIALS AND METHODS

This is a retrospective, observational, survey-based study at a New Jersey-based large academic institution which was approved by our institutional review board. This study involved a 1-time, 70-item, web-based questionnaire using Qualtrics. The survey was emailed to obstetrics and gynecology (OBGYN) and general surgery residency program coordinators which was distributed to the appropriate residents. A 1-time email reminder was sent 2 weeks after the initial survey distribution. The survey was closed 4 weeks after initial distribution. Participant consent was obtained electronically prior to the initiation of the questionnaire.

This questionnaire was similar in format to a questionnaire used in a prior study at our institution in 2016. This previous study included elements on demographics, JS, resident burnout, and self-care habits. The demographics included information on postgraduate year (PGY), medical school degree, type of residency program, race, religion, age, gender, marital status, children, number of hours per week worked, and career choice regret. The JS questionnaire used combined items from a modified WELCOA JS survey. The burnout questionnaire was a modified Maslach Burnout Inventory—Human Services Survey (MBI-HSS).

The modified JS survey consisted of 29 questions from publicly available WELCOA JS survey. JS questions were scored using yes and no responses and were assigned points where yes = 2 and no = 0. The scores were summed and averaged, then compared to JS data from 2016. The modified MBI-HSS survey was obtained through a leasing agreement and used to assess burnout subscales of EE, PA, and DP. The responses used a scale from low, moderate, and high, and were assigned points where low = 0, moderate = 3, and high = 6.

Data was collected using Qualtrics secure online survey software and was exported to an excel spreadsheet. Basic statistical analysis was performed using the Data Analysis Toolpak in Excel. A student's t-test was used to compare JS survey means between 2020 and 2016. A one-way ANOVA test was used for each individual burnout subscale to determine the difference between PGYs. A p-value of < 0.05 was determined to be statistically significant.

RESULTS

Our survey was sent to all OBGYN and general surgery residents based out of our institution (Table 1 ). The survey was sent to a total of 50 residents across the 2 programs. Forty total residents responded to the survey, with a response rate of 80%. Representation from PGY1 through PGY5, and PGY7 was present in the study. There were zero PGY6 respondents and one PGY7 respondent. There was a lower percentage of American allopathic respondents in 2020 than in 2016 (85% vs 97.3%, respectively), higher percentage of American osteopathic respondents (5.0% vs 0%), and more foreign medical graduate respondents (10.0% vs 2.7%). The majority of our respondents were between the ages of 31 and 33, female, white, atheist or agnostic, single, and had no children. This differs from our 2016 data, in which the majority of respondents were between the ages of 31-33, female, white, other religion, married, and had no children.

TABLE 1.

Participant Demographics

Characteristics		2020 (n = 40*)	(%)	2016 (n = 37)	(%)
Residency year	PGY1	12	30%	8	21.62%
	PGY2	6	15%	7	18.92%
	PGY3	10	25%	11	29.73%
	PGY4	5	12.50%	5	13.51%
	PGY5	2	5%	2	5.41%
	PGY6	0	0%	4	10.81%
	PGY7	1	2.50%	0	0%
Medical school	MD	34	85%	36	97.30%
	DO	2	5%	0	0%
	FMG	4	10%	1	2.70%
Residency program	OBGYN	14	35%	14	37.84%
	General surgery	26	65%	23	62.16%
Race	White	24	60%	18	48.65%
	Asian	7	17.50%	12	32.43%
	Hispanic	0	0%	3	8.11%
	Black	3	7.50%	1	2.70%
	2+ Races	5	12.50%	3	8.11%
Religion	Hindu	5	12.50%	3	8.11%
	Jewish	5	12.50%	1	2.70%
	Catholic	5	12.50%	7	18.92%
	Protestant	4	10%	4	10.81%
	Muslim	1	2.50%	3	8.11%
	Atheist	7	17.50%	4	10.81%
	Agnostic	7	17.50%	5	13.51%
	Other	4	10%	10	27.03%
	Buddhist	0	0%	0	0%
Age	25-27	12	30%	10	27.03%
	28-30	11	27.50%	12	32.43%
	31-33	16	40%	14	37.84%
	34-36	1	2.50%	1	2.70%
	37-39	0	0%	0	0%
	40+	0	0%	0	0%
Gender	Female	22	55%	30	81.08%
	Male	17	42.50%	7	18.92%
	Other	0	0%	-	-
Marital status	Single	16	40%	7	18.92%
	Relationship	8	20%	14	37.84%
	Married	15	37.50%	16	43.24%
	Civil union	0	0%	0	0%
Children	No	36	90%	31	83.78%
	Yes	3	7.50%	6	16.22%
	Expecting	1	2.50%	-	-

^⁎Given elective nature of the survey, some participants omitted parts of their demographic information. This results in some categories not totaling n = 40.

JS was compared between 2020 and 2016 (Fig. 1 ). The average satisfaction score for 2020 was higher than that of 2016 (45.2 vs 35.7, p = 0.0006). On JS, negative responses were a higher percentage in 2020 compared to 2016 for just one prompt: “I know what is expected of me at work.” Negative response percentages on JS were more than double in 2016 for 13 prompts, and more than triple for the following 5 prompts: “I'm engaged in meaningful work,” “I trust our leadership team,” “my leadership team cares about me as a person,” “my opinions count,” and “the residents and faculty are generally supportive.” JS was compared between PGYs showed no regular pattern, with PGY5 residents having the lowest average JS scores in 2020, and PGY4 residents in 2016.

FIGURE 1.

Job satisfaction responses 2020 vs 2016: A comparison of responses to WELCOA job satisfaction survey between 2020 and 2016 for individual survey items.

The average burnout scores for 2020 were 14.8 for EE, 35.8 for PA, 7.0 for DP, and 19.2 overall. The average scores for 2016 were 31.8 for EE, 38.3 for PA, and 13.0 for DP, and 27.7 overall (Fig. 2 ). The 3 burnout scales were compared between PGYs for both 2020 and 2016 and found to have no significant difference between PGYs (2020, EE p = 0.29, PA p = 0.88, and DP p = 0.14) (2016, EE p = 0.75, PA p = 0.26, and DP p = 0.59). For 2020, the highest EE score belonged to PGY5 (25.5) and lowest belonged to PGY1 (9.0). The highest PA score belonged to PGY3 (38.4), and lowest belonged to PGY5 (30.0). The highest DP score belonged to PGY2 (12) and excluding the single response from PGY7 (0), the lowest belonged to PGY1 (3.8). The highest average overall burnout score belonged to PGY2 (22.0) and excluding the single response from PGY7 (15.0), the lowest belonged to the PGY1 (16.4).

FIGURE 2.

Average burnout score comparison: evaluation of burnout score categories between 2020 and 2016. Comparison of burnout score categories between postgraduate year (PGY) residents from 2020.

For 2016, the highest EE score belonged to PGY3 (34.8) and lowest belonged to PGY5 (20.5). The highest PA score belonged to PGY5 (42.5), and lowest belonged to PGY3 (33.5). The highest DP score belonged to PGY3 (16.5) and the lowest belonged to PGY2 (10.4). The highest average overall burnout score belonged to PGY3 and PGY4 (both 28.3) and excluding the single response from PGY6 (28.4), the lowest belonged to the PGY5 (25.3).

The statements with the highest negative response rates in 2020 on the EE scale were “I feel used up at the end of the workday,” and “I feel fatigued when I get up in the morning and have to face another day on the job” for which 32.5% of respondents associated highly. The statement with the highest negative response rate on the PA scale was “I feel energetic,” for which 25.0% of respondents associated lowly. The statement with the highest negative response rate on the DP scale was “I worry that this job is hardening me emotionally,” for which 20.0% of respondents associated highly.

Wellness survey data collected in this study was compared to data collected in 2016 (Table 2 ). In 2020, 0% of participants worked fewer than 61 hours per week compared to 2.7% of participants in 2016. Residents in this study exercised more often, with 40% exercising 3 or more times per week compared to 21.6% in 2016. When asked if residents noted a change in alcohol consumption frequency, the majority of respondents in 2020 reported continued alcohol usage without change in frequency (60%) while the majority in 2016 reported decreased alcohol usage (35.1%). A total of 15% reported an increase in alcohol use in 2020 compared to 8.1% in 2016. The majority of respondents in both 2020 and 2016 reported no tobacco usage (95.7% and 94.6%, respectively). Fruit, vegetable, and fast-food consumption rates between 2020 and 2016 were similar with the majority of respondents eating fruits 2 to 3 times per week, vegetables daily, and fast-food 0 to 1 time per week.

TABLE 2.

Summary of Wellness Characteristics

Characteristics		2020 (n = 40)	(%)	2016 (n = 37)	(%)
Work hours	Less than 40	0	0%	0	0%
	41-60	0	0%	1	2.70%
	61-80	22	55%	18	48.65%
	81-100	18	45%	18	48.65%
Exercise	Never	12	30%	20	54.05%
	1-2x/wk	11	27.50%	9	24.32%
	3-4x/wk	14	35%	8	21.62%
	5-6x/wk	1	2.50%	0	0%
	Daily	1	2.50%	0	0%
Alcohol usage	No	7	17.50%	10	27.03%
	Yes, no change	24	60%	11	29.73%
	Yes, with increase	6	15%	3	8.11%
	Yes, with decrease	3	7.50%	13	35.14%
Tobacco usage	No	39	97.50%	35	94.59%
	Yes, no change	0	0%	1	2.70%
	Yes, with increase	1	2.50%	0	0%
	Yes, with decrease	0	0%	1	2.70%
Eat vegetables	0-1	3	7.50%	4	10.81%
	2-3	13	32.50%	9	24.32%
	4-6	6	15%	9	24.32%
	Daily	18	45%	15	40.54%
Eat fast food	0-1	23	57.50%	22	59.46%
	2-3	11	27.50%	12	32.43%
	4-6	3	7.50%	3	8.11%
	Daily	3	7.50%	0	0%
Eat fruits	0-1	6	15%	5	13.51%
	2-3	17	42.50%	12	32.43%
	4-6	6	15%	11	29.73%
	Daily	11	27.50%	9	24.32%
Regret specialty	No	36	90%	29	78.38%
	Yes	3	7.50%	8	21.62%
Change residency	No	28	70%	23	62.16%
	Yes	12	30%	14	37.84%
Change career	No	34	85%	20	54.05%
	Yes	6	15%	17	45.95%

When asked, “Do you regret choosing your specialty?” A total of 7.5% and 21.6% of residents responded "Yes" in 2020 and 2016, respectively. When asked, “Have you considered changing your residency over the past 6 months?” A total of 30% and 37.8% of residents responded "Yes" in 2020 and 2016, respectively. When asked, “Have you considered changing your career out of medicine over the past 6 months?” A total of 15% and 45.9% of residents responded "Yes" in 2020 and 2016, respectively.

CONCLUSIONS

The major significant finding of this study was the difference in JS scores between 2020 and 2016, with residents finding more satisfaction in their roles during the COVID-19 pandemic. Negative responses were higher in the majority of responses to the 2016 JS questionnaire. However, residents in 2020 more frequently endorse a lack of expectations of their role in the healthcare system during this time. Despite the difference in work responsibilities performed by the residents in 2020 due to the COVID-19 pandemic, our results show that residents were more likely to exercise, and similarly likely to consume fruits, vegetables, and fast food. Residents did not sacrifice wellness for their work. We noted a higher frequency of self-reported increase in alcohol usage in 2020 compared to 2016, which is a coping method for the high stress work environment related to the COVID-19 pandemic. While this does not quantify baseline alcohol consumption, residents self-report an increase in consumption more in 2020 than in 2016.

There are many factors that could have affected the results demonstrated in the study performed. In many institutions across the country, residents from one specialty had been required to cover the increasing demands of patient-loads in other areas of the hospital including the intensive care unit and the emergency department. Being shifted from one department to another can increase the ambiguity of work expectations as well as decrease the self-perceived meaningful work residents are performing. OBGYN residents at our institution were not shifted from their specialty to cover other units in the hospital. This could have served to increase motivation or prevent decreasing morale during the COVID-19 pandemic.

The Accreditation Council for Graduate Medical Education enforces work-hour restrictions for resident physicians. Eighty-hour work-week restrictions have been in place for the duration of both study periods. There were several shifts in demographics between the 2 studies, particularly gender, race, medical degree origin, and marital status that suggest a difference in study populations. When comparing the wellness surveys between 2020 and 2016, work hours were similar. Despite the COVID-19 pandemic placing incredible stress on patient-loads, work hours remained similar between the studies with most residents working over 60 hours. We hypothesized an increase in work hours given the new challenges of COVID-19, but this was not the case. It is intriguing that JS increased despite these new challenges. One explanation for this discrepancy could be the cancellation of elective surgeries during the COVID-19 pandemic. In March 2020, the Governor of New Jersey released an executive order to cease all elective surgical procedures in accordance with the centers for Medicare and Medicaid Services as well as the United States surgeon general.¹⁰ With this, the caseload of surgical subspecialties decreased at our institution.

When residents in 2020 were asked about regret, the rates of regret were low for specialty choice and career in medicine compared to residents in 2016. However, when asked if they had considered changing their residency, regret was much higher. This could be explained by the type of work that residents find meaningful. Specialty and career regret are low which could indicate that despite increased responsibilities and fear during a global pandemic, residents believe themselves to be involved in meaningful work. The increased regret in residency could be more of a reflection on increased stress; residents would consider their own specialty in a different location hopefully more manageable while still involved in the meaningful work of their chosen field. Despite the increased regret in residency choice, regret rates from 2016 were higher in all categories.

The burnout questions which score most negatively in both 2020 and 2016 surveys were those pertaining to physical and emotional fatigue. Residents report not having enough energy from the moment they wake, to interactions on the job, to returning home after work. They also endorse being emotionally drained and the development of a callous attitude. With increased work responsibilities, and questionable role on the healthcare team, residents are a more likely to become emotionally blunted, as demonstrated by the burnout subscale scores.

This study was performed to showcase the changes of surgical subspecialty residents during the COVID-19 pandemic. Another proposed study would be to examine how the medical field changes in response to COVID-19. Giving this survey again in several years, after the last cohort of COVID-19 residents has left would show more information on the reaction of the hospital to the pandemic. Would resident JS fall back to levels pre-COVID? Have changes in overall infrastructure been put in place to continue a rise in JS?

The usage of similar surveys with data available from 2016 allows for an appropriate control to evaluate the difference in wellness during the COVID-19 pandemic. Additionally, there was a high response rate within surgical subspecialties in both 2016 and 2020. One limitation of this study was that the Likert scale used in MBI-HSS for 2020 was out of 3 points, and the Likert scale used in 2016 was out of 7 points. This would have created a difference in comparison between the burnout scores that could not be compared with statistical validity, and thus only averages were compared between 2020 and 2016. Additionally, due to the elective nature of the survey, some participants omitted parts of their demographic information limiting related analyses.

Resident JS scores are significantly higher during the COVID-19 pandemic. In a time of uncertainty, working conditions have changed across the country. Residents feel less certain of their roles, leading to doubt in their residency positions. Despite increasing responsibilities due to the COVID-19 pandemic, residents have made wellness a priority in their lives, finding time for diet and exercise. Residents find meaning in their work, as well as a sense of camaraderie and trust with their coworkers. More efforts must be made to ensure resident wellness and support during a global pandemic and moving onward past these trying times.