Systematic Review of Mind–Body Modalities to Manage the Mental Health of Healthcare Workers during the COVID-19 Era

Healthcare workers (HCWs) have suffered physical and psychological threats since the beginning of the coronavirus disease 2019 (COVID-19) pandemic. Mind-body modalities (MBMs) can reduce the long-term adverse health effects associated with COVID-specific chronic stress. This systematic review aims to investigate the role of MBMs in managing the mental health of HCWs during the COVID-19 pandemic. A comprehensive search was conducted using 6 electronic databases, resulting in 18 clinical studies from 2019 to September 2021. Meta-analysis showed that MBMs significantly improved the perceived stress of HCWs (standardized mean difference, −0.37; 95% confidence intervals, −0.53 to −0.21). In addition, some MBMs had significant positive effects on psychological trauma, burnout, insomnia, anxiety, depression, self-compassion, mindfulness, quality of life, resilience, and well-being, but not psychological trauma and self-efficacy of HCWs. This review provides data supporting the potential of some MBMs to improve the mental health of HCWs during COVID-19. However, owing to poor methodological quality and heterogeneity of interventions and outcomes of the included studies, further high-quality clinical trials are needed on this topic in the future.


Introduction
Globally, the coronavirus disease 2019 (COVID- 19) pandemic has been a threat to mental and physical health of humanity [1]. Since the start of this pandemic, healthcare workers (HCWs) have suffered the physical and psychological threat of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) [2,3]. In this context, the prevalence of anxiety, depression, and stress among HCWs in the COVID-19 pandemic was reported to be as high as 67.55%, 55.89%, and 62.99%, respectively [3]. Moreover, women, younger nurses, frontline HCWs, and workers in areas with higher infection rates are more likely to be severely affected by mental health effects in HCW during the pandemic [3]. Mental health difficulties of HCWs can lead to burnout, worsening attitudes toward patient safety, and hindering the efficient and safe use of medical resources when they are important [4]. As a result, several countries are implementing initiatives to improve health and well-being in HCW in the context of the COVID-19 pandemic, the most common of which are mental health initiatives [5].
Mind-body modality (MBM) can be defined as "a health practice that combines mental focus, controlled breathing, and body movements to help relax the body and mind [6]". MBMs, including meditation, yoga, and mindfulness training, have been considered helpful in stress-related diseases by fostering resilience through self-care [7,8]. Researchers have also found that MBMs are effective in a variety of physical and psychological conditions, including chronic pain, anxiety, depression, cancer-related fatigue, tobacco addiction,

Data Sources and Search Strategy
The following 6 electronic databases were searched for studies published from December 2019 (when the first case of COVID-19 was identified [19]) to September 2021: Medline (via PubMed), EMBASE (via Elsevier), Cochrane Central Register of Controlled Trials, Cumulative Index to Nursing and Allied Health Literature (via EBSCO), Allied and Complementary Medicine Database (via EBSCO), and PsycARTICLES (via ProQuest). In addition, we searched the reference lists of relevant articles and conducted a manual search on Google Scholar to include all of the relevant articles. We included both the literature published in peer-reviewed journals and gray literature, such as dissertations. We designed search strategies for all databases based on the advice of experts in the systematic review (Supplementary Material File S2).

Types of Study Design
Given the urgency of COVID-19, we included all types of original prospective quantitative intervention studies, including randomized controlled trials (RCTs), non-randomized controlled clinical trials (CCTs), and before-after studies. Retrospective and qualitative studies were excluded. There were no restrictions on publication language or publication status.

Types of Participants
We included studies on all types of HCWs, such as physicians, nurses, hospital staff, and health managers, without restrictions on the sex, age, and ethnicity of the participants. However, we excluded studies that did not describe whether participants were directly or indirectly affected by COVID-19.

Types of Interventions
As treatment interventions, MBMs, including meditation, mindfulness-based intervention, autogenic training, yoga, tai chi, qigong, breathing exercises, music therapy, guided imagery, biofeedback, prayer, and faith-based techniques were allowed. As control interventions, no treatment, waitlist, sham control, attention control, or active comparators were allowed.

Types of Outcome Measures
The primary outcome was the level of perceived stress, assessed using validated tools, including the Perceived Stress Scale [20]. Secondary outcomes included mental healthrelated outcomes, such as depression, anxiety, burnout, and safety data of the intervention.

Study Selection
A study selection was conducted through a three-step screening process based on the eligibility criteria. First, titles and/or abstracts of the searched studies were screened to identify potentially eligible articles. Second, potentially eligible reports were sought for retrieval. Third, the full text of the retrieved reports was reviewed. Two researchers (C.-Y.K. and B.L.) independently conducted the study selection process, and disagreements between the researchers were resolved through discussion.

Data Extraction
Two researchers (C.-Y.K. and B.L.) independently extracted the following information using a predefined, pilot-tested excel form: the first author's name, year of publication, country, study design, sample size, details of participants, treatment and control interventions, a treatment period of intervention, outcome measures, results, and safety data. Any discrepancies between the researchers were resolved through discussion. When the data were insufficient or ambiguous, the corresponding authors of the original studies were contacted via e-mail.

Methodological Quality and Risk of Bias Assessment
The methodological quality of the included studies was assessed using the corresponding critical appraisal skills program tools, depending on the study type [21]. For RCTs, the Cochrane Collaboration risk of bias tool was used to assess the related risk of bias [22]. For CCTs and before-after studies, the Quality Assessment of Controlled Intervention Studies and the Quality Assessment Tool for Before-After (Pre-Post) studies with no control group by the National Heart, Lung, and Blood Institute were used to assess methodological quality [23]. Two researchers (C.-Y.K. and B.L.) independently assessed the methodological quality and risk of bias of the included studies and any disagreements between the researchers were resolved through discussion.

Data Analysis and Synthesis
Descriptive analyses of the participants, interventions, controls, and outcomes of all of the studies were performed. If there were 2 or more RCTs or CCTs with the same outcome measures, a meta-analysis was performed using RevMan 5.4 (the Cochrane Collaboration, London, UK). In the meta-analysis, dichotomous and continuous data were presented as risk ratios (RR) and standardized mean differences (SMD) with 95% confidence intervals (CIs). The I 2 values of ≥50% and ≥75% were considered substantial and statistically heterogeneous, respectively. In the meta-analysis, a random-effects model was used if the included studies had significant heterogeneity (I 2 value ≥ 50%), whereas the fixed-effect model was used when the heterogeneity was insignificant or the number of studies included in the meta-analysis was less than five [24]. If sufficient data were available, subgroup analyses were planned according to the (a) type of HCWs and (b) type of mind-body modality. In addition, sensitivity analyses were conducted to identify the robustness of the results by excluding (a) studies with a high risk of bias and (b) data outliers. Evidence of publication bias was assessed using funnel plots if at least ten RCTs were included in each meta-analysis. The results of the included before-and-after studies were only described without quantitative synthesis.

RCTs
Three RCTs [58,65,68] described appropriate random sequence generation methods such as computerized randomization; conversely, the other two studies [66,69] did not describe this method. No study has described the method of allocation concealment. Fiol-DeRoque et al. [58] conducted a study using a mobile application and reported that double-blind was implemented. The remaining four RCTs [65,66,68,69] did not report the implementation of blinding but were evaluated as high due to the nature of the intervention. Only one study [58] reported that blinding of the outcome assessment was not performed; in the remaining studies [65,66,68,69], blinding of the outcome assessor was not described. For incomplete outcome data, one study [58] performed an intention-to-treat analysis, and another study [66] with no dropouts was rated low in this domain. In the remaining three studies [65,68,69], dropouts existed, but the cause was not described, and a per-protocol (PP) analysis was performed; therefore, this domain was rated highly. With regard to selective reporting, the protocol was confirmed in only one study [58], and the pre-planned outcome was confirmed to be reported and evaluated as low. Other studies [65,66,68,69] evaluated selective reporting as unclear. Three studies [65,66,68], in which clinical and demographic homogeneity between the two groups was confirmed at baseline, were evaluated as low in other sources of bias, and the remaining studies were evaluated as unclear (Supplementary Material File S3).

CCTs
As the included CCTs were not RCTs, they were evaluated as "no" in Q1 and "not applicable (NA)" in Q2. Allocation concealment and blinding procedures were not reported in any of the studies. Therefore, Q3 and Q5 were evaluated as not reported (NR), and Q4 was evaluated as "no", considering that double-blinding was impossible due to the nature of the intervention. In a study [59], statistical heterogeneity of baseline characteristics was reported, and it was evaluated as "no" in Q6; the heterogeneity was not described in the remaining studies [57,62,70]. In another study [62], the overall dropout rate from the study at the endpoint was more than 20% of the number allocated to treatment, and the differential dropout rate at the endpoint was more than 15 percent. Therefore, Q7 and Q8 of this study [62] were evaluated as "no". As treatment adherence was not reported in all of the studies, Q9 was evaluated as NR. As only one study [57] recommended avoidance of other interventions in the groups, it was evaluated as "yes" in Q10. As validated outcomes were used in all of the studies, they were evaluated as "yes" in Q11. The sample size was calculated in only one study [70] and was evaluated as "yes" in Q12. The pre-registered protocol was confirmed in only one study [62], its Q13 was evaluated as "yes," and the remaining studies [57,59,70] were evaluated as cannot be determined. Two studies [57,70] with no dropouts were evaluated as NA in Q14, whereas the other two studies [59,62] with PP analysis were evaluated as "no" (Supplementary Material File S3).

Before-after Studies
The purpose of the studies was clearly described; therefore, it was evaluated as "yes" in Q1. As the selection criteria for the study population were not clearly described in the four studies [54][55][56]60], Q2 and CD were not evaluated in Q4. Except for one study [61] that included only geriatric fellows as participants, other studies [53][54][55][56]60,63,64,67] included two or more occupations, so their Q3 were evaluated as "yes." In only one study [61], the sample size was calculated and evaluated as "yes" in Q5. Because the intervention was insufficiently described in two studies [55,67], it was evaluated as "no" in their Q6. Two studies [53,63] that did not use a validated outcome were rated as "no" in their Q7. Only one study [64] reported that the outcome assessor was blinded. As not described in other studies [53][54][55][56]60,61,63,67], all were assessed as CD in their Q8. In two studies [54,60], loss to follow-up after baseline was >20%; therefore, it was evaluated as "no" in their Q9. One study [63] without a statistical test for pre-to post-changes, was rated as "no" in Q10. In only one study [56], the outcome assessment was performed three times, and it was evaluated as "yes" in Q11. In other studies [53][54][55]60,61,63,64,67], it was evaluated as "no" in Q11 because it was only evaluated twice: before and after. There was no intervention at the group level; therefore, all were evaluated as NA in Q12 (Supplementary Material File S3).

Main Results
The outcomes used among the included studies varied, but they can be classified into 12 categories: perceived stress (the primary outcome), psychological trauma, burnout, insomnia, self-efficacy, anxiety, depression, self-compassion, mindfulness, and quality of life (QOL), resilience, and well-being. Among these, meta-analyses of stress, depression, and anxiety are possible.

Safety Data
None of the included studies reported adverse events or safety data.

Publication Bias
As fewer than ten studies were included in the meta-analysis, publication bias through funnel plot generation was not evaluated.    Outcomes related to self-efficacy 1. General Self-Efficacy Scale (self-efficacy) CBT and mindfulness-based app vs. Psychoeducation app (14 days) NS (p > 0.05) Fiol-DeRoque 2021 [58] Outcomes related to anxiety

Safety Data
None of the included studies reported adverse events or safety data.

Publication Bias
As fewer than ten studies were included in the meta-analysis, publication bias through funnel plot generation was not evaluated.

Main Findings
This review was performed in order to investigate the benefits of MBMs on the mental health aspects of HCWs in the context of COVID-19. Through a comprehensive literature search, 18 studies [53][54][55][56][57][58][59][60][61][62][63][64][65][66][67][68][69][70] were included. According to the meta-analysis, MBMs had a significantly positive effect on the perceived stress of HCWs, which was the primary outcome of this study. Regarding the types of individual MBMs, the effects of yoga-and music-based interventions appeared to be the most prominent. In a meta-analysis of depression and anxiety, MBMs showed a significantly positive improvement compared to the control group. In this case, yoga-and music-based interventions had the largest effect size. However, the effect of self-compassion-based interventions on stress, depression, and anxiety was not significant. In other words, the effects of MBMs on mental health of HCWs may differ according to individual MBMs. For individual outcomes, some MBMs had significant positive effects on psychological trauma, burnout, insomnia, anxiety, depression, self-compassion, mindfulness, QOL, resilience, and well-being, but not psychological trauma and self-efficacy, compared to controls (in RCTs and CCTs) or baseline (in beforeafter studies). Although these results provide data that some MBMs may be useful options for mental health management of HCWs in the context of COVID-19, the methodological quality of the included studies was not optimal. In addition, the number of RCTs performed with rigorous design was insufficient, and CCT or before-after studies accounted for more than half of the studies. Therefore, the findings of this study could be greatly influenced by the results of large-scale rigorous clinical studies in the future.

Clinical Implications
The mental health of HCWs in the context of COVID-19 is a serious threat [3,4], and measures to manage it are urgently needed [5]. At present, to protect the mental health of this population and reduce stress as much as possible, there is an emphasis on establishing tailored, effective stress reduction interventions [71]. To develop effective antistress interventions, empirical evidence exists for some MBMs, such as mindfulness-based interventions, diaphragmatic respiration, and acting on self-efficacy [71]. Breath-focused mind-body therapies, as a strategy in precision medicine, have recently been claimed by some researchers to be effective in managing stress and anxiety [72,73]. Therefore, leveraging MBMs to manage the mental health of HCWs in the context of COVID-19 could be a promising strategy [10].
According to the results of the studies included in this review, some MBMs have shown positive effects in improving the mental health aspects of HCWs. When classified as individual outcomes, MBMs showed a relatively high rate of positive effects on selfcompassion, QOL, perceived stress, resilience, burnout, insomnia, and well-being in this population. In addition, the individual MBMs used were heterogeneous, and no studies have compared two or more MBMs in this population. Therefore, it was difficult to find the optimal MBMs for mental health and psychological stress management in HCWs during the COVID-19 era. Nevertheless, according to the meta-analysis of the review, yoga-and music-based interventions had a larger effect size on perceived stress (DASS) than CBT and mindfulness-based or self-compassion-based interventions. A recent meta-analysis also found that yoga may help relieve stress in people who live under high stress or negative emotions, including HCWs, and the mechanism may be related to the modulation of sympathetic-vagal balance [74,75]. Moreover, in a systematic review of MBMs for nurses, yoga may be helpful in improving burnout and perceived stress among nurses in hospital setting [76]. Important considerations should be taken into account when introducing interventions for mental health management of HCWs in the context of COVID-19. The reality of a busy clinical setting must be considered. Strategies for managing stress and mental health in this population should be feasible and accessible. In this context, more than half of the included studies [53,54,56,58,[60][61][62][63]65,[68][69][70] (12/18, 66.67%) provided participants with MBMs combined with ICT, including mobile phone applications, guided audio files, video files, and video conferences. In addition, three studies [55,64,67] integrated MBMs into the work environment of participants, such as the educational curriculum and time for a shift. These results suggest that MBMs may be introduced to reflect the work environment of HCWs in the context of COVID-19.

Limitations and Suggestions for Further Studies
This systematic review has several limitations. First, the MBMs and outcomes used in the included studies were heterogeneous. Therefore, quantitative synthesis was not possible for most outcomes in this study. However, given that the mental health of HCWs in the context of COVID-19 has important clinical relevance and MBMs are a promising option, future research in this field needs to be further standardized and refined. In particular, a head-to-head trial may be attempted to investigate the most effective MBMs for improving perceived stress among HCWs during the COVID-19 era. Second, the methodological quality of the included studies was suboptimal. These methodological limitations may negatively affect the reliability of this study's findings. In addition, it was rare among the included studies to be conducted with strict designs, including RCT (5/18, 27.78%). Future research in this field will be able to reflect a research design that reflects the characteristics of MBMs. For example, to investigate the effectiveness and safety of MBMs, n-of-1 trials may be ideal than traditional RCT [77]. Third, this study did not consider the temporal and environmental effects of COVID-19. Given that the impact of COVID-19 on the mental health of HCWs may vary depending on the time, place, and clinical setting, future research in this field will be able to develop customized interventions that take into account the temporal and environmental impacts of COVID-19 on HCWs.

Conclusions
This is the most comprehensive review available on the impact of MBMs on the mental health of HCWs during the COVID-19 era. This review provides data supporting the potential of some MBMs to improve the mental health of HCWs during COVID-19. There is evidence that yoga-and music-based interventions are helpful for improvement for perceived stress, the primary outcome. In addition, some MBMs had significant positive effects on psychological trauma, burnout, insomnia, anxiety, depression, self-compassion, mindfulness, QOL, resilience, and well-being, but not psychological trauma and selfefficacy of HCWs. However, owing to poor methodological quality and heterogeneity of interventions and outcomes of the included studies, further high-quality clinical trials are needed on this topic in the future.  Institutional Review Board Statement: Ethical review and approval were waived for this study, as this study is a systematic review of previously published studies.
Informed Consent Statement: Patient consent was waived, as this study is a systematic review of previously published studies.

Data Availability Statement:
The data presented in this study are available in the article and Supplementary Materials.

Conflicts of Interest:
The authors declare no conflict of interest.