Stress Management and Resiliency Training for Healthcare Professionals: a mixed-methods, quality-improvement, cohort study

Abstract

Objective:

To assess the effects of a multi-modal resilience program, the Stress Management and Resiliency Training (SMART) Program, on healthcare professional well-being and job satisfaction.

Methods:

This pilot, mixed-methods, quality improvement, cohort study assessed perceived stress, physical and mental health, job satisfaction, burnout, and value of the curriculum to attendees.

Results:

Participants experienced a significant reduction in perceived stress (p < 0.001) and significant improvements in global mental health (p = 0.001), physical health (p = 0.045), and job satisfaction (p = 0.047). There was no significant improvement in burnout. Qualitative analysis of free text responses revealed appreciation for the skills taught, increased resiliency, and a positive impact on relationships.

Conclusions:

Delivering the SMART Program to healthcare professionals is feasible and may serve as a useful tool for reducing stress and increasing resilience.

Introduction

Burnout is epidemic in healthcare and has negative consequences for clinicians, patients, and healthcare systems¹. The estimated cost of physician burnout in the United States is $4.6 billion annually². While multiple systems-drivers have been identified as contributing to this epidemic^3,4, numerous studies investigating programs that include mind-body skills, cognitive skills, narrative approaches, community building, and/or connecting with meaning and purpose in work have shown promise in reducing healthcare professional stress and burnout^5–8. However, there are few studies examining these tools in combination. Furthermore, there are almost no comparison studies to provide feedback on which components may be most helpful and for whom.

The Stress Management and Resiliency Training, Relaxation Response Resiliency Program (SMART-3RP or SMART Program) developed at the Benson-Henry Institute for Mind Body Medicine is an eight week, multimodal program that teaches a variety of different skills (e.g., mind-body, cognitive, and lifestyle) to reduce the physiologic, cognitive, emotional, and behavioral responses to stress and to enhance resiliency⁹. Resiliency is the ability to respond to stress in a healthy way that allows one to effectively function despite challenging circumstances with minimal psychological or physical cost^9,10. Resilient individuals have an enhanced capacity for dealing with adversity. The SMART Program has been used clinically and in research studies to help patients with a wide range of stress-related health conditions^11–14. Given that the SMART Program combines many of the tools that have been shown to reduce stress in healthcare professionals into one program, we reasoned that the SMART Program may be particularly helpful in this population. Thus, we conducted a pilot study examining the effects of the SMART Program on physicians, nurse practitioners, and research faculty at our academic medical center. We collected both quantitative and qualitative data to better understand the range of effects these tools have on healthcare professionals.

Methods

Study design and setting

We were invited by the Department of Medicine at Massachusetts General Hospital (MGH) to run a stress management program for faculty. We collected pre- and post-program questionnaire data from participants on an anonymous, volunteer basis as a quality improvement project to evaluate the program and help inform future iterations. Collection of data was approved by the Partners Human Research Committee/IRB. A total of three groups ran over three consecutive years during the winter months from November 2015 through February 2018.

The SMART Program

The SMART Program is a multi-modal program developed at the Benson-Henry Institute for Mind Body Medicine that teaches a variety of skills to reduce the physical, cognitive, emotional, and behavioral effects of stress⁹. This program contains 4 main components to build resilience to stress: mind-body skills (different meditation techniques, mini relaxations, and yoga), traditional stress management techniques, healthy lifestyle behaviors (sleep, exercise, nutrition, and social support), and cognitive reappraisal and adaptive coping skills (adapted from cognitive behavioral therapy, acceptance and commitment therapy, and positive psychology). Sessions include a variety of experiential exercises, didactic components, and small group activities with opportunities to debrief in pairs, small groups, and the larger group (see Table 1 for a partial list of skills taught). The SMART Program is usually conducted for 1.5 hours weekly over 8 weeks and has been tested in, and shown to be efficacious among, different patient populations^11–14. For this program, we scheduled sessions 2-3 weeks apart (depending upon holidays) to better accommodate clinicians’ schedules. Sessions were held in the late afternoon/early evening on a weekday in a conference room at the hospital and were co-run by two trained facilitators who are both physicians (MLD and DHM).

Table 1:

Examples of topics covered in the SMART Program Curriculum

Session	Topics/Skills Covered
1	Science of mind-body medicine Breath awareness & single-pointed focus meditations Appreciations Goal setting and tracking weekly practice
2	Body scan Recuperative sleep Stress warning signals exercise Introduction to “mini relaxations”
3	Mindfulness Mindful eating Identifying emotions & physical sensations Social support
4	Yoga Movement and exercise Negative automatic thoughts Thought distortions
5	Insight imagery Adaptive perspectives/cognitive restructuring Problem solving & acceptance Healthy eating behaviors
6	Contemplation practice Relaxation signals Comparing optimism & pessimism Promoting physical activity
7	Loving-kindness meditation Creative expression Empathy Self-compassion
8	Idealized-self meditation Humor and laughter yoga Energy battery Staying resilient

Participants

Participants were physicians, nurse practitioners, and faculty researchers from the Department of Medicine at MGH who volunteered to participate in the program. The program was advertised via email announcements from the department chair at MGH. Group sizes ranged from 15-20 participants each.

Measures

Before and immediately after the program, participants were sent an email with a web link to study questionnaires (SurveyMonkey.com). Participants were asked to create a unique ID to maintain the anonymity of their data while allowing us to link pre- and post-program responses. The following questionnaires were included: Perceived Stress Scale-10¹⁵ (PSS-10, score ranges 0-40), PROMIS Global-10 assessing both mental and physical health¹⁶ (each subscale score ranges from 4-20), 5 questions assessing Global Job Satisfaction from the Physician Worklife Survey¹⁷ (score ranges from 5-25), and single item measures of emotional exhaustion and depersonalization originally derived from the Maslach Burnout Inventory^18,19 (score ranges from 0-6 for each item). During the program, we also asked participants to complete a separate demographic survey which was purposefully not linked to their pre- and post-questionnaire responses to maintain anonymity. In addition, the post-program questionnaire contained Likert scale questions and free-text response questions regarding the relevance of the program to their life and work, the helpfulness of the skills taught, and how the program affected them personally and professionally.

Statistical and qualitative analyses

Descriptive statistics were calculated using SPSS v24 to assess participant demographics. Quantitative results from pre- and post-questionnaire data were compared using paired t-tests, and Cohen’s d was calculated as a measure of effect size.

For the qualitative analysis of free text responses, two researchers (EWN and CEN) not involved with the delivery of the intervention read the responses independently and noted major themes. They discussed their observations with a third researcher (MLD), themes were mapped onto the program model⁹, and the initial code manual drafted. The first two researchers used the draft manual to independently code the free text responses using NVivo software (v12, QSR International) and subsequently compared their codes and discussed areas of confusion or disagreement. Using an iterative process, the code manual was updated, free text independently coded by two researchers, and areas of discrepancy discussed, using a third researcher (MLD), to resolve discrepancies and settle on a final coding scheme.

Results

A total of 59 individuals enrolled in the program (see flow chart in Figure 1). Of these, 56 completed the pre-program questionnaires online. Fifty-four individuals attended at least 1 session after session 2. Within this group, the average number of sessions attended was 5.4 (standard deviation 1.7). A total of 41 individuals completed demographic questionnaires during the program and 42 individuals completed post-program questionnaires. We ended up with a total of 36 matched pre-post questionnaires.

Figure 1:

Participant flow chart.

Demographic characteristics of the participants are presented in Table 2. Participants were predominantly female (73%), white (85%), and physicians (73%). The average length of time in clinical practice was 17 years and the average number of hours worked per week was 55.

Table 2:

Participant Demographics^*

Age	45 (12)
Gender
Female	30
Race
Caucasian	35
Asian	5
African American	1
Hispanic Ethnicity	2
Credentials
MD/DO	30
NP	10
PhD	1
Years in Practice	17 (11)
Hours worked/week	55 (14)

^*N=41. Values in columns are either number of subjects or mean (standard deviation).

From the beginning to the end of the program, participants noted a significant reduction in perceived stress (p < 0.001, Cohen’s d = 0.78) and significant improvements in global mental health (p = 0.001, Cohen’s d = 0.61), physical health (p = 0.045, Cohen’s d = 0.35), and global job satisfaction (p = 0.047, Cohen’s d = 0.34; Table 3). There was not a statistically significant improvement in emotional exhaustion or depersonalization (p = 0.13, Cohen’s d = 0.26 and p = 0.26, Cohen’s d =0.19, respectively).

Table 3:

Pre and Post Program Questionnaire Data^*

	Pre	Post	p	Cohen’s d
Perceived Stress Scale	18.7 (6.4)	12.9 (4.4)	0.000	0.78
Global Mental Health	13.5 (3.2)	15.3 (2.7)	0.001	0.61
Global Physical Health	16.4 (1.5)	16.9 (1.6)	0.045	0.35
Global Job Satisfaction	18.1 (3.0)	19.2 (3.0)	0.047	0.34
Emotional Exhaustion	41.7%	27.8%	0.132	0.26
Depersonalization	19.4%	11.1%	0.257	0.19

^*N=36. Values in the pre and post columns are mean (standard deviation) or percent of total participants

The majority of participants (39/41) agreed that the program was relevant to their life and that the skills taught were helpful. Qualitative analysis of free text responses revealed the emergence of several themes: the power and usefulness of the course overall and specific skills taught, impact on their work with patients, resiliency, ripple effects into other areas of life, appreciation for the sense of community that arose within the groups, and greater awareness of stress and the ability to use tools to reduce stress (Supplementary Table).

Some individuals found all or most of the tools taught in the program helpful:

“Excellent program, very well-tailored to the needs of the participants.”

“Highly recommended. I didn’t want it to end! I could go to this class weekly!!!”

“This program has been truly a gift – and I am deeply grateful for the opportunity to participate….I wish I could take this course over and over again.”

Other participants mentioned specific tools that they found particularly helpful.

“The coping log was very helpful. It gave me insight into my own thought cycle and ways to reframe stressful situations.”

“The most important what SMART taught me – I experienced what mindfulness means and I think this … really changed my everyday life. Up until now as I constantly tried to catch up with everything, it felt like non ending marathon. Now I know how to ‘live in the moment’!”

“I enjoyed the meetings, meditation, and journaling.”

Participants also shared experiences of reduced stress and improved mental health as a result of participating in the program.

“I definitely feel myself coping with stressors that are out of my control better.”

“Meditation has helped me relax, de-stress and I have felt better at work.”

“I am much more aware of my own ability to feel better through the use of the skills I learned in the program.”

“I have also been better about recognizing automatic negative thinking which I think has improved my mood.”

Comments reflecting resiliency as an enhanced capacity for dealing with adversity also emerged.

“I have started a daily practice which is completely new, I have reflected and made changes in my life that make me even happier, and I have a skill set to fall back on when things seem like they may be out of control or overwhelming.”

“Helped with EPIC (continuing frustration), election results, difficult patient demands and personal challenges including constant pain.”

“Am more at ease with difficult situations.”

“I made a medical error during this course and I think that being in this course helped me through that difficult time.”

Many participants also commented on how the course had changed their interactions with patients.

“Being more present/mindful when seeing patients and not focusing on everything else I have to do after I leave the room; more empathy to their illness/struggle.”

“I have tried to be more in the moment and just listen to patients – which I think is helpful.”

“I have shared some of the content with them.”

“I have taught some of my patients the mini techniques and I have (usually) been able to approach my patients from a more compassionate place.”

These benefits of enhanced presence with patients also extended out to colleagues and home life.

“The course gave me many techniques to think carefully when giving feedback to others, to be kind and thoughtful, to pause when navigating conflict in the workplace. I was able to give my practice manager some feedback in a way that I think was more effective. It changed the tone of our interactions and we are now more kind and positive with one another.”

“I’ve been more positive in my daily life with family and in my workplace.”

“The ripple effect extends to coworkers, colleagues, patients, family members, and much of society!!”

Discussion

In this pilot study, physicians, nurse practitioners, and research faculty participating in a manualized, multimodal stress management program experienced significant decreases in perceived stress and improvements in mental health, physical health, and job satisfaction. Participation in this program and completion rates of the pre- and post-program questionnaires (61% of those who initially signed up and 67% of those who did not drop the program) is notable given that there was no incentive for completing the surveys and that clinicians had to carve time out of their schedules (12 hours) to participate. Indeed, many clinicians arrived late or missed some sessions altogether because of clinical or family responsibilities. They had no protected time to participate in these sessions, and despite the less than optimal participation, they nonetheless received significant value from the program.

The reductions in perceived stress from moderately high to minimal and corresponding improvements in mental health are particularly striking in this cohort. Of note, we saw greater decreases in perceived stress and improvements in job satisfaction in this cohort compared to another published multimodal intervention that included 19 hours of training over 9 months²⁰. No significant improvement in perceived stress was noted in that controlled study, however, there were significant decreases in depersonalization, emotional exhaustion, and overall burnout using the full Maslach Burnout Inventory (MBI). Another study involving 18 hours of mindfulness training over approximately 4 weeks found significant decreases in perceived stress (though smaller than in our cohort) as well as depersonalization and emotional exhaustion using the full MBI²¹. Although there was no significant improvement in emotional exhaustion or depersonalization in our cohort, we did not use the full MBI and the single-item measures that we used are known to be relatively insensitive to change over short time frames²². Moreover, one faculty member shared privately that they “downgraded” their responses to the burnout questions on the baseline survey due to concern about who might see the results or if somehow (despite their anonymity) they might be identified.

Our quantitative and qualitative results highlight the benefits that well-designed, multi-modal resiliency programs can offer healthcare professionals. As noted in our qualitative data, while some participants found most of the tools presented valuable, others tended to focus on specific tools (some meditation, others cognitive, etc.). Indeed, based on our experiences, the benefits of the SMART Program (compared to single modality programs) is that faculty are introduced to, and have the opportunity to experiment with, a rich toolbox and can identify the strategies that work best for them, often ones that they may not have necessarily thought about or signed up for as a single course.

Of note, nearly three-quarters of program participants were female. Whether this was due to increased interest in these tools among women compared to men^21,23 or related to higher rates of burnout in female faculty²⁴, or some combination of the two, we do not know. Several male faculty approached us at the beginning of the program and commented that they were “skeptics;” by the end of the program these faculty acknowledged that the skills taught are helpful, they better understood the rationale for such programs, and they were glad that they had participated.

Some commentators have depicted the role of resiliency programs as an attempt to “fix” clinicians rather than addressing a broken system. Others have labeled this approach as “victim blaming”. While such comments likely intend to shift attention and resources to addressing a system in need of significant changes, and to avoid stigmatizing clinicians (both of which we agree with), these comments do not account for stress physiology, the significant benefits that clinicians can gain from skillful practice of these tools, the slow pace of change in healthcare, and the reality that tackling healthcare professional burnout must be a “both and” not “either or” approach⁵. Indeed, the kinds of tools taught in this program can help generate “moral resilience” in the face of ongoing moral injury²⁵.

Given the pilot nature of this study and that completing the questionnaires was voluntary, we tried to minimize participant burden by using only a few short questionnaires, hence why we used a two-item burnout measure rather than the full Maslach Burnout Inventory. We also did not include a formal resiliency measure as there is considerable disagreement in the literature about what resiliency is and how to measure it^26,27. Nonetheless, our quantitative and qualitative data supports that participants not only experienced reduced stress and improved mental health as a result of participating in the program, but that they also developed an enhanced capacity for dealing with adversity. Future studies should also include questions about sleep, physical activity, diet, and other health behaviors to determine whether participation in the program encouraged healthy lifestyle changes.

Limitations of our work include a small sample size from a single institution, that all groups were run by two clinicians, and that there is no long-term follow-up. Nonetheless, this work reflects data from three different cohorts run over three years. Past participants continue to approach us with gratitude for this course’s value. Another limitation is the lack of a control group to control for secular trends. We suspect background stress levels remained high or may have increased for clinicians at our hospital, particularly during the first two cohorts. The first year we ran the program, the hospital was in the midst of switching from a homegrown electronic medical record to a well-known commercially available one. The second year we ran the program overlapped with the 2016 election and 2017 presidential inauguration. Both events were significant sources of stress for many clinicians, and we have good reason to believe that had we measured background stress in clinicians not attending the program it would have gone up during these periods of time^28,29. Additionally, there were no concurrent system-wide interventions to reduce clinician stress during these time periods, though some divisions had begun introducing interventions during the third cohort. Given the relatively low levels of burnout in our cohort at baseline, compared to the institution overall, we suspect that the most burned-out individuals may not have felt that they had time to participate. A final limitation is our response rate (61%). Although it is high compared to many physician surveys, our effect size estimates may be inflated due to missing data.

As a final note, sessions for the patient version of the SMART Program meet weekly. For this pilot study, we held sessions every 2-3 weeks rather than weekly in an attempt to better accommodate clinicians’ schedules. We have decided not to do so in the future. These are learned skills, and in our opinion, it is easier to learn and integrate these skills with weekly sessions/follow-up as opposed to every 2-3 weeks, and we have seen in subsequent groups that offering sessions weekly is feasible.

In summary, an eight-week Stress Management and Resiliency Training (SMART) Program is feasible to deliver to clinicians and shows promise in reducing perceived stress, improving mental and physical health, and job satisfaction. Clinicians felt that the skills they learned were useful and relevant to their lives, and they appreciated the diversity of skills taught. While systemic drivers of stress and burnout in healthcare must be addressed, multimodal programs such as SMART can be instrumental in improving quality of life for clinicians by giving them valuable tools to cope more effectively with their current lives. Given the promising results of this pilot trial, compared to other published studies, a randomized controlled trial further evaluating the efficacy of this program, the duration of its effects, and potential to reduce downstream consequences of burnout is needed. If demonstrated to be beneficial, it will be important for institutions to protect clinicians’ time to learn these skills, a modest investment given the high costs of clinician burnout².