Feasibility and effectiveness of the mindfulness-based stress reduction programs on relieving burnout of healthcare providers during the COVID-19 pandemic: a pilot randomized controlled trial in China

Abstract

Background

The 2019 coronavirus disease (COVID-19) pandemic brings critical health problems to workers in many occupations, particularly healthcare providers. The aim of the study was to examine the feasibility and effectiveness of the mindfulness-based stress reduction (MBSR) program on relieving burnout of healthcare providers during the pandemic. The roles of positive and negative emotion as well as emotion regulation strategy in the intervention effects were also investigated.

Methods

A sample of 112 healthcare providers in China were recruited and randomly assigned to the MBSR (n = 56) or the control condition (n = 56). Measures were collected at pre-intervention, mid-intervention, and post-intervention, assessing mindfulness level, emotion regulation strategy, positive and negative emotion, and burnout.

Results

The MBSR program showed acceptable feasibility. Compared to the control group, healthcare providers in the MBSR group showed significant increase in personal accomplishment and decrease in emotional exhaustion after the intervention. No significant difference was detected on the dimension of depersonalization. Results of mediation analyses implied that cognitive reappraisal and positive affect partially mediated the intervention effects on personal accomplishment.

Conclusions

The study provided preliminary evidence that the MBSR programs might be effective in reducing healthcare providers’ burnout, even during the pandemic. Cognitive reappraisal and positive emotion might be important mechanisms of how the training took effect.

Introduction

The epidemic of coronavirus disease 2019 (COVID-19) was first reported in Wuhan, China in December, 2019 and is now rapidly spreading around the world [6]. In addition to physical damage, COVID-19 has also caused unbearable psychological pressure to people worldwide, especially healthcare providers, who represent a high-risk group of infection and are more vulnerable to burnout [66]. Previous studies have identified that healthcare providers encounter multifaceted challenges, encompassing the management of isolation, fear, and heightened anxiety; adaptation to evolving healthcare practices and policies; addressing the emotional and physical needs of patients and their families, as well as navigating workplace safety concerns [45]. Studies have also reported that burnout is highly frequent in this population and the prevalence can be high up to 40% ~ 63% [14, 16, 64].

Burnout is a psychological syndrome due to prolong exposure to emotional and interpersonal stressors on the job [43]. The well-established conceptual model developed by Maslach defined burnout into three dimensions: emotional exhaustion, depersonalization, and personal accomplishment [43]. Emotional exhaustion refers to someone is overwhelmed by stress, and has depleted his or her emotional or physical resources. Depersonalization is manifested as someone tends to take a detached and impersonal response to job demands, being indifferent towards interpersonal contact. Personal accomplishment is represented by a devalued self-evaluation, feeling a lack of achievement and productivity at work.

Several studies have found that burnout is linked with a number of physical and mental problems, including coronary heart disease [59], musculoskeletal pain [1], insomnia [60], depression [24], and anxiety [34]. In addition to the severe impact on the well-being of individuals, the potential occupational influences of burnout can also not be underestimated. Burnout is found to be associated with lower care quality [47] and higher medical errors [52]. Job dissatisfaction [58] and turnover intention [4] are also associated with burnout in healthcare providers.

In view of the consequences, approaches targeting at reducing burnout and promoting well-being of healthcare providers are recommended, especially during the pandemic. Among the most frequently adopted and empirically validated interventions, mindfulness-based interventions (MBIs) show promise with regard to this goal. The MBIs provide a diverse range of methodological approaches that can be tailored to accommodate various needs and preferences, including structured programs, personalized sessions, or digital platforms. Among the most widely recognized structured programs within the realm of MBIs are mindfulness-based stress reduction (MBSR) and mindfulness-based cognitive therapy (MBCT).

The training program is usually a structured group program that focuses upon the progressive acquisition of mindfulness awareness [21]. The program is traditionally conducted as an 8–10-week course, in which a group of participants up to 30 meet weekly for practice and discussion [2]. A typical week session is usually 2–2.5 h, and there is an all-day session held around the sixth week [2]. Each session covers a particular topic with several mindfulness meditation skills being taught, such as body scan, sitting meditation, and hatha yoga [2, 21]. Participants are instructed to carry out mindfulness meditation skills outside group meeting for 6 days per week with at least 45 min per day [2]. In addition to regular and formal practice, participants are also encouraged to apply mindfulness to situation during ordinary activities in everyday life, like eating, standing, and walking [2, 21].

A variety of research of MBIs have been conducted in clinical and non-clinical setting to test the effectiveness and the results showed promising effect [21, 31, 32]. Overall, these studies have found that MBIs can effectively reduce depression, anxiety, and stress [31]. In addition, MBIs have been found to improve empathy, forgiveness, self-compassion, satisfaction with life, and quality of life [31, 32]. These results indicated that it may be of benefit for a broad range of individuals to cope with clinical and non-clinical problems [21].

Despite these promising findings, two issues regarding the delivery of MBIs for burnout among healthcare providers need to be further addressed. One issue being discussed is whether MBIs are beneficial tools for burnout among healthcare providers. Extensive research has been conducted but mixed results were reported. Goodman [18] found that a continuing education course based on mindfulness-based stress reduction (MBSR) could significantly decrease burnout and improve mental well-being for healthcare providers. Research exploring the effects of MBSR on nurses also found significantly more reduced scores on burnout for treatment group participants than wait-list controls [7]. However, a study [36] adopting MBSR on surgery residents also suggested no intervention influence on burnout as little difference was found between the MBSR and control participants. Another study [61] conducted in residents also did not support the effectiveness of MBSR for reducing burnout. These mixed findings call for more research in exploring the effectiveness of MBIs in the healthcare settings. In addition, limited research evaluating the efficacy of MBSR on burnout have been conducted in China, it was of crucial importance to add evidence to this field.

In addition to test the effectiveness of MBIs of alleviating burnout in Chinese healthcare providers, the underlying mechanisms of this intervention strategy remain unknown. One possible pathway linking mindfulness and burnout is established from the cognitive–affective perspective, where emotion regulation, as well as positive and negative affect (PNA) are emphasized and should be well addressed. In the process model proposed by Gross [20], the most commonly emotion regulation strategies people use in everyday life are cognitive reappraisal and expressive suppression. Cognitive reappraisal refers to a cognitive change by reconstructing an emotion-eliciting situation as to change its emotional impact; while expressive suppression refers to a response modulation by inhibiting ongoing emotion-expressive behavior. On the affective aspect, positive and negative affect are two dominant affect dimensions. Positive affect (PA) refers to a person’s feelings of energetic, enthusiastic, active, concentrated, and engaged, while negative affect (NA) denotes to a sense of fear, sad, anger, disgust, and guilt [9, 63].

Previous studies indicated that emotion regulation can be enhanced by cultivating mindfulness [19, 22, 57]. These studies suggested that the mindfulness can elicit mindful emotion regulation strategies, which can thus increase the capacity for effective emotion regulation. On the other hand, several studies have pointed out that emotion regulation has a close relationship with burnout [26, 35]. These studies showed that the capacity of adopting self-modulate emotion regulation strategies can help facilitate appropriate emotions and decrease undesirable emotions, thus providing supporting evidence of emotion regulation in the reduction of burnout. In view of the relationships between mindfulness, emotion regulation, and burnout, studies are recommended to deeply investigate the associations linking these three factors.

Several studies [11, 17, 27, 49,50,51] have also reported the relationships of mindfulness and PNA and provided promising results for further investigation. In all, these studies revealed a correlation of mindfulness with PNA, and suggested that the implement of MBIs could promote PA and reduce NA. Meanwhile, PNA may have impact on one's well-being, including burnout. Research have found negative relationship linking PA and burnout and positive association between NA and burnout, indicating the potential role of PNA as risk factors in burnout prediction [23, 44]. Despite the associations verified between mindfulness, PNA, and burnout, more research is called for in exploration of further casual relationships of these variables.

This study aims to explore the feasibility, effectiveness as well as the underlying mechanisms of MBSR on burnout among healthcare providers in China during the COVID-19 pandemic. It is hypothesized that the MBSR will show good feasibility in the studied population, and compared to the control condition, participants receiving MBSR will experience decreases in burnout. Moreover, we hypothesize that the reduction of burnout is facilitated by emotion regulation and PNA.

Methods

Participants and procedure

Participants were recruited from two large urban hospitals located in eastern China in 2022, namely a general hospital and a stomatological hospital. The selection of the two hospitals was based on convenience sampling, primarily due to the pre-existing research collaborations between the hospitals and the research group. Both hospitals have approximately 1000 staff members. The participants consisted of healthcare providers who directly interacted with patients within their respective services.

With approval of the administration boards and ethics committees, information regarding the study was disseminated to healthcare providers through WeChat, which is widely used for communication in China. Interested individuals were subsequently invited to join a dedicated WeChat group specifically created for this study, where they completed an eligibility survey. Eligibility for participation was determined based on meeting the following criteria: 1) active engagement in direct patient care within the past 3 months; 2) expressed willingness to participate and provide written informed consent; 3) abstention from using any psychotropic medications within the past 3 months; and 4) non-enrollment in any other comparable training programs concurrently.

A total of 128 healthcare providers responded to the study advertisement, but 16 declined participation and were subsequently excluded. The remaining participants were randomly assigned to either the mindfulness training condition (n = 56) or the waitlist control condition (n = 56) using computer-generated random numbers. The first author conducted the randomization process without blinding participants to their group assignment. Two participants assigned to the mindfulness training condition dropped out before the intervention due to time conflict. Those randomized to the mindfulness training condition completed the 8-week mindfulness training program in spring and summer of 2022; those assigned to the waitlist control condition would receive the same program in autumn and winter of 2022.

Participants were required to complete assessments three times in both conditions: pre-intervention (T1), mid-intervention (T2), and post-intervention (T3). Participants in the intervention group also filled a questionnaire designed for the purposes to assess the at-home practice and the acceptability and feasibility of the program. The data collection was carried out by two research assistants who were unaware of the group assignment and had previously undergone 1-day training for implementing the assessment. Of the participants in the intervention group, 2 rejected to complete mid-intervention questionnaires, 8 rejected to return post-intervention questionnaires, while the number in the control group were 10 and 14 at mid-intervention and post-intervention. The flow diagram is shown in Fig. 1.

Fig. 1

Flow diagram of intervention

All participants provided written informed consent before taking part in this study. Participation in the study was voluntary and no monetary compensation was provided. The study was approved by the Research Ethics Committee of the Department of Psychology and Behavioral Sciences, Zhejiang University (Approval number: 2021[027]).

Intervention

The program delivered was an 8-week mindfulness training program based on the mindfulness-based stress reduction (MBSR) by Kabat-Zinn and his colleagues [30], which feasibility and effectiveness in the young adult population has been reported in our previous research [15]. The length of the program was adapted to healthcare providers’ working schedule. Each session introduced a key theme and one to two different practice. The eight sessions focused on the following themes: (1) simple awareness, (2) attention and the brain; (3) dealing with thoughts; (4) stress: responding vs. reacting; (5) dealing with difficult emotions or physical pain; (6) mindfulness and communication; (7) mindfulness and compassion; and (8) conclusion. Several mindfulness exercises were taught during the sessions, including body scan, sitting mediation, mindful yoga, and lovingkindness meditation. Participants were given video tapes and were encouraged to practice regularly outside the session for 6 days per week with at least 30 min per day. In addition, informal practice was also advised to participants’ everyday life, including mindful eating, sitting, standing, and walking.

The program was delivered by the first author, who has more than 2 years of personal practice and training delivery experiences of mindfulness in college students and hospital nurses, with the supervision of the second author, who has received mindfulness training from the Harvard Medical School and has more than 10 years of personal practice and training delivery experiences.

Measures

Demographic characteristics

Demographic characteristics were assessed only at baseline. The information collected included age, gender, educational level, major, marital status, religious belief, hospital type, hospital level, post, work experience, and previous mindfulness experience.

Five facet mindfulness questionnaire

Mindfulness was assessed with the Five Facet Mindfulness Questionnaire (FFMQ; [3]. The FFMQ was a 39-item questionnaire measuring five dimensions of mindfulness: Observing (8 items, example is “I pay attention to sensations, such as the wind in my hair or sun on my face”), describing (8 items, example is “I can easily put my beliefs, opinions, and expectations into words”), acting with awareness (8 items, example is “I find it difficult to stay focused on what’s happening in the present”), nonjudging of inner experience (8 items, example is “I tell myself I shouldn’t be feeling the way I’m feeling”), and nonreactivity to inner experience (7 items, example is “I watch my feelings without getting lost in them”). Each item was rated on a 5-point Likert scale, ranging from 1 (never or very rarely true) to 5 (very often or always true). The current study adopted the Chinese version of MMFQ developed by Deng and her colleagues [10]. The alpha reliability coefficient of the scale is 0.825.

Emotion regulation questionnaire

Emotion regulation was measured adopting Emotion Regulation Questionnaire (ERQ; [20]. It was a 10-item self-report measure assessing the habitual use of two specific emotion regulation strategies: cognitive reappraisal (6 items, example is “When I want to feel more positive emotion (such as joy or amusement), I change what I’m thinking about”) and expressive suppression (4 items, example is “I keep my emotions to myself”). Items were rated using a 5-point Likert scale from 1 (strongly disagree) to 5 (strongly agree). The scores of two subscales were summed separately. The ERQ has demonstrated adequate psychometric properties in prior research [20]. The Chinese version of ERQ was developed by Wang et al. [62] and was used in this study. In the present study, the internal consistency of Cronbach’s alphas was 0.662 for the full scale, with 0.807 for the subscale of cognitive reappraisal and 0.743 for the expressive suppression subscale.

Positive and negative affect schedule

Positive and negative emotions were measured by the Positive and Negative Affect Schedule (PANAS; [63]. The scale consisted of 20 items describing different feelings and emotions, and was identified into two scales: Positive affect (10 items, example are "interested", "excited") and negative affect (10 items, example are "distressed", “upset”). All items were rated from 1 (very slightly or not at all) to 5 (extremely) to indicate the extent to which participants have felt a particular emotion during the past week. The Chinese version of PANAS was translated by Huang's team [25] and was conducted in this study. The Cronbach’s alphas were 0.875 for PA subscale and 0.918 for NA subscale, and 0.832 for the whole scale.

Maslach burnout inventory

Burnout was evaluated using Maslach burnout inventory–human service survey (MBI–HSS; [42]. MBI–HSS was a 22-item questionnaire measuring the frequency of symptoms in three domains: emotional exhaustion (9 items, example is “I feel emotionally drained from my work”), depersonalization (5 items, example is “I feel I treat some recipients as if they were impersonal objects”), and personal accomplishment (8 items, example is “I can easily understand who my recipients feel about things”). MBI was generally considered as the gold standard tool for measuring burnout, with higher scores in emotional exhaustion and depersonalization and lower scores in personal accomplishment indicating burnout [43]. Items were assessed on 7-point Likert scale, from 0 (never) to 6 (everyday). The Chinese version developed by Li and his colleagues [38] was adopted in current study. The reliability and validity of the MBI–HSS have been well established in previous research [39, 40]. In current research, the Cronbach’s alphas were 0.890, 0.784, 0.835 for three subscales, respectively, and 0.801 for the whole scale.

Program acceptability and feasibility measures

At the end of the training program, healthcare providers completed a mindfulness training program evaluation survey. Post-program mindfulness training survey were collected from 46 of 56 (83.9%) healthcare providers who were in the mindfulness training. Survey questions were used to assess the acceptability of the program in terms of healthcare providers' perceptions of (a) the quality of the mindfulness training program they have received (1 = bad, 4 = excellent); (b) whether they get the mindfulness training program they were hoping for (1 = not at all, 4 = certainly yes); (c) the degree the mindfulness training program meets their needs (1 = not at all, 4 = nearly all); (d) Whether they would advise their friends to choose this program if their friends need the same mindfulness training program (1 = certainly no, 4 = certainly yes); (e) the satisfaction for the level of help they received (1 = very dissatisfied, 4 = very satisfied); (f) whether the mindfulness training program helps them solve some of their problems (1 = no, they make things worse, 4 = yes, they provide a lot of help); (g) the satisfaction the mindfulness training program they received in general (1 = very dissatisfied, 4 = very satisfied); and (h) whether they would come back to the mindfulness training program if they need help again (1 = certainly no, 4 = certainly yes).

Measures of program feasibility included (a) healthcare providers' self-reported days of weekly practice and (b) healthcare providers' weekly attendance at sessions.

Statistical analyses

A priori power analyses were conducted using G*Power software Version 3.1 to determine the sample size. A total of 86 participants were required to achieve a statistical power of 0.8 for detecting significant interaction effects between condition and time, with a two-tailed α level of 0.05 and an effect size of 0.25 [considered medium based on previous studies (Johannes C [12, 13, 33],Suleiman‐Martos et al., 2020)], in relation to burnout outcomes. The sample size of 122 in this study conforms to the criteria.

The statistical procedures were performed with SPSS version 26.0. Differences at baseline were first examined between the intervention and control groups on all demographic and clinical characteristics. A series of 2 (condition) by 2 (time) repeated measures analysis of variance (ANOVA) were conducted to explore the effect of potential outcomes, while considering covariates such as age, type of healthcare profession, and work experience that differed between the intervention and control groups. Missing data were imputed using the last-observation-carried-forward method as the intention-to-treat (ITT) approach was adopted in current study.

Partial eta squares (η²) was calculated for effect sizes estimates for the ANOVAs, with 0.01 indicating a small effect size, 0.06 a medium effect size, and 0.14 a large effect size [56]. Cohen’s d was calculated for effect sizes estimates for the paired samples t tests, with 0.2 indicating a small effect, 0.5 a medium effect and 0.8 a large effect [8].

We then explored whether or not the effect of mindfulness training program on the reduction of burnout was mediated by emotion regulation and affect. No mediational analyses were conducted for expressive suppression, negative affect, and depersonalization, given that the changes of these outcomes were not significant at post-intervention. The mediation analysis satisfied the key statistical assumptions, including linearity, homoscedasticity, absence of multivariate outliers, normality of residuals, independence, and collinearity diagnostics as indicated by the SPSS outputs. Four models were finally tested using bootstrapping model via SPSS PROCESS macro, which was developed by Preacher and Hayes [48]:

1) Mindfulness (T1) → Cognitive reappraisal (T2) → Emotional exhaustion (T3).

2) Mindfulness (T1) → Positive affect (T2) → Emotional exhaustion (T3).

3) Mindfulness (T1) → Cognitive reappraisal (T2) → Personal accomplishment (T3).

4) Mindfulness (T1) → Positive affect (T2) → Personal accomplishment (T3).

All analyses were performed two-sided in this study and statistical significance was declared at p < 0.05.

Results

Demographic and clinical characteristics of participants

Independent t tests and chi-square analyses showed no statistically significant differences between the participants who dropped out before the program versus those who completed the study for demographic variables or for any of the dependent variables (mindfulness, cognitive reappraisal, expression suppression, PA, NA, emotional exhaustion, depersonalization, and personal accomplishment) measured at pre-intervention.

Participant characteristics are demonstrated in Table 1. The sample age ranged from 23 to 53 years (M = 33.21 years, SD = 8.71 years) and 89.3% were females (100 females, 12 males). Most of the participants completed a bachelor's degree (94.6%). More than half of the participants were married (57.1%) and most of them did not have a religious belief (89.3%). Most healthcare providers participated were in a third-class (95.5%) specialized hospital (76.8%), with work experience ranged from less than 1 year to more than 20 years. The majority of the participants had no mindfulness experience before (88.4%).

Table 1 Demographic and clinical characteristics of the participants (N = 112)

Baseline differences between intervention and control groups

Chi-squared analyses revealed that there were no significant differences between the intervention group and control group regarding gender (${\chi }^{2}$=0.373, p = 0.541), educational level (${\chi }^{2}$=3.221, p = 0.196), marital status (${\chi }^{2}$=5.036, p = 0.055), hospital type (${\chi }^{2}$=1.803, p = 0.179), and previous mindfulness experience (${\chi }^{2}$=0.087, p = 0.768), though significant differences were found in terms of age (t = 4.639, p < 0.001), position (${\chi }^{2}$=4.356, p = 0.037), work experience (${\chi }^{2}$=13.960, p = 0.001). In addition, no differences were found in any of the psychological outcome variables, namely mindfulness (t = 0.087, p = 0.768), cognitive reappraisal (t = 1.054, p = 0.294), expression suppression (t = -1.205, p = 0.231), PA (t = 0.337, p = 0.737), NA (t = -0.051, p = 0.959), emotional exhaustion (t = -0.518, p = 0.606), depersonalization (t = -1.591, p = 0.114), and personal accomplishment (t = 0.849, p = 0.398). As a result, age, position, and work experience were included as covariates in further analysis.

Program acceptability and feasibility

Results from the program evaluation survey showed that healthcare providers, on average, said the mindfulness training program “was of good quality” (M = 3.45, SD = 0.62), “was what they were hoping for” (M = 3.64, SD = 0.49), “mostly met their needs” (M = 3.11, SD = 0.52), “provided satisfactory help” (M = 3.40, SD = 0.68), and “helped solve some of their problems” (M = 3.60, SD = 0.50). In general, the participants were satisfied with most parts of this mindfulness training program (M = 3.49, SD = 0.69). In addition, most participants said they would recommend this mindfulness training program to their friends (M = 3.62, SD = 0.53) and would receive the program again (M = 3.64, SD = 0.57) if needed.

Analysis of attendance data showed that, on average, participants attended 5.20 sessions (SD = 2.29) during 8-week MBSR program, while 27 of 54 (50.0%) completed the program which was defined as attending 6 or more of the 8 sessions. Amount of home practice was calculated for 47 (83.9%) healthcare providers who returned their weekly mindfulness practice report. Results showed an average of 4.32 days (SD = 1.60) of home practice across the whole intervention period, while only 11 (23.4%) participants showed compliance with the home practice guidelines of 6 days per week.

Effects of the intervention

Repeated measures ANOVAs were conducted to examine changes across time between the intervention and control groups on measures of mindfulness, emotion regulation, affect, and burnout with age, post, and work experience as covariates. Table 2 summarizes descriptive statistics of mean scores and standard deviations of pre-, mid-, and post-measures and significance of time, group and time-group interaction, as well as Cohen’s d of pre-mid and pre–post-intervention.

Table 2 Outcomes of the mindfulness-based intervention compared to the control group

We found a significant time and group interaction effect only for mindfulness (F = 10.578, p = 0.002, η² = 0.090) at mid-intervention. Significant time and group interaction effects were found for mindfulness (F = 30.926, p < 0.001, η² = 0.220), cognitive reappraisal (F = 10.286, p = 0.002, η² = 0.088), positive affect (F = 11.632, p = 0.001, η² = 0.098), emotional exhaustion (F = 5.398, p = 0.022, η² = 0.048), and personal accomplishment (F = 7.621, p = 0.007, η² = 0.066) at post-intervention, with medium and large effect sizes. The analyses indicated that healthcare providers in intervention group scored significantly higher in mindfulness, cognitive reappraisal, PA, personal accomplishment, and significantly lower in emotional exhaustion at post-intervention than those in the control group.

Mediation analyses of intervention effects

Results of the mediation analyses are presented in Table 3. Mediation analyses showed that cognitive reappraisal and positive affect at mid-intervention were significant mediators of the effects of mindfulness on healthcare providers’ personal accomplishment at post-intervention (see Fig. 1).

Table 3 Direct and indirect effects of cognitive reappraisal and positive affect in the intervention

Discussion

The main aim of this research was to determine whether the MBSR program was feasible and efficacious with respect to helping healthcare providers reduce burnout during the COVID-19 pandemic. Furthermore, we were interested in knowing if cognitive–affective process (emotion regulation and affect) might help explain the effect of mindfulness trainings. Overall, the results suggested that (a) the MBSR program was of well acceptability and moderate feasibility; (b) the MBSR program could effectively decrease emotional exhaustion and increase personal accomplishment among healthcare providers; (c) the improvement of personal accomplishment through the training program could be explained by cognitive reappraisal and PA.

With regard to program acceptability, participants indicated strong acceptance of the mindfulness training program in terms of program goals, quality, and benefits. The majority of participants said the program was satisfactory in general and would recommend it to their friends. However, the feasibility of the training program was moderate. Only half of the participants attended 75% or more of sessions and less than one fourth of the participants showed compliance with the home practice guidelines. These results suggested that though mindfulness training program was highly acceptable among healthcare providers, the adherence and compliance of the program among the recipients should be addressed and needed further improvement. A more flexible program schedule adapted for this population and a more suitable home practice plan designed for them were recommended in the future.

Results of this study confirmed our hypotheses concerning intervention effects on mindfulness, emotion regulation, affect, and burnout. Healthcare providers in the mindfulness training condition reported greater mindfulness, cognitive reappraisal, positive affect, personal accomplishment, and less emotional exhaustion at post-intervention than those in the waitlist control condition, with medium effect sizes. These findings are consistent with other studies showing the beneficial effects of mindfulness intervention on mindfulness [46], emotion regulation [19, 22, 57], affect [27, 49], and burnout (Johannes C. [12, 13, 41, 53, 54]) (Fig. 2).

Fig. 2

Mediation model for the effect of mindfulness on personal accomplishment. a Direct effect of mindfulness on personal accomplishment; b mediation model with the indirect effect of pre-intervention (T1) mindfulness via mid-intervention (T2) cognitive reappraisal on post-intervention (T3) personal accomplishment, decreasing the direct effect of group on change in burnout, indicating partial mediation; c mediation model with the indirect effect of pre-intervention (T1) mindfulness via mid-intervention (T2) positive affect on post-intervention (T3) personal accomplishment, decreasing the direct effect of group on change in burnout, indicating partial mediation

In addition to finding evidence of effect differences in these indicators at post-intervention, we also found support for the hypothesis that cognitive reappraisal and positive affect could plausibly account for the effects of the mindfulness training program on improvement of personal accomplishment. These results are in line with the logic model that guides this work: mindfulness requires one to pay attention purposefully and nonjudgmentally to the unfolding of experience in the present moment [29]. The emotional acceptance cultivated by mindfulness training is crucial in fostering cognitive reappraisal and positive affect. The cognitive reconstruct and affective awareness to the experience field increase the capacity for effectively feeling a sense of personal accomplishment and fulfilment.

It is important to note that, in contrast to these positive findings, no statistically significant differences were found in expressive suppression, negative affect, and depersonalization between intervention and control groups in healthcare providers at post-intervention in the study. While previous investigations have yielded varying evidence on depersonalization (Johannes C [12, 13, 33],Suleiman‐Martos et al., 2020), consistent findings from prior research indicate significant outcomes in terms of expressive suppression and negative affect [28, 65]. This may be the result of the small sample sizes in this report as an examination of the effect sizes did show small intervention effects that were in the predicted direction on these measures. On the other hand, implications may be that mindfulness, instead of facilitating suppressing expressions of negative emotions, exerted more beneficial effects on promoting cognitive reconstructing functioning and enhancing positive emotional responding. That may explain why cognitive reappraisal and positive affect showed promising results while expressive suppression and negative affect did not. In addition, mindfulness was more likely to deal with “self” rather than “others”, thus the depersonalization subscale of burnout is not easily to take effect, as depersonalization dimensions described interpersonal interactions with others [29, 43]. Investigating the implications of such findings is needed for the next step in this work.

We note that the mediation effects of cognitive reappraisal and positive affect for the mindfulness training program was only showed in personal accomplishment but not emotional exhaustion. This may add evidence to the theory that the relationship of personal accomplishment to the other two aspects of burnout is parallel [43], rather than sequentially [37]. Though there are research indicating the sequential process of the three dimensions of burnout [5, 55], this study found that emotional exhaustion, instead of being antecedent syndrome of reduced personal accomplishment, may work independently with personal accomplishment. This parallel process of burnout may be well suited in the context of China: In China, people work hard to obtain one’s goals. Even if they have been overwhelmed by work overload and depleted of emotional resources, they still gain a sense of achievement when a job is well done or the work is perfectly completed. This phenomenon indicates that a person can feel a sense of personal accomplishment along with emotional exhaustion, meaning that these two aspects can go side by side. The interpretation of our findings can be examined in future research.

Limitations

Several limitations are important to note. First, the main intervention effects results reported here were based primarily on healthcare providers self-report data. Further research may combine a broader range of measures, such as observational, behavioral, and biological measures. Second, the research conducted here adopted a waitlist control condition, future studies using an active control condition can be applied to increase the rigor of the study design. Third, the casual mediation models we drew here may not be optimal. More research exploring the alternative pathways should be added to the present study. Finally, the study included a motivated and engaged sample of healthcare providers. The degree of the appeal of this kind of mindfulness training program needs further investigation among general hospital healthcare providers.

Conclusion

The results suggest that mindfulness training programs hold promise for the relief of burnout among healthcare providers in China during the COVID-19 pandemic. To be specific, mindfulness training programs can effectively decrease emotional exhaustion and increase personal accomplishment among the recipients. Furthermore, emotion regulation and affect play as key factors that partially mediate the effect of mindfulness trainings on burnout over time. Mindfulness training programs may help healthcare providers facilitate cognitive reappraisal and positive affect to develop personal accomplishment. Examining potential downstream effects of reductions in healthcare provider’s burnout on healthcare settings represent a key next step in this emerging line of research.