Article Text
Abstract
There is a growing need for interventions to improve well-being in healthcare workers, particularly since the onset of COVID-19.
Objectives To synthesise evidence since 2015 on the impact of interventions designed to address well-being and burnout in physicians, nurses and allied healthcare professionals.
Design Systematic literature review.
Data sources Medline, Embase, Emcare, CINAHL, PsycInfo and Google Scholar were searched in May–October 2022.
Eligibility criteria for selecting studies Studies that primarily investigated burnout and/or well-being and reported quantifiable preintervention and postintervention outcomes using validated well-being measures were included.
Data extraction and synthesis Full-text articles in English were independently screened and quality assessed by two researchers using the Medical Education Research Study Quality Instrument. Results were synthesised and presented in both quantitative and narrative formats. Meta-analysis was not possible due to variations in study designs and outcomes.
Results A total of 1663 articles were screened for eligibility, with 33 meeting inclusion criterium. Thirty studies used individually focused interventions, while three were organisationally focused. Thirty-one studies used secondary level interventions (managed stress in individuals) and two were primary level (eliminated stress causes). Mindfulness-based practices were adopted in 20 studies; the remainder used meditation, yoga and acupuncture. Other interventions promoted a positive mindset (gratitude journaling, choirs, coaching) while organisational interventions centred on workload reduction, job crafting and peer networks. Effective outcomes were reported in 29 studies, with significant improvements in well-being, work engagement, quality of life and resilience, and reductions in burnout, perceived stress, anxiety and depression.
Conclusion The review found that interventions benefitted healthcare workers by increasing well-being, engagement and resilience, and reducing burnout. It is noted that the outcomes of numerous studies were impacted by design limitations that is, no control/waitlist control, and/or no post intervention follow-up. Suggestions are made for future research.
- Quality in health care
- MENTAL HEALTH
- Quality of Life
- Health & safety
- Interprofessional Relations
Data availability statement
All data relevant to the study are included in the article or uploaded as supplementary information. Other data are available on reasonable request to the corresponding author.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
Strengths and limitations of this study
This review analysed various workplace well-being strategies, both individually and organisationally focused interventions for physicians, nurses and allied healthcare workers.
The reported evidence in this review builds on previous evidence in the literature that focused on specific well-being interventions such as mindfulness-based practices and contrasts these results against other types of workplace well-being interventions.
A quantitative meta-analysis was not possible due to the marked variation in study designs and reported outcomes implemented by the eligible studies.
Organisationally focused interventional studies were under-represented in this review with only three meeting the inclusion criteria.
Introduction
Widely acknowledged as the cost of caring, healthcare workers report higher rates of absences due to psychological distress and job burnout than workers in other sectors.1–6 Psychosocial hazards such as chronic exposure to occupational stress, place healthcare workers at greater risk of presenteeism (reduced productivity), anxiety and depression.3 7–9 Occupational stress relates to workplace interferences that can disturb a worker’s well-being physically and mentally, thereby fostering the potential for burnout.1 10 Moreover, the financial constraints of healthcare systems within developed countries are further pressurised by an ageing patient population, technological advances and poor worker retention.10–12 The Australian Institute of Health and Safety estimated in 2019, that worker’s absenteeism due to poor mental health, cost between $13 and $17 billion per year.13 These figures were exponentially compounded throughout the recent COVID-19 pandemic.9 10 14 For example, in a study conducted in England during the first wave of COVID-19, sickness absences due to poor mental health in National Health Service staff, increased from 519 807 days in March–April of 2019 to 899 730 days 12 months later.14
The reduced productivity of healthcare professionals in the workplace can lead to suboptimal delivery of care to patients and therefore poorer treatment outcomes.6 10 Consequently, there is growing interest in boosting healthcare worker wellness, resiliency and self-care, globally.6 The framework for occupational health and safety is moving beyond traditional workplace hazards and now seeks to include emerging demographic factors such as mental health conditions.15 This view is reflected in the 2030 United Nations Goal that employment respects the ‘… physical and mental integrity of the worker in the exercise of his or her employment’.15 Additionally, in July 2022, Safe Work Australia released new codes of practice for managing psychosocial hazards at work, including risks to workers’ mental health.16 Using theoretical frameworks such as the Job Demands-Resources (JD-R) model or Watson’s Human Caring theory, researchers are developing support strategies that aim to improve well-being as well as understand worker’s wellness and exhaustion.7 17 Both these validated frameworks are invaluable to well-being researchers as the JD-R model explains how increasing job resources can buffer job strain in workers, while Watson’s theory of Human Caring aims to create a holistic or mind-body-soul approach to human care.17–19 The literature demonstrates that workers who receive treatment for absenteeism and presenteeism, for example, through workplace well-being interventions, report higher levels of productivity, which can then lead to higher levels of job satisfaction.3 Furthermore, workplace interventions that focus on self-care, worker empowerment and access to mental health services such as mindfulness sessions or meditation, have been found to decrease levels of psychological stress, anxiety and burnout in healthcare workers.20
Workplace interventions may be designed and applied based on the level at which behavioural changes are targeted, for example, the individual worker, manager or the organisation itself.21 Interventions designed to target health behaviours for the individual, such as mindfulness-based practices, gratitude journaling, meditation or yoga are increasingly popular among researchers.8 22–24 As such, several recent systematic reviews have reported the benefit of individual focused interventions in healthcare, predominantly highlighting the effectiveness of mindfulness-based practices to improve well-being.22–24 Organisational interventions designed to target the source of occupational stress such as reducing workloads, increasing autonomy or job crafting (physical and cognitive changes individuals make in the task or relational boundaries of their work25), are less explored.21 The predominant view in the literature is that interventions designed to alter health behaviours in the individual, may be a reactive strategy to occupational stress, whereas organisational change may be far more proactive in promoting worker well-being in the long term.21 25 A recent surge in the understanding of workplace wellness and causes of occupational stress further promotes the need to explore and invest in organisational strategies.26 27 As it stands, institutions may be reluctant to implement large organisational changes without strong literature evidence supporting the effectiveness and long-term benefits to employees.26 27
A second approach to implementing workplace well-being interventions can be achieved through first identifying the presence and likely cause of occupational stress and then applying the intervention at either the primary, secondary or tertiary level.28 Primary interventions may be implemented in a largely preventative manner as they aim to eliminate occupational stress or change the cause of stress such as through workload reductions.28 Secondary and tertiary interventions are designed to treat workers who are already showing signs of occupational stress. Secondary interventions aim to help workers minimise the effects of occupational stress through methods such as relaxation training and tertiary interventions aim to treat workers who have already developed stress-related health issues such as anxiety or depression.28
To date, a multitude of systematic reviews have investigated the effects of mindfulness-based education or yoga interventions for healthcare professionals in a wide array of contexts. For example, Lomas et al22 conducted a meta-analysis investigating the impact of mindfulness-based interventions on healthcare workers, Cocchiara et al29 investigated the use of yoga to manage stress and burnout in healthcare workers and Klein et al30 investigated the benefits of mindfulness-based interventions on burnout among health professionals. Other systematic reviews have focused on specific populations, for example, DeChant et al31 investigated the effect of organisation-directed workplace interventions on physician burnout and Murray et al32 investigated interventions to improve the psychological well-being of general practitioners. To the authors knowledge, no systematic review has been conducted to provide an overview of all types of well-being interventions for allied healthcare professionals, including physicians and nurses. The authors also acknowledge the effects that COVID-19 has had on healthcare practices as well as healthcare worker mental well-being.14 Therefore, in order to include all above-stated professional groups and types of interventions in this review, as well as sampling the most recent research (including 5 years prior to the onset of COVID-19), the search criterium for this review was restricted to 2015.
This systematic review therefore aims to identify and analyse all positive outcome measures produced by workplace interventional strategies designed to support well-being and reduce burnout for nurses, physicians and allied health professionals since 2015.
Methodology
A search strategy was developed by the research team and validated by an institutional research librarian using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines prior to searching the literature and we used the PRISMA checklist when writing our report.33 34 The full search protocol, including the search terms, can be found in online supplemental appendix. Title and abstract screening as well as full-text screening was carried out by two members of the research team (CC, JC). Where there were disagreements or uncertainty, a third, senior member of the research team was available for resolution if reconciliation sessions were unsuccessful.
Supplemental material
Patient and public involvement
There was no patient or public involvement in this research.
Eligibility criteria
Population
All studies in which the sampled population included either allied health personnel, physicians or nurses were included in this review. Studies that contained mixed populations that is, healthcare workers as well as patients, were included, however, data not pertaining to healthcare workers were excluded. Studies that did not confirm that healthcare workers, physicians or nurses were involved in the intervention, were also excluded. A summary of the allied health professions included in this review, both regulated by the Australian Health Practitioner Regulation Agency and self-regulated, as per the Australian Government Department of Health can be found in online supplemental file 1.35
Interventions
This review accepted all studies reporting the outcomes of either organisationally or individually focused workplace well-being interventions. Studies whereby improving employee well-being or reducing burnout was not the primary focus of the intervention, were excluded. Studies that suggested workplace interventions but did not implement them were also excluded.
Outcome measures
All study designs were accepted if objective and quantifiable preintervention and postintervention outcome measures such as observer ratings were reported, using a valid and reliable well-being or burnout survey inventory. Systematic literature reviews were not included in the initial synthesis; however, they were included in the discussion and summary of study findings. Studies published prior to 2015 were also excluded to ensure the most up to date literature was reviewed (while still including 5 years prior to the onset of COVID-19) as well as including all well-being intervention types for nurses, physicians and allied healthcare professionals.
Literature search
Five healthcare/medicine databases (CINAHL, Embase, Emcare, Medline, PsycInfo) were searched on 2 May 2022 and again on 5 October 2022 using a three-step (Population, Intervention, Outcome) systematic approach (see inclusion criteria for details). A secondary search of Google Scholar was also performed with results limited to the first five pages of articles (n=50). The outcomes of the literature search are outlined in figure 1.
Study selection
Data extraction was performed by the first author (CC) and verified by all other members of the research team (EB, JC, SP and MT). The extracted data was summarised into an Excel spreadsheet and comprised study design, objectives, location, participants (numbers and characteristics), sample size justification, number of institutions, outcome measurement tools, intervention used (type, repeatability, random or blinded), data collection timeframe, ethical approval, rate of attrition, quantitative data results and study outcomes.
Quality assessment
Quality assessment of the articles was also carried out independently by two members of the research team (CC and JC) using the Medical Education Research Study Quality Instrument (MERSQI).36 The MERSQI is a long-standing and validated quality assessment tool that is applicable to a wide array of study designs.35 Papers were given scores (out of a maximum of 18) based on study design, number of institutions, response, type, validity of the methods, sophistication of data analysis, appropriateness and reported outcomes.36
Meta-analysis
It was not possible to perform a meta-analysis due to a marked variation in study designs, interventions employed, survey instruments used and statistical analyses performed.
Results
The database searches produced 1596 total articles for screening. A grey literature search of Google Scholar as well as Pearling reference lists generated a further 67 papers. In total, 545 articles were marked as duplicates. Of the remaining 1118 articles, 921 were removed during title and abstract screening. Of the 197 full-texts articles, 40 were included in this review. Following quality assessment, seven papers that scored 12 points or lower were excluded from data synthesis leaving 33 articles for analysis.
Characteristics and quality of studies
The full results of the quality assessment are available in online supplementary table 1. All articles in this review were published between 2015 and 2022 with most studies (n=22) based in the USA.1 4 7 17 26 37–53 Two studies were based in Australia,54 55 two in the Netherlands56 57 and one each from Ireland,58 Italy,59 Portugal,60 Brazil,61 Hong Kong,12 Japan62 and Iran.63 Twenty-seven1 4 7 12 26 37–52 54 58–61 63 studies had participants that were predominantly female, two56 57 reported more male than female participants and four17 46 53 55 did not specify gender. A total of 1617 37 39 42 44 46–48 50–54 60 61 63 articles reported that nurses were the sole participants in their studies, 526 38 43 56 57 reported physicians only, 159 reported sampling both physicians and nurses, 51 12 40 49 55 sampled a variety of healthcare workers (including allied health) and 64 7 41 45 58 62 articles used generic terminology such as healthcare professionals, hospital staff or clinical/non-clinical roles.
Supplemental material
In the five studies that sampled various allied healthcare professionals, three1 40 55 studies listed social workers, three12 49 55 included occupational therapists, two1 55 mentioned psychologists/therapists, two12 49 reported physiotherapists, two40 49 included pharmacists and one49 study mentioned technologists/technicians. Sample sizes ranged from 955 to 22849 with 117 12 39 43 44 48 49 52 57 60 62 studies providing a power analysis to sample size justification.
Eight4 12 37 41 43 63 studies used randomised controlled trials (RCTs), six26 40 43 44 57 60 used quasi-experimental study designs and nineteen1 7 17 42 45–55 58 59 61 62 used a single group pretest/post-test study design method. Of the 14 studies that implemented a control group, 34 12 44 used an active control method, 426 39 43 63 chose no treatment control groups and 737 38 40 41 56 57 60 implemented waitlist control groups. Seventeen4 37 38 40 42 44 45 47 50 51 53–55 57 59 60 63 studies implemented interventions incorporating mindfulness based education (MBE) such as stress management, resiliency training, emotional intelligence training, or improving mindfulness and self-compassion. Three7 17 41 studies opted for MBE combined with yoga and one39 study used yoga only. The remaining studies implemented a myriad of well-being interventions including gratitude journaling,12 49 62 meditation sessions,48 61 workload reduction,26 56 massage chairs,46 choir singing,58 acupuncture,1 professional coaching43 and a peer support network (PSN).52
The longest intervention period was 6 months,44 and the shortest intervention period was 90 min.50 In each study, participants were issued surveys at baseline and immediately following the intervention. Several studies also issued additional follow-up surveys at either 1, 3, 6 or 12 months postintervention. Overall, attrition rates were low across the studies with 1912 17 39–44 46 47 49–51 55 58 59 61–63 reporting they experienced 0% participant dropout and only 51 45 52 53 60 recording 25% or greater attrition. For some studies, it was difficult to quantify attrition rates, for example, Bevan and Emerson42 reported 0% attrition yet also stated that two out of three mindfulness sessions were not attended by all participants (session one 53% and session two 85% attendance). In another web-based study,37 10 out of 52 experimental participants did not withdraw from the study but never actually logged online to use the programme. All studies provided sufficient details regarding the interventions used, to ensure repeatability. All outcome measures relating to well-being or burnout were evaluated for this review, however burnout, perceived stress, depression, and anxiety as well as resilience, quality of life and work engagement levels were of particular interest to the authors. A full summary of the wellbeing interventions analysed in this review are outlined below in table 1.
Organisationally focused interventions
Three26 52 56 of the thirty-three studies implemented workplace interventions focused on the organisation altering daily practices and protocols to enhance worker well-being. Two26 56 of these studies used interventions designed to eradicate or eliminate stress in the workplace and were therefore categorised as primary level interventions and one52 was secondary that is, participants were managed for symptoms of occupational stress. Gordon et al56 implemented general job crafting strategies based around weekly goals to promote teamwork and increase quality of care for medical specialists and nurses, through training and setting personal goals. In the study conducted by Gregory et al26 workloads for physicians were reduced, by employing a certified medical assistant to run appointments, manage pharmaceutical refill requests, triage and coordinate care activities.
Finally, a three-tiered PSN programme was delivered by Wahl et al52 which consisted of departmental peer supporters, a trained peer support team and mental health experts.52 The key goals of this study were to develop a PSN, identify at-risk colleagues, and understand and deliver therapeutic communications by developing keywords and phrases.52
The duration of these interventions varied from 3 weeks56 to 3 months52 with only one26 conducting an additional follow-up, 3 months after the intervention. All three studies showed positive results; however, comparison and contrast of efficacy is difficult due to the significant heterogeneity of data collection methods and outcomes measured (see for a summary of all study outcomes). Both Gregory et al26 and Gordon et al56 implemented quasi-experimental study designs with non-randomised, no treatment and waitlist control groups. Wahl et al52 implemented a single group pretest/post-test study design method and recorded a significantly smaller sample size of just 20 participants compared with the other two organisational studies (10756 and 69,26 respectively). All three studies reported on multiple well-being outcomes by using more than one validated survey instrument, however, there was not a wellness inventory (survey tool) that was implemented in all three studies. Gregory et al26 and Gordon et al56 measured levels of emotional exhaustion (EE) for their participants using the Maslach Burnout Inventory (MBI) and the Oldenburg Burnout Inventory (OLBI).
Both studies demonstrated statistically significant reductions in EE for participants following the intervention in comparison to the control groups, the reduction in EE was also sustained at the 6-month follow-up in the study by Gregory et al.26 In the study by Wahl et al52 a reduction in the burnout subscale of the Professional Quality of Life Scale (ProQOL-5) was reported by participants following the PSN intervention but it was not significant.52 A significant improvement in mean scores was reported for the Compassion Satisfaction (CS) subscale of the Compassion Practice Instrument (CPI), however, it was noted that comparison of the CS subscales for the ProQOL-5 and the CPI revealed contradictory results in that CS scores decreased following the intervention as measured by the ProQOL-5. It is thought that the CPI may be more sensitive for measuring CS and therefore able to discern more subtle differences.52
Individually focused interventions
Thirty1 4 7 12 17 37–51 53–55 57–63 studies in this review implemented secondary level well-being interventions that targeted individual behavioural changes by aiming to minimise the effects of stress in participants through such techniques as relaxation training or promoting a positive mindset.
Relaxation techniques
Various relaxation techniques accounted for 25 of the individually focused intervention studies. MBE was the most consistently implemented (n=20),4 7 17 37 38 40–42 44 45 47 50 51 53–55 57 59 60 63 followed by meditation,45 48 yoga,39 acupuncture1 and massage chairs.46
Of the 20 intervention studies involving MBE, 104 7 38 40–42 44 50 54 63 held in-person educational sessions, 717 50 51 53 55 59 60 applied a combination of in-person education sessions followed by individual practices at home and 44 37 45 47 used online delivery methods either via a smartphone application or web-based programmes. In the study performed by Mistretta et al,4 two intervention groups were structured to implement MBE both in-person and via a smartphone application and therefore has been listed two times. Study durations for MBE ranged from 2 to 16 weeks, with a single outlier study running for 6 months (McNulty et al44 implemented four MBE sessions throughout a 6-month nursing residency programme). The number of sessions delivered in each MBE programme also varied greatly from a single seminar50 to a 12-session education programme.62 Most delivered between 4 and 12 sessions. Nine out of the twenty studies implementing MBE interventions used control groups and of these, six37 38 40 41 50 60 implemented a waitlist style control, two4 44 were active and one37 group received no treatment. The largest sample size was reported by McNulty et al44 with 131 participants in the intervention group and 69 in the active control group. The smallest sample was 13.42 All programmes that delivered in-person group-based MBE education reported at least one statistically significant, positive outcome. Four4 7 38 50 studies documented EE levels as measured by the MBI with three4 38 50 reporting improvements following the intervention. Although not statistically significant, EE levels did improve in participants in the study by Ofei-Dodoo et al7 who also noted that baseline MBI scores were low. This study did show significant reductions in depression, anxiety and stress scores as measured by the Depression Anxiety and Stress Scale (DASS-21) following the intervention, as well as significant improvements in levels of Personal Accomplishment. Changes in stress (DASS-21), as well as levels of perceived stress (Perceived Stress Scale (PSS)) were reported in 74 7 38 41 44 45 54 of the 11 studies. Of these seven, six were statistically significant with Schroeder et al.38 Werneburg et al45 and Mistretta et al4 all reporting that reductions in levels of stress were maintained at the 3-month follow-up. Mistretta et al4 also reported improvements in well-being as measured by the WHO Well-being Index (WHO-5) for the in-person intervention group. Schroeder et al38 and Colgan et al40 both reported improvements in resiliency levels using the Brief Resilience Scale for participants following the MBE sessions compared with the control groups. A substantial increase in resiliency levels was reported by Colgan et al40 following the intervention, however, the increase in resiliency for participants in the study by Schroeder et al38 was not enough to be significant. Werneburg et al45 and Ofei-Dodoo et al7 also both reported changes in resilience levels using the Connor Davidson Resilience Scale (CD-RISC) and the Resilience Scale (RS-14), respectively. Neither study employed a control group however both reported substantial improvements in resiliency following the interventions which were maintained at the 3-month follow-up in Werneburg et al’s study.45
Of the seven studies that implemented a combination of in-person MBE followed by self-directed practice, six17 51 53 57 59 60 reported on levels of burnout using either the MBI or the ProQOL-5 and of these, two50 60 employed a waitlist control group. Statistically significant reductions in burnout were reported postintervention in the studies by Duarte and Pinto-Gouveia60 and Hevezi.51 There were significant reductions in EE as well as Depersonalisation (DP) levels but not overall burnout in the study by Caponnetto et al59 with Verweij et al57 documenting significantly reduced DP levels only. Van Horne et al53 reported a reduction in burnout for participants who attended three or more OASIS sessions versus those who reported not attending any sessions, however, this study did not report all mean scores. Dobie et al55 and Duarte and Pinto-Gouveia60 documented significantly reduced stress levels in participants following the intervention as measured by the DASS-21 and Caponnetto et al59 also reported a significant reduction in perceived stress as measured by the PSS. It is worth noting however, that the study by Dobie et al55 had strong limitations in that they recruited the smallest sample size of just nine participants. Bianchini and Copeland17 employed the largest sample size (143 participants) yet did not report any statistically significant change in outcome measures following the 12-week didactic intervention. Of the four studies that implemented MBI practices in an online or solely self-directed manner, two37 47 used the web-based BREATHE programme designed to help nurses manage stress. Both studies reported statistically significant changes in levels of stress as measured by the Nurse Stress Scale (NSS) despite the varied study designs.37 47 Hersch et al37 implemented an RCT design with a total of 104 participants while Dutton and Kozachik47 used a single group pretest/posttest study design with 31 participants. Participants in both studies were encouraged to login and use the online programme as much as possible within a 2-month47 and 3-month37 period.
The smartphone intervention group in the study by Mistretta et al4 reported less significant changes in outcome measures compared with the in-person MBE intervention group. Significant improvements in well-being were reported as measured by the WHO-5, however, this was not maintained at the 6-month follow-up.4 Reductions in stress, anxiety and depression were not enough to be significant for the smartphone group. Two48 61 of the thirty individually focused studies implemented a meditation intervention and one39 implemented yoga. Three7 17 62 of the twenty MBE studies also incorporated yoga as part of the overall mindfulness education programme with one study by Duchemin et al41 using a combination of MBE, yoga and meditation. Neither of the meditation studies used a control group and both had relatively small sample sizes (13 and 27 participants). Burnout levels were reported to be significantly reduced in both studies using the MBI58 and the ProQOL-5 burnout subscale.48 dos Santos et al61 additionally reported significant reductions in perceived stress as well as trait anxiety using the State-Trait Anxiety Inventory (STAI) which were all further maintained at the 6-week follow-up. Bonamer and Aquino-Russell48 reported that resiliency scores (CD-RISC) significantly increased following the meditation intervention and demonstrated a correlation between resiliency combined with burnout and CS (ProQOL-5).
In the single yoga study by Alexander et al39 the intervention group reported significant improvements in self-care as measured by the Health Promoting Lifestyle Profile (HPLP II) and mindfulness as measured by the Freiburg Mindfulness Inventory (FMI) as well as significant reductions in EE and DP postintervention. Yoga was also incorporated into three of the MBE studies with Ofei-Dodoo et al7 implementing group mindfulness-based yoga sessions 1 hour a week for 8 weeks. The 43 participants in this study reported significant improvements in depression, anxiety and stress (DASS-21) as well as increased resilience (RS-14) and compassion as measured by the Santa Clara Brief Compassion Scale (SCBCS).7 The other two17 41 MBE studies that incorporated yoga both implemented a control group, and both reported no significant change in burnout (MBI) or perceived stress. Statistically significant reductions in both state anxiety and trait anxiety (STAI) were reported following the auricular (around the ear) acupuncture intervention, and total work engagement scores as measured by the Utrecht Work Engagement Scale (UWES) also showed significant improvement for the 112 participants.1 The study using a massage chair intervention reported substantial reductions in all outcome measures (perceived stress, heart rate, systolic and diastolic blood pressure) for the 51 study participants.
Promoting a positive mindset
Five of the thirty individually focused studies implemented interventions designed to promote a positive mindset in participants including gratitude journaling,12 49 62 choir singing58 and professional coaching.43 Gratitude journaling interventions are designed to shift thoughts away from the negativity of stress and towards a more positive primary appraisal of a situation.12 Similarly, choral singing research has shown that participants who regularly engage, may experience cognitive benefits through the addition of social routine and meaningful activity to everyday life.58 The professional coaching intervention was also specifically implemented for participants to design and achieve goals, thereby strengthening personal resilience and mindset positivity.43 All three studies that implemented a gratitude intervention had large recruitment numbers with the smallest sample size being 102.12 Cheng et al12 implemented a three-arm randomised control trial and overall scored the highest on the MERSQI for quality and design. In addition to a gratitude and control group, Cheng et al12 employed a ‘hassle group’ to record only negative work-related events throughout the 4-week intervention. Postintervention, no significant interaction effects with time or other changes were found for the hassle group, suggesting it was indistinguishable from the control group.12 The gratitude group reported significantly less perceived stress, following the intervention, which was also documented at the 3-month follow-up. A reduction in depressive symptoms was also seen over time for the gratitude group compared with the control group. Participants in the study by Komase et al62 attended a single 50 min gratitude workshop and were then asked to send lists of three work-related things they were grateful for, three times a week for three consecutive weeks. Gratitude, self-efficacy, psychological distress and job performance improved significantly postintervention. The gratitude study by Sexton and Adair49 also reported significant improvements in both depressive symptoms (Center for Epidemiologic Studies Depression Scale (CES-D)) and subjective happiness (Subjective Happiness Scale (SHS)) following participants keeping a log of everything that went well for 15 days. However, it was also noted in this study, the extent to which participants expressed their gratitude was not known as participants could opt to share their diaries or not.
The study by Moss and O’Donoghue58 reported no significant improvement for any quantifiable outcome measure in this study, however, qualitative findings reported that a workplace choir can promote social connectedness, enjoyment at work and staff engagement.58 Dyrbye et al43 implemented professional coaching for 44 participants and assigned the same number of participants to a control group (no treatment). The proportion of participants in the intervention group with high EE, at 5-month follow-up decreased by 19.5% and increased by 9% in the control group. The intervention group also had a significant overall improvement in quality of life as measured by the Quality of Life Scale as well as resilience (CD-RISC) compared with the control group.
Individual job crafting
Finally, the second intervention group within the study by Gordon et al56 implemented individual job crafting to improve work engagement and lower burnout. Participants in the individual job crafting group reported improved work engagement (UWES) as well as significantly lowered levels of EE (OLBI) compared with the waitlist control group.56 Results were comparable to the first group who received a generalised job crafting intervention and therefore no variation in positive study outcomes was reported between the two job crafting treatment groups.56 A detailed breakdown of all study outcomes is outlined below in table 2.
As it is difficult to compare results to determine the overall impact of these studies, table 3 was created to highlight four positive attributes of the studies reviewed, partially informed by the MERSQI guidelines for quality studies. As demonstrated, only four4 12 26 60 studies reported statistically significant positive outcomes while implementing a large sample size, control group and postintervention follow-up.
Discussion
Overall, the studies included in this review scored moderate-highly on the MERSQI rating system yet were markedly heterogeneous pertaining to intervention types, study designs, outcome measures and sample sizes implemented. The literature suggests that organisationally focused well-being interventions may be more effective in promoting and maintaining worker well-being however, with only three studies falling under the organisational umbrella, this review cannot confirm or refute this notion. Each organisational intervention reported at least one statistically significant positive outcome, however only the study by Gregory et al26 conducted a follow-up beyond the finalisation of the intervention whereby one outcome (workload) had nearly reverted to the baseline measurement (2.5-2.99-2.62).
Both Gordon et al56 and Gregory et al26 also implemented waitlist style control groups and the final organisational peer support study (Wahl et al52) did not implement a control group at all. The literature reports growing concerns over the use of waitlist style control groups in psychotherapy research, as the outcomes of the interventions may be over-estimated with this type of control.64
Organisational intervention studies that target work conditions and create a preventative or primary approach to reducing occupational stress are more effective in promoting healthier workplaces yet are far more difficult to feasibly implement.21 Perhaps this explains why all but three of the intervention studies in this review used secondary, individually focused interventions designed to manage the effects of stress in individual workers. These interventions may be implemented for employers who are concerned about staff productivity and retention or to mitigate employee health costs, however, do not address any underlying contributors to occupational stress such as work conditions or environments.21 Three out of 26 studies implementing relaxation techniques, did not report a positively significant outcome nor did one study that promoted a positive mindset. Overall, MBE interventions were successful with all but two reporting significantly positive outcomes, however, out of 20 MBE studies, only 6 conducted follow-ups beyond the intervention period and only two implemented active style control groups.
Many of the studies in this review that implemented MBE interventions reported reductions in participant levels of burnout, perceived stress and anxiety as well as increased resiliency. However, there were limited validation factors implemented to suggest these improvements were the true result of the intervention itself (ie, comparison against an appropriate control group and follow-up conducted beyond the intervention period). Multiple systematic reviews conducted by Lomas et al22 23 achieved similar results with overall study findings reporting that mindfulness was generally associated with positive outcomes in relation to most measures, however, the quality of assessed studies was inconsistent, especially noting the lack of RCT study designs that were eligible for inclusion. Kriakous et al24 also found that mindfulness based stress reduction techniques were effective in reducing healthcare professionals’ experiences of anxiety, depression and stress but were less effective at reducing burnout or improving resiliency. MBE interventions can be implemented quite sustainably as they are generally performed in the workplace (during scheduled breaks) or through self-directed practices.22–24 Interventions that do not greatly disrupt daily productivity are more easily implemented for longer time periods and make longer term follow-up assessments more achievable.30–32 Despite this, most studies (n=20) in this review did not collect postintervention follow-up data and only one57 study conducted a long-term follow-up (12 months post the gratitude intervention), in which reported improvements were sustained. It is therefore difficult to suggest which interventions show long-term benefits and therefore may be more effective. With the acknowledgement that healthcare workers face especially difficult psychosocial hazards in the workplace, this review suggests that healthcare employees are likely to benefit from a stronger focus on well-being promotion in the workplace. However, with such a large volume of studies attempting to relieve symptoms of burnout by encouraging new health behaviours in employees themselves, rather than affecting change at an institutional level, further interventional research is demanded. Particularly in the form of RCTs with adequate long-term follow-up data collection.
Conclusion
The results of this review suggest that healthcare workers benefited from workplace well-being interventions, with a wide array of positive outcomes (improvements in well-being, work engagement, quality of life and mindfulness as well as reductions in burnout, perceived stress, anxiety and depressive symptoms) reported. Relaxation techniques targeted at the secondary interventional level (designed to manage stress in the individual worker) were the predominant well-being strategy of choice. The literature suggests this is due to the feasibility of implementing such a study. Reported positive outcomes included: improvement in EE, resilience, mindfulness, well-being, quality of life, and work engagement and/or reduction in burnout, perceived stress, anxiety and depressive symptoms. However, many of these reported positive outcomes, were somewhat diluted by the limitations of the various study designs that is, no control group, usage of a waitlist style control group and/or lack of postintervention follow-up surveys. Despite this, healthcare institutions can implement interventions pertaining to mindfulness-based education, promoting positive mindsets or organisational changes that are shown to illicit positive results in the well-being of their staff. Employers must also address the bigger picture for healthcare worker well-being. A short-term inconvenience, can cultivate an environment condusive to improved well-being and reduced burnout, thereby fostering long-term productivity and retention of staff.
Limitations and future research
Due to the large variation in study designs and data collection methods implemented, quantitative meta-analysis or comprehensive comparison of all study findings was not possible. Theoretical and practical implementations of the results are therefore complicated by the marked heterogeneity in intervention types, intervention timeframes, participant groups and conduction of postintervention follow-ups. This review was also limited by the lack of organisational studies that met the inclusion criteria. For example, several studies that implemented interventions involving the use of medical scribes or the electronic health record were excluded from this review as they did not focus on well-being as a primary outcome or did not use validated measures of well-being. Twenty-six of the thirty-three intervention studies reviewed opted to implement employee relaxation techniques to manage symptoms of occupational stress in the individual. With most studies adopting this one-dimensional view to employee well-being, it is difficult to accurately comment on the overall impact of well-being strategies for healthcare workers. Future research is needed to implement more primary or proactive organisational well-being interventions for healthcare workers, using robust study designs including appropriate control groups as well as further data collection beyond the conclusion of the intervention timeframe.
Data availability statement
All data relevant to the study are included in the article or uploaded as supplementary information. Other data are available on reasonable request to the corresponding author.
Ethics statements
Patient consent for publication
Ethics approval
Not applicable.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Contributors CC: primary contributor to the content within this manuscript, developed and implemented a search strategy, screening of articles, data extraction, quality assessment, synthesising of results. Article formulation, multiple draft revisions and preparing manuscript for submission. SP (cosupervisor): content expert regarding workplace well-being interventions and therefore played a pivotal role in shaping the direction of the paper relating to this content such as the introduction and discussion of results pertaining to the literature. Provided key assistance as to the flow of the manuscript through extensive grammatical and content revisions as the various drafts progressed. EB (primary supervisor): content expert relating to the processes of systematic reviews. Provided extensive guidance throughout text screening, quality assessment, data extraction and synthesis of results stages. As well as extensive feedback throughout multiple drafts during article formulation. MT (end-user advisor): content expert relating to the effects of workplace wellness in a clinical sense. Contributed to the idea for the article as well as guiding the search strategy (exclusion/inclusion criterium). Guidance and feedback throughout the duration of article formulation, as well as feedback provided for draft copies of the final manuscript. JC (cosupervisor and guarantor): second independent author who performed all stages of text screening, quality assessment and review of data extraction tables. Considerable contribution to synthesising of results as well as formulation of the final article.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.