Pediatrician Burnout Before and After the COVID-19 Pandemic

Abstract

Objective:

During the surge of the COVID-19 pandemic, burnout among physicians increased significantly. In the spring of 2023, the COVID national emergency was terminated in the U.S. To investigate whether provider burnout rates have returned to pre-pandemic levels, the current study compared dimensions of burnout among pediatricians pre- and post-pandemic.

Method:

As part of 2 separate behavioral health trainings held at a Midwest academic health center in 2019 and virtually in 2023, data on burnout was collected from 52 pediatricians pre-pandemic and 38 pediatricians post-pandemic. Participants completed an online survey during the trainings and responded to items reflecting 3 dimensions of burnout: emotional exhaustion, depersonalization, and personal accomplishment.

Results:

There were no statistically significant differences in pre- and post-pandemic burnout amongst pediatricians in terms of total scores, number of pediatricians who met the clinical cutoff for each dimension, number of cutoffs met, or number of providers reporting elevated burnout on at least 1 dimension (p > .05 for all comparisons). Participants were 1.77 times more likely to meet the cutoff for emotional exhaustion post-pandemic than pre-pandemic. Over half of providers met this cutoff post-pandemic, compared to only 35% pre-pandemic.

Conclusions:

While post-pandemic rates of burnout among pediatricians appear to be statistically similar to pre-pandemic levels, there appear to be clinically significant differences in emotional exhaustion between groups. With 63% of the post-pandemic group meeting the cutoff score for at least 1 dimension, it is imperative for the healthcare system to consider ways to mitigate burnout.

Burnout has been described as “a syndrome of emotional exhaustion and cynicism that occurs frequently among individuals who do ‘people-work’ of some kind” ¹ (p. 99). It is associated with emotional exhaustion or feeling drained by one’s work, depersonalization or feeling disconnected to one’s work, and a low sense of personal accomplishment, competence, or success in one’s work. ¹ The World Health Organization (WHO) defines burnout as “a syndrome conceptualized as resulting from chronic workplace stress that has not been successfully managed.” ² Burnout is associated with negative outcomes for providers, patients, and the healthcare system. ³ Pediatricians who report burnout have higher rates of mood disorders, substance abuse, and motor vehicle crashes. ³ For patients, provider burnout is linked with decreased satisfaction with care and lower adherence to medical advice. ³ Burnout is also associated with higher administrative costs, provider turnover rates, and medical errors. ³

There is notable variability in the prevalence of burnout among physicians across studies, with estimates ranging from 0% to 80.5%. ⁴ At the onset of the COVID-19 pandemic, there was concern about the negative effects on physician mental health and burnout. ⁵ Existing research has shown mixed results, with some studies finding no change in physician burnout early in the pandemic,^6,7 and others demonstrating significant increases.^{8 -10}

One survey of over 2000 physicians found that significantly more providers reported burnout in 2021 compared to 2020, 2017, 2014, and 2011. ⁸ Between 2020 and 2021, physicians’ average emotional exhaustion scores increased by 38.6% and average depersonalization scores increased by 60.7%. ⁸ Physicians of all specialties surveyed had greater emotional exhaustion scores in 2021 than in 2020. ⁸ Compared to those working in internal medicine, physicians working in emergency medicine, family medicine, and general pediatrics were 4.59, 1.57, and 2.44 times more likely to be burned out, respectively. ⁸ In contrast, pediatric subspecialists were less likely to be burned out compared to those in internal medicine. ⁸

Other studies focused on pediatricians found no change in physician burnout early in the pandemic.^6,7 One study compared burnout among pediatricians in February, April, and September of 2020, defining April 2020 as the surge of the pandemic. ⁶ Burnout rates ranged from 20% to 26%, with no statistically significant differences across these 3 time points. ⁶ Interestingly, they also found that emotional exhaustion scores were lower in April than in February, and that emotional exhaustion did not significantly increase by September 2020. ⁶ Another study found similar results, with no significant differences in burnout among pediatricians pre-pandemic compared to during the pandemic, June to July 2020. ⁷ During the pandemic, higher burnout was associated with greater compassion fatigue, emotional depletion, as well as lower prioritization of self-care and perception that their institution valued their efforts. ⁷

In April 2023, weekly hospital admissions and deaths related to COVID-19 in the United States were at the lowest levels since March 2020. ¹¹ At that time, President Joe Biden signed a resolution to end the COVID-19 national emergency. ¹² Despite the resolution, stress related to working in healthcare over the course of a 3-year pandemic may have residual, long-term effects. The primary aim of the current study is to describe burnout among a group of pediatricians at a Midwest academic health system post-pandemic. The secondary aim is to compare burnout to pre-pandemic burnout among a group of pediatricians at the same Midwest academic health system. To our knowledge, this is the first study to investigate pediatrician burnout after the termination of the U.S. national emergency.

Methods

Pediatricians at Michigan Medicine were invited to participate in 2 trainings, one held in-person in 2019 on the management of adolescent depression and one held virtually in 2023 on the management of anxiety in the pediatric population. As part of larger studies that investigated the effectiveness of the trainings on provider knowledge, perception of feasibility, and comfort levels in managing depression ¹³ and anxiety, data was also collected regarding pediatrician burnout. Both studies were reviewed and deemed exempt by the institution’s internal review board. During the final 10 minutes of the trainings, pediatricians were provided with a link and asked to complete an online questionnaire via Qualtrics survey platform. Using burnout data from both trainings, the current study compares 3 dimensions of burnout before and after the onset and peak of the COVID-19 pandemic. Descriptive statistics are presented to illustrate burnout in both groups. Independent samples t-tests were used to compare mean burnout dimension scores, and chi-square tests were used to compare frequency of physicians meeting clinical cut-offs at both time points.

The Maslach Burnout Inventory Human Services Survey for Medical Professionals (MBI-HSS-MP) was used to measure burnout. ¹ The MBI-HSS-MP is a 22-item validated measure of burnout with 3 subscales: emotional exhaustion, depersonalization, and personal accomplishment.^1,14 Previous research has supported the reliability and validity of the measure.^1,14 Items are rated on a 7-point Likert-scale ranging from 0 (never) to 6 (every day). On the emotional exhaustion and depersonalization subscales, higher scores indicate higher burnout, and on the personal accomplishment subscale, lower scores represent higher burnout. While there is variability regarding suggested cutoff scores for the 3 MBI dimensions, the current study used the most commonly reported cutoffs described in the literature. ⁴ Total scores of at least 27 on emotional exhaustion, of at least 10 on depersonalization, and of no greater than 33 on personal accomplishment were considered to meet the cutoff. ⁴

Results

Participants included pediatricians working in Michigan Medicine’s 9 primary care offices. Fifty-seven pediatricians attended the 2019 pre-pandemic training, and 52 (91%) completed the post-training survey. Forty-seven pediatricians attended the 2023 post-pandemic training, and 38 (81%) completed the post-training survey. The average age and years in practice for the pre-pandemic group have been published as part of a previous study. ¹⁵ For the post-pandemic group, participants were 43.76 years old on average (SD = 8.51), and had been in practice for 14.05 years (SD = 8.34). Additional demographic data for the groups are presented in Table 1. There were no significant differences between groups on age (p = .37), years in practice (p = .56), gender (p = .40), or education (p = 1.0). Significantly more participants endorsed having behavioral health providers at their clinic post-pandemic than pre-pandemic (p < .05).

Table 1.

Participant Characteristics by Group.

Variable	Pre-pandemic group	Post-pandemic group
Gender, n (%)
Female	49 (94)	36 (95)
Male	3 (6)	1 (3)
Other gender identity	0 (0)	1 (3)
Education, n (%)
Medical doctor	51 (98)	37 (97)
Doctor of osteopathic medicine	1 (2)	1 (3)
Clinic includes behavioral health providers, n (%)
Yes	26 (51)	30 (79)
No	25 (49)	8 (21)

Correlations between years in practice and scores on each dimension of burnout for the pre-pandemic group have been previously published. ¹⁵ For the post-pandemic group, higher emotional exhaustion was associated with greater depersonalization; other dimensions were not significantly correlated (Table 2). For both groups, burnout dimension scores were not significantly correlated with years in practice. ¹⁵

Table 2.

Post-pandemic Correlations Between Years in Practice and Burnout Dimensions.

	Emotional exhaustion	Depersonalization	Personal accomplishment
Years in practice	.18	.05	.04
Emotional exhaustion	—	.54*	−.06
Depersonalization		—	−.27
Personal accomplishment			—

^*p < .01.

Prior to the pandemic, 35% of providers met the emotional exhaustion cutoff, 14% met the cutoff for depersonalization, and 14% met the cutoff for low personal accomplishment. ¹⁵ Post-pandemic, 53% of providers met the emotional exhaustion cutoff, 16% met the cutoff for depersonalization, and 18% met the cutoff for low personal accomplishment.

Participants endorsed higher levels of emotional exhaustion in post-pandemic (M = 27.84, SD = 11.56) compared to pre-pandemic (M = 23.19, SD = 10.67). However, this difference was not statistically significant, t(76.04) = −1.95, p = .06, r = .22. While a greater proportion of providers met the cutoff for emotional exhaustion post-pandemic than pre-pandemic, the association between the year of the training and whether participants met the emotional exhaustion cutoff was not statistically significant, χ²(1) = 1.79, p = .18. The odds of participants meeting the emotional exhaustion cutoff was 1.77 (odds ratio [OR] 95% CI 0.70-4.52) times higher post-pandemic than pre-pandemic.

Participants had slightly higher depersonalization total scores post-pandemic (M = 5.95, SD = 5.12) compared to pre-pandemic (M = 5.04, SD = 4.13), but this difference was not significant, t(69.26) = −0.90, p = .37, r = .11. The association between the year of the training and whether participants met the depersonalization cutoff was not significant, χ²(1) = 0.10, p = .76. Participants were 1.20 (OR, 95% CI 0.30-4.63) times more likely to meet the depersonalization cutoff post-pandemic than pre-pandemic.

Participants’ personal accomplishment total scores pre-pandemic (M = 39.54, SD = 5.92) were not significantly different than total scores post-pandemic (M = 39.21, SD = 6.29), t(76.96) = 0.25, p = .80, r = .03. The association between the year of the training and whether participants met the personal accomplishment cutoff was not significant, χ²(1) = 0.02, p = .89. The odds of participants meeting the cutoff were 1.08 (OR, 95% CI 0.31-3.66) times higher post-pandemic than pre-pandemic.

Twenty-four (46%) participants met the cutoff for at least 1 burnout dimension pre-pandemic. Twenty-four (63%) met the cutoff for at least 1 dimension post-pandemic. The association between the year of the training and whether a provider met the criteria for at least 1 dimension of burnout was not significant, χ²(1) = 2.55, p = .11. Participants were 1.98 (OR, 95% CI 0.78-5.16) times more likely to meet at least 1 dimension cutoff post-pandemic than pre-pandemic. Table 3 presents the total number of burnout dimensions met by providers pre- and post-pandemic. There was no significant difference between the total number of cutoffs met pre-pandemic and post-pandemic, t(82.06) = −1.00, p = .32, r = .11.

Table 3.

Number of Burnout Dimensions Met Across Participant Groups.

Number of dimensions met	Pre-pandemic group (n, %)	Post-pandemic group (n, %)
0	28 (54)	14 (37)
1	13 (25)	16 (42)
2	10 (19)	7 (18)
3	1 (2)	1 (3)

Discussion

Burnout is an important concern for pediatricians, with negative impacts on the patient, provider, and healthcare system more broadly. ³ Given the increased work-related stress in the context of the COVID-19 pandemic, it is necessary to consider the long-term effect on providers’ wellbeing. Previous research looking at physician burnout during the pandemic has been somewhat mixed, with differing results across provider types and time points.^{6 -10} The current study adds to the existing literature by comparing burnout in pediatricians pre-pandemic in 2019 and post-pandemic in 2023. There were no statistically significant differences in emotional exhaustion, depersonalization, or personal accomplishment at these 2 timepoints. There was also no statistically significant difference in the number of total dimension cutoffs met or the number of providers who met at least 1 cutoff.

While not statistically significant, an argument could be made that differences between pre- and post-pandemic burnout scores appear to be clinically significant. Pre-pandemic, less than half of providers met the cutoff for at least 1 dimension of burnout, but nearly two-thirds of the post-pandemic group did so. Looking closer at each dimension, there appear to be important differences in emotional exhaustion scores, which is the dimension considered the most appropriate representation of burnout. ¹⁶ The average emotional exhaustion total score post-pandemic was nearly 5 points higher than pre-pandemic, placing the average post-pandemic total score above the clinical cutoff while the average pre-pandemic total score fell below the cutoff.

Despite the termination of the national emergency in the U.S., ¹² COVID-19 remains impactful, with thousands of new cases each day. ¹⁷ The current results suggest that providers continue to experience clinically significant levels of emotional exhaustion post-pandemic. While healthcare organizations may be returning to pre-pandemic policies and procedures, it is important to recognize the ongoing impact of the pandemic on physician wellbeing. Interventions to alleviate burnout are vital to better support providers. A 2023 review found that while organizational-level interventions aimed at reducing workplace stress may be more effective at improving employee wellbeing, such interventions were less common than individual-level interventions, such as relaxation training and fostering a positive mindset. ¹⁸ Another study looking at the influence of a variety of supportive resources and interventions on physician emotional exhaustion during the pandemic suggested that institutions should implement interventions that target physician sense of autonomy, hope, optimism, resilience, and self-efficacy to reduce emotional exhaustion. ¹⁹ A 2020 review of the literature on factors associated with burnout during the pandemic made similar recommendations. ⁹ Specifically, they suggest that organizations implement burnout reduction programs, offer access to professional helplines, and consider support programs for provider families. ⁹ They also recommend the inclusion of stress management skills in medical training and consideration of gender differences in experiences of burnout. ⁹ Future research should continue to investigate interventions that are most beneficial for reducing and preventing physician burnout, particularly in the context of the increased emotional exhaustion post-pandemic.

The study has a number of limitations. First, our sample was relatively small, which limited the analyses. The study was underpowered and thus may not have been able to detect true significance. Based on a post hoc power analysis, the study was underpowered to detect a medium effect. Given this limitation and the apparent clinical significance of some of the current findings, future research with larger samples and a range of provider types should continue to investigate how residual effects of COVID-19 pandemic affect physician wellbeing and mental health. Relatedly, our sample was predominantly female, and data regarding race and ethnicity was not collected pre-pandemic and therefore could not be compared or reported. Future studies with larger and more diverse samples may benefit from investigating whether individual factors such as age, gender, race, ethnicity, or marital status, and professional variables, such as patient profile, hours worked per week, or work setting characteristics are associated with different experiences of burnout post-pandemic. Further, future research should collect qualitative descriptions of providers’ experiences of burnout for a more nuanced understanding of burnout. In addition, because participants were recruited from the same pool of providers, it is possible that a number of pediatricians participated in both the pre- and post-pandemic training surveys. As a result, the groups may not be truly independent. While it would be preferable to perform these analyses with paired samples t-tests, identifying information was not collected from providers and responses could not be matched pre-and post-pandemic. As burnout can be a sensitive and personal topic, ensuring anonymity and confidentiality for providers was paramount. Next, burnout was assessed using the MBI, which has limitations. Some have argued that the MBI depersonalization dimension represents a strategy to cope with burnout rather than a symptom of burnout, and that the low personal accomplishment could result from other concerns unrelated to burnout. ¹⁶ Given these criticisms, emotional exhaustion has been suggested as the most relevant or accurate dimension in considering burnout. ¹⁶ In addition, the clinical cutoffs used for MBI dimensions vary considerably in the literature. ⁴ Additional research is needed to establish consistent cutoff scores and build the empirical support for the 3 MBI dimensions as core aspects of burnout. An additional limitation is that data on burnout was not collected during the peak of the pandemic. Although research has demonstrated increased burnout among physicians during the pandemic, we cannot be sure that this was true for the providers in our sample. Finally, it is possible that the providers experiencing the highest levels of burnout were the ones who opted to not participate in the trainings or did not complete the surveys as a result of the exhaustion and stress they are experiencing. If this is the case, the reported levels of burnout may be lower than is actually present in the broader population.

Conclusion

This study found no statistically significant differences in the emotional exhaustion, depersonalization, or personal accomplishment dimensions of burnout among pediatricians post-pandemic compared to pre-pandemic. However, it is notable that the majority of the sample met the criteria for at least 1 dimension of burnout, and that average emotional exhaustion scores post-pandemic exceeded the clinical cutoff. Given its impact on providers, patients, and the healthcare system, burnout remains an important issue in need of intervention.