The Feasibility and Impact of an Asynchronous Interprofessional Well-Being Course on Burnout in Health Care Professionals

Abstract

Background

Well-being initiatives are essential components in fostering an engaged workforce and creating an effective health care ecosystem. Health care professional (HCP) burnout is widespread and has worsened since the COVID-19 pandemic. In 2014, with Health Resources and Services Administration funding support, the Andrew Weil Center for Integrative Medicine created an online course for HCP well-being. It was subsequently studied in medical residents and revised in 2020. In this study, we explore the impact of the course across larger systems, as well as the long-term impact on HCPs.

Methods

The Health Care Professional Well-Being course is 4.5 hours of interactive online education that explores personal well-being, promoters and detractors of well-being, and systemic factors that influence the overall impact of well-being in health care systems. Participants were recruited through institutional members of the Academic Consortium for Integrative Medicine and Health and were randomized to either active or waitlist control groups. Assessments were taken pre-course, 1-month post-course, and 6-months post-course in the areas of burnout, compassion, resiliency, and lifestyle behaviors.

Results

Burnout measures of depersonalization and emotional exhaustion showed a significant improvement amongst active participants, sustained for 6 months after the course. However, no significant improvement in either the resiliency or the compassion measurements was noted for the active group. Initially, the active group showed improvement in personal accomplishment; however, both groups showed a decline overall. Most noteworthy, a large number of active participants demonstrated adoption of new health-promoting behavior; 95% incorporated at least 1 new lifestyle behavior learned from the course.

Conclusion

This study of a brief, asynchronous, online well-being course with interprofessional HCPs, demonstrates that the course is associated with improvement in individual burnout measures and can educate HCPs about healthy behaviors and a framework for professional engagement.

Introduction

Burnout in health care professionals (HCPs) was first described in the mid-1970s¹ and recognized to be a critical issue affecting HCPs and the care of their patients. Since that time, numerous studies have documented the incidence of burnout in physicians,² nurses,³ pharmacists⁴, medical students⁵, residents,⁶ and other health care staff,⁷ indicating the breadth of this issue across health care systems. The incidence of reported burnout accelerated rapidly with the outbreak of the recent COVID-19 pandemic, ⁸ further impacting the delivery of optimal health care to patients and also affecting the well-being of individual workers, their families, and health care organizations. One meta-analysis found HCP burnout contributed to poor career engagement, increased staff, increased patient safety concerns, and decreased patient satisfaction.⁹ In addition, this same meta-analysis revealed that burnout also creates a reduction in work productivity, less patient access to health care, more malpractice filings, and increased health care costs.

According to the WHO International Classification of Diseases (ICD-11), burnout is a syndrome resulting from “chronic workspace stress that has not been successfully managed” ¹⁰ and characterized by Maslach et al as emotional exhaustion, depersonalization, and a sense of reduced accomplishment in day-to-day work.¹¹

Various factors, both internally and externally generated, can contribute to the development of burnout in HCPs. Internal factors include having high self-expectations, poor work and home integration, or chronically foregoing one’s own needs for the care of others. External factors might comprise high work demands, lack of resources, or moral injury from the sense of being unable to provide the care patients need.¹² Strategies found to help mitigate burnout include individual practices that promote stress management, mindfulness, and resiliency, as well as activities to engage group socialization and community building. On the organizational level, mitigation strategies of optimizing the electronic health record, respecting work and home integration through mindful scheduling for work hours and meetings, and engaging HCPs in shared decision-making have been helpful.¹³ Limiting excessive workload can also contribute to improved well-being.

As burnout is experienced by health care workers at all levels, it would be prudent to endorse known mitigation strategies for all members of the health care team. Interpersonal collaboration has come to the forefront of health care practice to eliminate the silos of task responsibilities and promote cohesive communication, which improves patient care. Promoting beneficial strategies interprofessionally to prevent or reduce the incidence of HCP burnout can foster a culture of inclusion and respect critical for workplace morale.¹⁴

In 2014, the Andrew Weil Center for Integrative Medicine (AWCIM) received Health Resources and Services Administration funding to create an interprofessional course teaching Integrative Medicine and included a module on well-being for HCPs. The course was shown to be an effective tool for teaching Integrative Medicine and HCP well-being principles to interprofessional learners.¹⁵ The well-being course was then evaluated with trainees and demonstrated improvements in both resiliency and burnout when offered to first-year residents in 15 residency programs within a single health care system.¹⁶ The course in this study is an updated version of the course first created in 2014, with a revised structure and content based on learner feedback and new evidence. It was our hypothesis that providing an asynchronous, web-based, well-being course to interprofessional groups of HCPs in various health care systems would improve aspects of both personal health and professional well-being, such as improved personal resiliency, reduction in professional burnout, and implementation of healthy lifestyle behaviors.

Methods

Study Design and Sample

The AWCIM Health Professional Well-Being (HPW) course was designed as an asynchronous online course to build knowledge and skills around HCP well-being. The 4.5-hour online course teaches the foundations of well-being (sleep, nutrition, exercise, resiliency, and mindfulness), includes interactive instruction on the many detractors and promotors of well-being, and gives the learner an introduction to the complex concepts of burnout and engagement within health care systems. Strategies for building resilience, managing stress, preventing burnout, and developing mindfulness practices are explored.

Institutions affiliated with the Academic Consortium for Integrative Medicine and Health (ACIMH) were invited to enroll all HCPs and trainees in the AWCIM HPW course. The ACIMH is a professional organization of approximately 75 academic medical centers and health systems with expertise in integrative medicine and health. The ACIMH’s purpose is to advance the principles and practices of integrative medicine within academic institutions with a focus on optimal health and healing. Institutions were randomized to active or waitlist control. Ten institutions volunteered to participate in the study. One institution opted to implement a mass enrollment of their workforce, with a total of 3804 enrollees. Except for this institution, the randomization process was done at the institutional level with 4 institutions assigned to the active group and 5 to the control group. The 3804 from the mass enrollment were randomly assigned to separate active and control groups with 1902 in each. The total number of participants was 5204; 2441 were invited to participate in the active group and 2763 to the control group.

Participants completed validated burnout, resiliency, compassion, and lifestyle measures at the start of the course and at 1-month and 6-month follow-up periods to evaluate the immediate and long-term impact of the course. In the course enrollment letter, participants were informed that incentives would be available for the first 400 participants to complete the surveys. These incentives for the course were sent as electronic gift cards. The waitlist control participants received $5, $10, and $20 for completing the baseline, 1-month, and 6-month assessments, respectively. Active participants did not receive an incentive for the baseline assessment as it was a part of the course, but they did receive $10 and $20 for the 1-month and 6-month follow-up assessments. The HPW course uses technology to push email or text alerts (based on learner preference) at 1 and 6 months after completion to remind the participant to complete the assessments. The study was reviewed and approved by the University of Arizona Institutional Review Board.

Measures

Maslach Burnout Inventory (MBI)

The MBI is the leading measure of burnout.¹¹ The 22-item scale measures 3 indicators of burnout: 1) Emotional exhaustion (EE) (9 items) measures feelings of being emotionally overextended and exhausted by one’s work. Scores range from 0–54, with higher scores indicating a greater risk of burnout; 2) Depersonalization (DP) (5 items) measures an unfeeling and impersonal response to patients. Scores range from 0–30, with higher scores indicating a greater risk of burnout; and 3) Personal accomplishment (PA) (8 items) measures feelings of competence and successful achievement in one’s work. Scores range from 0–48, with lower scores indicating a greater risk of burnout.

Connor-Davidson Resiliency Scale (CDR-RISC)

The CDR-RISC is a 10-item scale measuring the ability to cope with adversity, including hardiness and persistence.¹⁷ It reflects the ability to tolerate change, personal problems, illness, pressure, failure, and painful feelings. It also reflects an ability to bounce back from a variety of challenges that can arise in life. Scores range from 10–50, with higher scores indicating greater resilience.

Professional Quality of Life Scale (ProQOL)

The Compassion Satisfaction subscale from ProQOL measures the pleasure you derive from being able to do your work well.¹⁸ For example, you may feel like it is a pleasure to help others through your work. You may feel positive about your colleagues or your ability to contribute to the work setting or even the greater good of society. Scores range from 10 to 50. Higher scores on this scale represent greater satisfaction related to your ability to be an effective caregiver in your job.

Wellness Behavior Survey

This is a 1 question survey asking if the participant had incorporated any wellness behaviors learned in the course into their lives. The questions states: “Based on what you learned in the Health care Provider Well-Being course, did you incorporate any of the following wellness behaviors into your life? Check all that apply. Response choices: Diet, Exercise, Sleep, Meditation, Journaling, Gratitude practice, Mindfulness practice, other, or none.”

Sample Description

Only the participants who completed all 3 surveys at baseline and at 1-month and 6-month follow-ups were included in the analyses in this study, and the final sample included 64 participants for the active group and 108 for the control group. There were some significant differences between the active and control groups in terms of gender, race, educational background, and professional roles. The active group had a higher percentage of females (91.9% vs 78.5%), a higher percentage of White participants (93.2% vs 75%), and a lower percentage of Asian participants than the control group (3.4% vs 19%). Regarding educational background, the active group had a lower percentage of MDs and other physician degrees (24% for active group vs 50.5% for control), and the active group had a higher percentage of master’s or bachelor’s degrees (51.7% vs 38.4%). In terms of professional roles, the active group had a lower percentage of fellows or faculty (33.3% for active group vs 55.1% for control), and the active group had a higher percentage of staff members (59.6% vs 29.6%) (Table 1).

Table 1.

Sample Characteristics

Group			Active		Control
Characteristics		n	Mean (SD)/Frequency (%)	n	Mean (SD)/Frequency (%)
Age		62	43.0 (11.1)	104	45.7 (12.2)
Gender*	Male	62	5† (8.1%)	107	23‡ (21.5%)
	Female		57† (91.9%)		84‡ (78.5%)
Ethnicity	Hispanic	62	2 (3.2%)	106	10 (9.4%)
	Not Hispanic		60 (96.8%)		96 (90.6%)
Race*	White	59	55† (93.2%)	100	75‡ (75.0%)
	Asian		2† (3.4%)		19‡ (19.0%)
	Other		2 (3.4%)		6 (6.0%)
Educational Background	MD, DO, other physician	58	14† (24%)	99	50‡ (50.5%)
	Master’s, PhD or advanced		30† (51.7%)		38‡ (38.4 %)
	Bachelor’s degree		14† (24%)		11‡ (11.1%)
Role*	Student or resident	57	4 (7.0%)	98	15 (15.3%)
	Fellow or faculty		19† (33.3%)		54‡ (55.1%)
	Staff		34† (59.6%)		29‡ (29.6%)

Active total n = 64 and Control total n = 108

^*Indicates significant difference between the active and control groups

^{† and ‡}indicate significant difference between the subgroups.

Abbreviation: SD = standard deviation

Statistical Analysis

Independent samples t-tests and chi-square tests were used to compare the sample characteristics. An independent t test was also performed to examine the baseline differences between the active and control groups. A 1-way analysis of variance (ANOVA) was carried out to explore the relationship between role and outcome measures at baseline. Repeated-measures general linear models were used to investigate the change over time by group interactions on the outcome measures. All the analyses were performed using SPSS Statistics 28.¹⁹

Results

Group Differences in Outcome Measures at Baseline

There was a marginally significant difference between groups (P = .076) for MBI EE. The active group’s baseline EE was lower than the control group’s baseline EE (Table 2).

Table 2.

Group Controls of Outcome Measures at Baseline

Measure	Group	n	Mean	P value
Professional Quality of Life Scale Compassion total pretest	Active	64	42.27	.363
	Control	108	41.91
Connor Davidson Resiliency Scale total pretest	Active	64	39.08	.451
	Control	108	39.19
Maslach Burnout Inventory Personal Accomplishment Scale pretest	Active	64	37.05	.23
	Control	108	36.33
Maslach Burnout Inventory Depersonalization Scale pretest	Active	64	6.44	.447
	Control	108	6.32
Maslach Burnout Inventory Emotional Exhaustion Scale pretest	Active	64	22.66	.076
	Control	108	25.36

Impact of Role on Baseline Outcome Measures

One post hoc comparison showed a statistically significant (P = .044) difference in baseline MBI PA between medical students (n = 5) and other uncategorized HCPs (n = 17). Given the small sample size of these 2 groups, role was not controlled for in subsequent analyses. Despite the difference between the numbers of faculty and staff in the active and control groups, there were no statistically significant differences for faculty or staff for any of the baseline measures.

Change Over Time in Outcome Measures by Group

There were no statistically significant differences between the active and control groups for the ProQOL compassion scale. For the CDR-RISC scale, there was a marginal significance of the time by group interaction. The post hoc comparisons showed that the active group’s baseline resiliency rose at 1-month (P = .027) and 6-month (P = .003) follow-ups, indicating an increase in resiliency that was maintained at 6 months for the active group (Table 3).

Table 3.

Change Over Time on Outcome Measures

Variable	Group	n	Pre-mean	1-month mean	6-month mean	Group by time P value
Professional Quality of Life Scale Compassion	Active	64	42.27	43.19	42.28	.428
	Control	108	41.91	41.95	41.79
Connor Davidson Resiliency Scale*	Active	63	38.97	40.33	40.86	.076
	Control	106	39.31	39.26	39.90
Maslach Burnout Inventory Personal Accomplishment †	Active	64	37.05	38.30	30.09	.091
	Control	108	36.33	35.83
Maslach Burnout Inventory Depersonalization‡	Active	64	6.44	5.19	2.92	.03
	Control	108	6.32	6.84	4.20
Maslach Burnout Inventory Emotional Exhaustion§	Active	64	22.66	18.39	13.27	.006
	Control	108	25.36	24.45	15.72

^*Active group pretest was significantly lower than 1-month (P = .027) and 6-month (P = 0.003).

^†For both groups the 6-month follow-up was significantly lower than pretest and 1-month follow-up (P < .001); however, the difference between 6-month follow-up and pretest was greater for the control group, while the difference between the 6-month and 1-month follow-ups was greater for the active group.

^‡While depersonalization at the 6-month follow-up was significantly lower than pretest and 1-month follow-up for both groups, the difference between 6-month and pretest was greater for the active group (P < .001), and the difference between 6-month and 1-month follow-ups was greater for the control group (P < .001)

^§In the active group, there was a statistically significant decrease in emotional exhaustion between each of the time points (P < .001). In the control group, the 6-month follow-up was lower than the pretest and the 1-month follow-up (P < .001).

For the MBI PA scale, the interaction of the time by group also reached marginal significance. The post hoc comparisons showed that both groups decreased significantly from baseline to 6-month follow-up and from 1-month to 6-month follow-ups. The mean DP at 6-month follow-up was significantly lower than both the baseline and the 1-month follow-up (P < .001). For the MBI EE scale, both groups improved and thus decreased their score in EE over the 6 months (P < .001). Mean EE scores decreased, and thus improved, significantly between all 3 time points (P < .001) for both groups. The post hoc comparisons demonstrated a statistically significant decrease in EE for the active group between all 3 time points, while the control group only had a statistical difference between 1-month and 6-month assessments (P < .001) (Table 3).

Incorporation of Wellness Behaviors

The active group participants were also given a brief 1-question survey with the 1-month and 6-month assessments. Based on the participants who completed all follow-up assessments, sleep and exercise were the top 2 wellness behaviors addressed and incorporated at both 1-month and 6-month follow-ups. There was a very high level of incorporation of new wellness behaviors at 1-month (95.3%) and even higher at 6-months (98.4%) after the course by the active group (Table 4).

Table 4.

Active Group – Incorporation of Wellness Behaviors Into Your Life (n = 64)

Based on what you learned in the Health Care Provider Well-being course, did you incorporate any of the following wellness behaviors into your life? Check all that apply.
1-Month Follow-Up			6-Month Follow-Up
Behavior	n	Percent	Behavior	n	Percent
Sleep	45	70.3%	Sleep	50	78.1%
Exercise	39	60.9%	Exercise	48	75.0%
Mindfulness	39	60.9%	Mindfulness	38	59.4%
Gratitude practice	38	59.4%	Gratitude practice	28	43.8%
Diet	30	46.9%	Diet	42	65.6%
Meditation	25	39.1%	Meditation	26	40.6%
Journaling	8	12.5%	Journaling	9	14.1%
Other, specify*	3	4.7%	Other, specify*	3	4.7%
None yet	3	4.7%	None yet	1	1.6%

^*Other: breath work practice, changing my job and employer, making a job change

Discussion

This concise, asynchronous, online well-being course involving interprofessional health care professionals correlates with improvement in individual burnout measures. The active and control groups were followed over 6 months and showed persistent change. This course teaches health care professionals about healthy behaviors and provides a framework for professional engagement. This intervention holds promise for extensive adoption within a broader organizational approach to health care professional well-being.

At baseline, both groups scored at the same level for most measures of burnout and resilience. The active participants scored slightly lower at baseline MBI EE than the waitlist control group. When we look at the change in MBI EE over time, the marginally significant difference at baseline is unlikely to have impacted the reduction in EE demonstrated at 1 month and sustained over time after completing the course. However, despite being statistically significant, it is not clear that this alone would be impactful on an individual’s personal burnout but contribute as 1 tool in the setting of other resources and interventions.

The active group did not show a significant change in the resiliency or compassion measurement as compared to the waitlist control group. For MBI EE and MBI DP, the active groups showed a significant change at 1 month after the course and a sustained and continued change at 6 months. This sustained change is potentially related to the skills and knowledge gained through the course. The course content is focused on 7 foundational pillars of personal health and approaches to improving professional fulfillment, including mindfulness, empathy, self-compassion, meaning, and purpose. In our previous studies, we have shown that building skills in these areas can improve EE and DP.¹⁶ Although the control group trended toward improvement in DP, the active group showed quantitatively more change, which was also then sustained. This trend was also observed with EE, where the active group showed statistically significant change and persistence of change. The control group showed no improvement at 1 month after the course. All participants trended toward improvement in EE at the 6-month assessment. Despite this trend, the active group showed a greater improvement at both time points in EE.

One explanation for the trends for both groups toward improvement in EE and DP measures relates to the timing of the course implementation. Our assessments took place during the COVID-19 pandemic between fall 2021 and spring 2022 when the Delta variant predominated. ²⁰ All participants are HCPs and thus were likely significantly impacted by the ongoing pandemic. In the spring of 2022, HCPs may have been seeing some impact from the vaccination efforts, an increase in personal protective equipment, an increasing availability of treatment options, and a break in the overall frequency of infections as the country headed into the summer.

For MBI PA, the active group had an initial 1-month improvement compared to the waitlist control group; however, both groups had a significant decrease in this area at the 6-month follow-up. One possible explanation for this drop is the timing of the study during the COVID-19 pandemic. PA could have been impacted by the ongoing sense of futility of care and a sense of moral injury during the pandemic. ²¹ The sense of PA is often tied to professional efficacy whereas the feelings of EE and DP are more impacted by workload and systemic stress. Therefore, despite decreases in patient volumes in the spring of 2022, the ongoing and cumulative impact of working through the pandemic with a narrow locus of control was likely to impact the sense of PA.²²

The active participants completed a survey reporting the incorporation of self-care behaviors in several categories, and more than 95% reported incorporating at least 1 new behavior that they learned from the course at both 1 and 6 months. The incorporation of a new self-care behavior is a very meaningful finding for a 4.5-hour online course to spur participant behavioral change. The rate of wellness behavior incorporation was not assessed with the baseline measures, and it was not assessed with the control group, so we do not know if this was greater than the spontaneous incorporation of behavior change present in the general population, although at 95% it seems likely higher than spontaneous change. In a future study, it would be essential to track behavior changes over time for both groups. Additionally, it is not clear as to whether these behaviors are directly related to the trend toward less burnout, but this will be an important area for future research. As an intervention to help HCPs learn skills that can be applied to their own well-being, this course is a feasible strategy for implementing these skills.

We offered this course through the ACIMH to its members. Of the 10 participating institutions, most reached out to their HCPs and offered this course and enrolled only those with interest. In contrast, 1 institution mass enrolled their entire health care workforce. This resulted in the appearance of very low participation numbers and showed that mass-enrollment without other forms of employee engagement is not a recommended pathway for distributing an optional online course. Additionally, as both groups, waitlisted control and active participants, self-selected into this study, this could impact the results.

This project shows that a course with asynchronous online modules can reach a wide range of HCPs and trainees across a large geographic area to share evidence-based approaches to personal and professional well-being. Although there is a relatively low participation rate due to the mass enrollment, the total number of participants, 172, still represents a significant figure. While the online distribution method allows for widespread dissemination, it will likely be most meaningful when implemented in the context of a broader, institution-specific campaign or approach to HCP well-being. Both personal and professional well-being interventions have the potential to improve not only the well-being of HCPs engaged in the interventions, but also may lead to downstream effects improving the patient experience, patient outcomes, and retention of the health care workforce.

This course provided education on health behaviors linked to improved personal health, including nutrition, physical activity, sleep, and stress management, which most participants used to improve their own health. Additional information was provided on frameworks for considering workplace well-being and engagement. Regardless of impact on burnout, the observed changes in health-promoting behaviors are likely to reduce the development of chronic disease. As burnout is multifactorial and largely driven by workforce factors, we might expect limited impact through health behaviors alone. However, we hypothesize that improvement in health behaviors may increase agency to improve the work environment. Additionally, this could include impacts on work culture, such as improved relationships with colleagues and increased capacity to alter workplace process efficiency. As self-care improves, this may also lead to a virtuous cycle where individuals are more able to create a workplace culture of self-care and collegiality, which in turn makes it easier to improve work systems. Together these may decrease turnover of the workforce, making work even more productive and efficient.

Conclusion

An asynchronous web-based learning module can educate many health care workers in health behaviors and frameworks to improve professional engagement as well as successfully reduce burnout amongst participants. This study showed the impact of a brief, online module with an interprofessional group of HCPs. The active group showed improvements in burnout at 1 and 6 months, and 95% of the participants incorporated a new healthy behavior learned from the course. The study did show limitations of a large-scale intervention, including the impact of mass-enrollment of HCPs without local context resulting in lower participation rates. Overall, this study demonstrates that as an integrated part of a larger institutional initiative, there is broad applicability for a brief web-based course for HCPs that addresses both personal and professional well-being.