Rural Healthcare Workers’ Well-Being: A Systematic Review of Support Interventions

Abstract

Although there is increased awareness about healthcare workers’ stress and burnout after the COVID-19 pandemic, support interventions should be tailored according to the needs of healthcare workers. Given the unique challenges rural healthcare workers face, we sought to systematically identify the types of interventions specifically designed and utilized to support the well-being of healthcare workers practicing in rural settings.

Methods:

We conducted a comprehensive search of the existing literature through electronic databases to identify quantitative, qualitative, and mixed methods studies describing supportive interventions for rural healthcare workers with well-being related outcomes between January 1 and March 31, 2023. We used the Effective Public Health Practice Project (EPHPP), Mixed Methods Assessment Tool (MMAT) and Joanna Briggs Institute (JBI) Critical Appraisal Checklist to evaluate the study quality.

Findings:

Out of 1583 identified records, 25 studies were included in the analysis. The studies described a wide range of supportive interventions and outcomes. The overall quality of the studies was weak to moderate. None of the studies were randomized and only six included controls. Included interventions were generally well- accepted. Quantitative and qualitative themes identified shared decision-making, effective supervision, and proactive cultural change as promising interventions that warrant further exploration. Financial interventions alone were not effective. Most of the studies were either unfunded or were funded internally by the institutions.

Conclusions:

There is limited research in support interventions for rural healthcare workers. Larger, well-designed studies are needed to explore promising interventions to promote wellbeing of rural healthcare workforce.

Introduction:

There is a shortage of healthcare workers (HCW) in rural areas. The United States Census Bureau considers areas with a population of less than 50,000 as rural (United States Census Bureau, 2022). Shortage of healthcare staff in rural areas has been a pernicious problem. For example, only about 10 % of the nation’s physicians practice in rural areas. Even though approximately 1 in 5 (18 %) of US population resides in rural areas which comprise of the 97 % of the US land area (Weinhold & Gurtner, 2014)

Globally, 50% of the world’s population lives in remote areas, but only 38% of nursing staff and less than 25% of physicians work in those areas (Kumar & Kumar, 2018). Rural populations tend to have higher rates of the elderly residents and chronic illness, and tend to be poorer (James et al., 2018). Rural patients often have to travel farther for their care and due to their occupations, many are reluctant to take time off from work (Buzza et al., 2011). As a result, they often delay seeking care. Rural areas have higher proportion of Indigenous people such as Native Americans with significant healthcare disparities (James et al., 2017) . All of these factors have an adverse effect on the health outcomes of the rural populations.

Rural healthcare workers also face unique challenges. Rural healthcare providers are asked to treat a wider range of conditions with limited access to sophisticated technology or subspeciality support. Many travel long distances to cover large geographic areas and perform a wide variety of procedures, often without specialized training. Many of them care for other HCW or their families. Despite these demands, they have limited access to support and training. They also have limited administrative support and career choices (Coombs et al., 2022) (O’Toole et al., 2010). Retention of rural healthcare workers is particularly challenging for healthcare organizations because of a number of factors such as shortage of HCW, higher workloads and limited resources (US Government Accountability Office, 2017). These challenges have intensified during the COVID-19 epidemic. Barriers such as restricted reimbursement policies and broadband availability have hindered telehealth adoption in some rural areas during the COVID-19 pandemic. (Hirko et al., 2020)

High rates of stress and burnout in healthcare workers have been reported consistently even before the COVID-19 pandemic. The 2019 National Academy of Medicine report found burnout had reached “crisis” levels, affecting up to 54% in nurses and physicians and up to 60% in medical students and residents (National Academy of Medicine, 2019). These trends appear to have worsened during the pandemic. For example, in a meta-analysis of 4,419 healthcare workers, Parandeh and colleagues reported that 82% of participants reported moderate or severe emotional exhaustion, which is a component of burnout (Parandeh et al., 2022). Previous studies suggest that HCWs might also be differentially impacted by the pandemic depending on whether they practice in rural, regional, remote or metropolitan settings. Rural HCW tend to have higher rates of depression, anxiety, stress, post-traumatic stress disorder (PTSD) and burnout as compared to HCW in urban settings (Tham et al., 2022).

Despite increased research and awareness about healthcare workers’ stress and burnout during the COVID −19 pandemic, there is a lack of clear, evidence-based intervention focused on HCW mental health and well-being. (Harkanen et al., 2023) A comprehensive scoping review of 10,529 published references found only 7 prospective intervention studies during COVID-19, none of which were high-quality or theory-based (Cairns et al., 2021). Moreover, most of the support interventions are directed towards HCW in urban areas. This is because these interventions tend to be developed at large tertiary medical centers and require substantial resources for implementation, which are typically lacking in rural settings. In addition, these interventions generally fail to address the unique contributors of professional stress and burnout experience by healthcare workers in rural settings. Therefore, the purpose of this systematic literature review was to identify the types of interventions specifically designed and utilized to support the well-being of healthcare workers practicing in rural settings and evaluate their effectiveness and limitations.

Methods:

This was a systematic review of the published literature. We conducted this study in accordance to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) (Page et al., 2021). Prior to conducting this review, a protocol was developed and registered with PROSPERO (International prospective register of systematic reviews, record CRD42023382408).

Eligibility criteria

We included papers that reported original, published, quantitative, qualitative or mixed methods research describing rural healthcare workers’ well-being in the English language. We included studies on all healthcare professionals, including students and trainees located in rural settings. We included any study design (e.g., observational, cohort, case-control, randomized controlled trials, focus groups) that included an intervention to support healthcare workers in the rural health care setting and provided outcome/evaluation data. The eligibility criteria for our review are listed in Table 1.

Table 1:

Inclusion and Exclusion Criteria

This table provides inclusion and exclusion criteria to determine eligibility

Targeted Studies	Quantitative research (randomized controlled trials, controlled clinical trials, controlled before and after studies, uncontrolled before and after studies), qualitative research, systematic reviews, meta-analyses
Inclusion Criteria	Includes intervention for Healthcare Workers Studies include quantitative or qualitative evaluation of the intervention Study outcomes are related to healthcare wellbeing directly or indirectly Study location identified as rural setting Accessible in full text form Studies in English language
Exclusion Criteria	Studies not involving Healthcare Workers in rural settings Studies not describing an intervention Studies not reporting on wellbeing-related outcomes No abstract for review Article is a commentary, opinion piece, or editorial

Search strategy

Between January 1 and March 31, 2023, the following databases were searched without time and study design limit: Medline, Embase, PsycINFO, Web of Science, CINAHL, Emerald, SCOPUS databases, Google Scholar, ProQuest Dissertations and Theses Global, American Doctoral Dissertations, Open Access Theses and Dissertations. The searches were undertaken using a combination of Medical Subject Headings (MeSH) and Boolean operators, using keywords such as burnout, stress, anxiety, depression, rural vs. urban, peer support, hospital, ambulatory care, primary care clinic, health professionals, evaluation, wellness, staff well-being, critical incident, traumatic stress etc. We also hand-searched the reference lists of eligible articles to identify any additional citations. The search strategy is summarized in Table 2.

Table 2:

Search Strategy

Examples of Search Terms Used

Mentor + health + rural OR remote Professional support + health + rural OR remote Second victim + rural + support + healthcare worker Supervision + health + rural OR remote Professional development + health + rural OR remote Continuing professional education + health + rural OR remote Continuing medical education + health + rural OR remote Preceptorship + health + rural OR remote Medical + supervision + rural OR remote Allied health + rural OR remote

Electronic sources

Medline, Embase, PsycINFO, Web of Science, CINAHL, Emerald, SCOPUS databases, Google Scholar, ProQuest Dissertations and Theses Global, American Doctoral Dissertations, Open Access Theses and Dissertations.

Data extraction

After removing duplicate articles, two sets of reviewers (MM, MP and MM, HB) screened titles and abstracts of retrieved papers for relevance. The text of the selected articles was then reviewed in entirety by the two independent reviewers. Any discrepancies in the study inclusion were resolved through discussion between the reviewers. The two reviewers extracted data from the selected papers onto a pre-designed data extraction template. We used online systematic review software (DistillerSR, Evidence Partners, Ottawa, Canada; https://www.evidencepartners.com) to facilitate literature screening.

Data synthesis

Due to the heterogeneity in the study designs, settings, objectives, interventions, measurement instruments and outcomes measured, it was not possible to conduct a meta-analysis. Therefore, a narrative synthesis was completed. A spreadsheet was used to capture relevant variables for each study question. Key points were then identified, highlighted and used to generate themes, using data-based convergent synthesis method (Hong et al., 2017). In this method, the quantitative and qualitative evidence is collected and analyzed during the same phase of the research process in a parallel or a complementary manner. Both sets of reviewers contributed to data extraction and synthesis.

Quality assessment

Included studies were categorized as quantitative (randomized, non-randomized and descriptive), mixed methods (combining quantitative and qualitative approach) or qualitative.

For quantitative studies, quality was assessed using the Effective Public Health Practice Project (EPHPP) tool (Effective Public Healthcare Panacea Project, 2024). EPHPP has been shown to have good inter-rater reliability across a variety of quantitative study designs (Armijo-Olivo et al., 2012). Studies were assessed on: (1) selection bias, (2) study design, (3) confounders, (4) blinding, (5) data collection methods and (6) withdrawals and dropouts. Components were scored as 1 (‘strong’), 2 (‘moderate’), or 3 (‘weak’). EPHPP guidelines were used to generate a global score as follows: no ‘weak’ component ratings= ‘strong’, one ‘weak’ component rating= ‘moderate’ and two or more ‘weak’ component ratings= ‘weak’.

For mixed methods studies, Mixed Methods Assessment Tool (MMAT 2018) was used (Hong et al., 2018). MMAT version 2018, designed for the assessment of five different study types (qualitative research, mixed methods research, quantitative descriptive studies, randomized controlled trials, and non-randomized studies), consists of a series of screening questions that can be answered as ‘yes’, ‘no’, or ‘can’t tell’. MMAT comprises of two screening items for mixed methods research, followed by 25 appraisal items in five sections, including: (1) five items on the qualitative component, (2) five items each on the quantitative components (i.e., randomized controlled, non-randomized, or descriptive), and (3) five items on mixed methods. In accordance with the user guide, only the sections relevant to a particular study were rated.

For qualitative studies, the Joanna Briggs Institute (JBI) Critical Appraisal Checklist for qualitative research (2017) was used (Joanna Briggs Institute, 2023). This tool consists of 10 items, with four response options: yes, no, unclear or not applicable.

Two sets of reviewers independently assessed all studies. Differences were resolved through discussion to achieve consensus. Cohen’s kappa (Cohen, 1960) was calculated to determine inter-rater reliability.

Results:

The literature search identified 1583 articles published before March 1, 2023, from which 141 duplicates were removed. After screening titles and abstracts, 85 articles were eligible for full-text assessment, of which 28 were included (Figure 1). An additional 3 articles were excluded after review. Two excluded studies described implementation of educational strategies (telemedicine and clinical supervision) but did not include any well-being related outcomes. The third excluded study reported a task-shifting program for village health volunteers in rural Thailand and did not include any data on health care workers. There was moderate agreement for the quality assessment at the full-text screening stage between reviewers (κ = 0.60).

Figure 1:

PRISMA: Preferred Reporting Items for Systematic Reviews and Meta-Analysis. This figure illustrates the process of searching for relevant studies from the databases including identification, selection, eligibility, and inclusion of the studies depending on the predefined criteria.

Study characteristics:

25 studies were included in the final analysis, which are listed in Table 4. Of these, most (70%) were conducted in three countries – the United States, Canada, or Australia. Of the included studies, 20 studies were peer-reviewed published articles, 4 were doctoral theses and one was a web report.

Table 4:

Methodology, Findings, and Appraisal of the Included Studies

This table summarizes demographics of the population, study design, interventions, outcomes and quality appraisal for the included studies

MMAT: Mixed Method Appraisal Tool, JBI: Joanna Briggs Institute Critical Appraisal Checklist, EPHPP: Effective Public Health Practice Project

Authors	Study Title	Population and Participant Demographics	Study Design	Intervention	Measurement Instruments	Outcomes	Limitations	Lessons Learnt	Attrition rate	Quality Appraisal
Adam et al 2021 Funding: None reported	Coping with COVID: Developing a Rapid-cycle Frontline Quality-improvement Process to Support Employee Well-being and Drive Institutional Responsiveness in a Tertiary Care Faith-based Hospital in Rural Kenya.	122 self-selected clinical and non-clinical staff at a tertiary hospital in rural Kenya No Control Group No demographics reported beyond professions	Qualitative	17 focus group discussions over two-week period exploring coping skills and support needs	“Sticky notes” from participants	The five most popular personal coping strategies reported were prayer, reading the Bible, listening to music, watching movies, and interacting/sharing with family members.	Thematic analyses using sticky notes without complete recordings and transcript analyses	Examination of coping skills can help healthcare staff adopt and learn new skills	Not reported	Reliability: Weak Risk of Bias: High Overall Quality:7/10 Instrument used JBI
Chitwood 2019 Funding: None reported	Second Victim: Support for the Healthcare Team	not-for-profit healthcare organization based in rural Virginia 307 employees referred to the TRUST Team from sources such as first responders, SafeWatch, supervisors, the office of quality and patient safety, or self-referred after viewing a brochure Demographics reported were age (63% between 25 and 45), profession and educational level	Quantitative descriptive Post Implementation evaluation	The TRUST Team (TTT) program Details of program components not provided	pre-post Second Victim Experience and Support Tool (SVEST)	Generally positive evaluations	TTT was implemented in fall 2014 and the program evaluation occurred in December 2018 38 out of 307 employee referrals answered the survey and only a small sample size of 16 respondents had used TTT services. Thesis report Not peer reviewed	A focus on staff nurses with more than five years of experience would be helpful to improve utilization of services and support for all healthcare team members.	Not reported	Reliability: Weak Risk of Bias: High Overall Quality: Weak Instrument used: EPHPP
Craigie et al 2016 Funding: Nursing Executive Committee at the study hospital	A pilot evaluation of a mindful self-care and resiliency (MSCR) intervention for nurses	Teaching hospital in Western Australia Twenty-five nurses expressed interest in the MSCR program, and of these, 24 met inclusion criteria. After final scheduling of the intervention groups, a total of 21 nurses commenced the program. Demographic data reported included qualifications, managerial level and employment type	Quantitative descriptive Pre and Post evaluation	1-day compassion fatigue prevention educational workshop, followed by a series of weekly mindfulness training seminars conducted over 4 weeks (12 h total intervention time). Participants completed a number of standardized measures at pre, post, and 1-month follow-up	The Patient Health Questionnarie-9 The Professional Quality of Life Scale version 5 Depression Anxiety Stress Scales Spielberger State-Trait Anxiety Inventory Passion for Work Scale	pre-intervention, 45 % of the sample had high burnout scores, but this reduced to just 15 % by post-intervention. No significant changes were observed for general resilience, anxiety, or secondary traumatic stress post-intervention or at follow-up.	No control group, non-randomized self-selected group. Participants were experienced nurses, thus limited generalizability	relatively brief workplace intervention may represent a feasible approach to improving resilience and well-being among nurses.	12 % 21 out of 24 participants completed the study	Reliability: Moderate Risk of Bias: High Overall Quality: Moderate Instrument used: EPHPP
Gardiner et al 2013 Funding: The Rural Doctors Workforce Agency	Effectiveness of cognitive behavioral coaching in improving the well-being and retention of rural general practitioners.	69 self-selected GP’s working in rural Australia Control group consisting of 205 rural GP’s Only demographic data reported is sex distribution (32 % in the coaching group, 29 % in control group)	Quantitative Non-randomized Case Control Longitudinal Follow-up	9-hour initial cognitive behavioral coaching program advertised as a work–life balance, followed by 6 weeks of email coaching	Pre and post questionnaire (3–42 months after the program) Rural doctor Distress Scale 10-item (75 % of participants showed some distress reduction) Doctors’ intention to leave rural general practice, 1–7 Likert Scale 3-year Retention rate	Statistically significant improvement on all items of RDD scale Modest but statistically significant reduction in Intention to leave rural general practice At 3 years, 94% of participants stayed compared with 80% of the general rural doctor population	Self-selection Of intervention group Matching variables not described Higher dropout rate in control group	Highlighting positive aspects of rural practice, fostering self-management skills and make attitudinal changes and improve wellbeing and retention	37 % In intervention group 41 % in control group 40 out of 48 completed at follow up could be matched to those completed before the intervention, leaving an ultimate response rate of 63%. control group consisting of 205 of the 440 GPs working in rural South Australia (51% response rate)	Reliability: Moderate Risk of Bias: High Overall Quality: Moderate Instrument used: EPHPP
Gardiner et al. 2006 Funding: Rural Doctors Workforce Agency.	Impact of support initiatives on retaining rural general practitioners	221 rural GPs South Australia male (71.0%), with 74.8% aged between 30 and 50 years	Qualitative Evaluation of DrDOC program after two years on implementation	peer-support network for GP’s, visiting health checkups for rural doctors and their families, crisis plans for distressed GP’s, an emergency support line, rural retreats for GP’s	Intention to leave rural practice Quality of work life-6 item Likert scale Rural doctor distress-10 item Likert scale	Intention to leave rural practice in the short to medium term (from 30% to 25%) 10 % increase in willingness to discuss personal issues Social isolation down from 16 % to 10.2%	No control group Self-reported measures Actual retention of GPs not tracked.	Overall modest improvements. Improving psychological well-being might influence rural GPs’ intentions to leave rural practice	43 % Eighty-seven of the original 187 at time 1 who completed the questionnaire at time 2	Reliability: Weak Risk of Bias: High Overall Quality:5/10 Instrument used: JBI
Gorsche and , Woloschuk 2012 Funding: Rural Physician Action Plan of Alberta.	Rural physicians’ skills enrichment program: a cohort control study of retention in Alberta	29 physicians working in rural Alberta with a matched cohort median duration of enrichment training was 5 weeks 5 years follow up Age range: 31 to 80 years (M = 47), 31 % females, 28 % foreign medical graduates	Quantitative non-randomized Longitudinal, matched, case control study and program evaluation	Self-identified personal skills training program with 3 aims	Retention	All 29 program participants stayed in rural areas at 5 year follow up as compared to only 22 of 29 matched control: relative risk 1.31, confidence interval 1.06–1.62 P < 0.05.	Self-selected participation Important variables such as gender omitted in matching Duration of additional skills training varied, and any skills training completed by the cohort group outside of the formal EP was not recorded.	additional skills training by established physicians can improve retention in rural practice	0 % All 29 program participants stayed in rural areas at 5 year follow up	Reliability: Moderate Risk of Bias: High Overall Quality: Moderate Instrument used: EPHPP
Healey-Ogden et al 2012 Funding: Interior Health Authority Canada	British Columbia: improving retention and recruitment in smaller communities.	14 nurses on the pediatric unit a rural hospital, British Columbia. 8-month pilot Demographic data not provided	Qualitative Implementation Of a staff support program	80/20 program: 20% of their salaried time off from direct patient care in order to pursue various types of professional development activities.	Participant interviews	50 % stated that because of their experiences in the 80/20 project, they would definitely remain working at the rural settings	Findings based on single interviews Focus groups not used Website report, not peer reviewed	Participation in employer-supported professional development can lead to the retention of nurses	10% Team collected and analyzed qualitative data from individual phone interviews with 14 of the 15 participating nurses.	Reliability: Weak Risk of Bias: High Overall Quality 4/10 Instrument used: JBI
Hsu et al 2021 Funding: None reported	Effects of Zentangle art workplace health promotion activities on rural healthcare workers.	40 healthcare workers in a rural hospital in Taiwan No Control Group Median age 32 (23–41.75), 70 % had university education and 65 % were single.	Quantitative descriptive	One 5-part session of Zentangle art, encouraging meditation through drawing	Brief Symptom Rating Scale-5 Work stress management effectiveness self-rating scale-20 item Likert scale General Self-Efficacy Scale- 10 item Likert scale Workplace Spirituality Scale-28 items	Statistically significant improvement in BSRS-5, stress management and self-efficiency	No control group or randomization Follow up period not defined Small sample size with high turnover	Single exposure to art therapy can relieve stress and improve wellbeing	Not reported	Reliability: Weak Risk of Bias: High Overall Quality: Weak Instrument used: EPHPP
Kanste et al. 2010 Funding: Hospital district and Finnish Cultural Foundation	Effects of network development on attitudes towards work and well-being at work among health care staff in northern Finland.	14 health centers in sparsely populated and rural areas, as well as 4 hospitals in Northern Finland. Mean age was 43.7 in control group and 43.3 in interventional group All participants were female. 78 % were nurses	Quantitative non-randomized quasi-experimental design with intervention (n=33) and control (n=23) groups. 2 years	5 main elements: regional networking, self-ruling teamwork, collaborative learning, staff education and guidance.	organizational commitment, occupational commitment (4 items from Organizational Commitment Questionnaire (OCQ) and 4 items from the Organizational Commitment Scale (OCS) job involvement (3 items from Kanungo’s Job Involvement Questionnaire (JIQ) growth satisfaction and internal work motivation; and well-being (Maslach Burnout Inventory)	statistically significant improvements in attitudes towards work No effect on job involvement and motivation no statistically significant effect of the intervention on vigor, affective well-being and mental resources	Small size Much higher dop out rate I the control group intent-to-treat analysis not used	regional network development between health centers and hospitals can help with improving work attitudes	For the second round 41 % in intervention group and 60 % in the control group Only 34 (59%) responded in the intervention group and 23 (40%) in the control group.	Reliability: Moderate Risk of Bias: High Overall Quality: Moderate Instrument used:EPHPP
Kirkland 2022 Funding: None reported	Using Peer Support Groups as an Innovative Approach to Improve Wellness and Self-Care Among Community Health Workers	9 CHW’s from rural Ohio Participant demographics not reported	Qualitative Survey evaluation	HANS KAI pilot program, which consisted of education, Physical Activity, wellness or self-care education, and a healthy snack 60 min biweekly virtual meetings for 4 months	Physical activity input from vívoactive HR watches Subjective questions about burnout	Data not analyzed due to small number	Thesis report, not peer reviewed Very small sample size	There is a continued need for HCW-specific support programs	Not reported	Reliability: Weak Risk of Bias: High Overall Quality: 3/10 Instrument used: JBI
Kok et al 2018 Funding: European Union through the Seventh Framework Program	Does supportive supervision enhance community health worker motivation? A mixed-methods study in four African countries.	1689 Community Health workers and supervisors in rural areas of 4 sub-Saharan countries Ethiopia, Kenya, Malawi and Mozambique 1 year follow up No Control Group Participants were primarily female (60/63) with an age range from 20 to 72 years	Mixed Methods Quantitative non-randomized and qualitative	5–6 days of supervision training followed by monthly individual and group supervision	278 CHW pre and post questionnaires Motivational outcomes were assessed by using a 12-item, self-reported measure, adapted from the Motivational Outcome Scale The six-item Perceived Supervision Scale-5 point Likert Scale Participant interviews	Significant changes in job satisfaction scores were observed, but not in organizational and community commitment Qualitative analysis showed strong improvement in motivation and supervision	qualitative and quantitative findings differed Improvement was not substantiated by observed changes in motivation-related outcomes	supportive supervision training can contribute to improved job satisfaction and motivation	Not reported	Reliability: Weak Risk of Bias: High Overall Quality: 2/5 Instrument used: 2/5
Krasner et al. 2009 Funding: Physicians Foundation for Health Systems Excellence	Association of an educational program in mindful communication with burnout, empathy, and attitudes among primary care physicians.	70 primary care physicians in Rochester, New York from 871 invited 54 % were males, 49% Internal medicine physicians, mean experience 15.69 years	Quantitative Non-Randomized Before-and-after survey No control group	Peer support, mindfulness awareness exercises, narratives about meaningful clinical experiences, appreciative interviews, didactic material, and discussion	5 sets of self-administered surveys 2-Factor Mindfulness Scale Maslach Burnout Inventory Physician Belief Scale Profile of Mood States (POMS)	Completion rate ( 60) 86% Participation in a mindful communication program was associated with short-term improvements in well-being and attitudes associated with patient-centered care	Non-randomized self-selection	Rural practitioners less likely to practice in research Model of CME delivery May contribute to professional satisfaction and wellbeing	14 %	Reliability: Moderate Risk of bias: High Overall Quality: Moderate Instrument used: EPHPP
Lam et al. 2011 Funding: Not reported	Telephone-administered cognitive-behavioral therapy for clients with depressive symptoms in an employee assistance program: a pilot study.	39 Self-referred employees of Interior Health Authority in rural British Columbia to employee assistance program average age 45.1 ± 8.9 years. 80 % were female	Quantitative non-randomized pre and post intervention No control Group	8-session telephone-administered CBT program	9-item Personal Health Questionnaire (PHQ-9), Global Assessment of Functioning (GAF), and clinician ratings of work absence and performance impairment.	79 % completion rate.significant improvement on the PHQ-9 from a mean score of 13.11 ± 4.87 at baseline to 4.57 ± 3.54 after the intervention (t = 12.35; df = 27; P < .0001) Significant improvement was also found with the GAF score and the performance impairment measure, but not with work absence	No control group no checks on fidelity of the manualized CBT program.	Support interventions can be remotely delivered effectively to rural health care employees	8 of 39, or 21%	Reliability: Moderate Risk of Bias: High Overall Quality: Moderate Instrument used:EPHPP
Lee 2014 Funding: None reported	Implementation of a second victim program: HOPE Team	68 nurses identified from reported medical errors in a rural hospital in North Carolina Demographic data not reported	Quantitative descriptive Pre-implementation survey	Second victim support program, it was started but not completed Reportedly provided emotional support for the second victim by assisting the employee in managing responses that might threaten personal and professional identity, and cause the nurse to leave the profession.	Medically Induced Trauma Support Services (MITSS) Staff Support Survey 69 items	66.6 reported medical errors during last year 60 reported no formal or informal support	Only 6 participants completed the evaluation. Due to the low response on the survey, the researcher could not make a correlation using the data collected. The researcher obtained a total of six anonymous responses Thesis report Not peer reviewed	Need for cultural change in which an error is looked upon as an opportunity for improvement and not blame or shame will require leaders to reconsider how power is perceived and used in the organization	Not reported	Reliability: Weak Risk of Bias: High Overall Quality: Weak Instrument used: EHPP
Majee et al 2019 Funding: None reported	A Self-Management Training Intervention: Perceptions and Practices of Community Health Workers in South Africa.	20 CHW’s from two rural communities of South Africa’s Western Cape region Men age: 41, 95 % females, 30% college education, 100 % identified as colored.	Qualitative	Act Healthy Self-Management Groups- 50 minutes once a week as part of a 6-week series, co-led by trained coworker volunteers. Group meetings focus on defining a health goal, developing a weekly action plan, reporting on the previous week’s action plan, problem solving to create an action plan the participant is confident in achieving, and sharing health resources.	Semi-structured interviews	CHWs felt empowered to change their health behavior by skills such as goal setting and action planning, and by growth in self-awareness and confidence.	Study conducted only pos-training interviews. All participants except one were female some of the participants were not fluent with English and that could have caused a communication barrier	Self-management training can lead to healthy behavior change for healthcare workers	5 % 19 (95%) out of 20 elected to participate	Reliability: Moderate Risk of Bias: High Overall Quality: 4/10 Instrument used: JBI
Mitchell et al. Year of publication 2000 Funding Source:Australian Government’s Rural Health Support, Education and Training Program (RHSET)	An evaluation of a network for professional development in child and adolescent mental health in rural and remote communities.	A network to provide professional development for mental health staff in rural and remote parts of South Australia and the Northern Territory 45 participants, who ranged from psychiatrists, psychologists, nurses, social workers and family therapists Participant demographics not reported	Mixed Methods Quantitative descriptive and Qualitative Outcomes of the first 12 month	A total of 36 telemedicine sessions were held, ranging in duration from 45 to 90 min (average 56 min) The most common types of session were case discussions (47%), followed by specialist seminars (36%) and administrative and introductory sessions (17%).	Activity log, 14-item questionnaire, interviews and action research (identify concepts that might be explored as the network development).	29 (60%) participants Completed survey The ability to access second opinions; the ability to access specialists highlighted.	Non-validated measures Self-reported survey	Not long enough to affect staff retention and recruitment Barriers included cost, technical support, induction and training	Not reported The number of sessions per participant ranged from 1 to 15	Reliability: Weak Risk of bias: High Overall Quality: 2/5 Instrument used: MMAT
Olcoń et al 2022 Funding: National Health and Medical Research Council	A Narrative Inquiry into the Practices of Healthcare Workers’ Wellness Program: The SEED Experience in New South Wales, Australia.	33 HCW’s (44 invited) from a rural hospital in Australia Convenience sample Mean age: 49.9 (32–65) 88% females, 79 % Caucasians.	Qualitative	Hospital staff wellness program named SEED (Stability, Encompassing, Endurance and Direction) Interventions included weekly wellness sessions, communal journal, quiet room and planned coffee breaks etc.	110 to 120 min focus groups Semi-structured interviews	eight themes were identified: (1) responsive and compassionate leading, (2) co-designing wellness activities with staff, (3) listening to understand, (4) creating a safe and healing space, (5) connecting with others, (6) collective caring, (7) diversifying and localizing wellness activities, and (8) striving for sustainability.	Timeline of SEED program and follow up evaluation not described participants were largely female and White study did not formally evaluate the outcomes of the SEED program.	Local context and needs have to be incorporated in developing support programs	Not reported	Reliability: Weak Risk of Bias: High Overall Quality :6/10 Instrument used: JBI
Rees et al 2020 Funding: Central Highlands Health and Central Queensland University, Australia	Pilot study of the effectiveness of a Mindful Self-Care and Resiliency program for rural doctors in Australia.	13 GP’s in rural Australia 1 month follow up after training No control group Mean age: 40 (range: 22–64) 23 % females Experience range was from 1 to 39 years	Mixed Methods Quantitative descriptive and qualitative	4-hour face-to-face session and three 1-hour video-conference follow-ups mindfulness and self-care training	10-item Burnout Measure 10 items Positive and Negative Affect Scale Well-being Index General Health Questionnaire-12 Participant Interviews	7 participants completed all evaluations, 4 only did qualitative interviews Reductions in burnout, psychological strain and negative affect Qualitatively participants gained new skills: self-awareness, reflection and self-care	Small sample size Self-selection bias No control group Low completion rate	Mindfulness training may improve stress in rural healthcare providers	46% 7 out of 13 participants completed all evaluations	Reliability: Weak Risk of bias: High Overall Quality :2/5 Instrument used: MMAT
Slatyer, et al 2018 Funding: Nursing Executive Committee at the study hospital	Evaluating the effectiveness of a brief mindful self-care and resiliency (MSCR) intervention for nurses: A controlled trial	Teaching hospital in Western Australia 91 nurses participated in the study (n = 65 intervention condition; n = 26 control condition 67 % females, 50% married and 60% Australian nationals	Quantitative non-randomized Case-Control pre-test, post-test and 6-month follow-up	full-day educational workshop comprising four sessions of about 1.5 h each, and three follow-up sessions of 1.75 h each held weekly (total duration of 11.5 h)	Professional Quality of Life Scale version 5 Depression Anxiety Stress Scales WHO (Five) Well-being Index [WHO Five] Self-Compassion Scale-short form [SCS-SF] General Self-Efficacy Scale [GSES]	significant reductions in burnout and depressed mood upon completion of the MSCR. Follow-up data indicated that these reductions persisted at 6-month post-MSCR for the intervention group	post-test between-groups model estimates were underpowered participants were not randomly allocated No control comparison data at the 6-month follow-up	significant and persistent improvements in burnout after relatively brief workplace intervention	16 % 76 of 91 participants completed the study	Reliability: Moderate Risk of bias: High Overall Quality: Moderate Instrument used: EPHPP
Teasley et al. 2007 Funding: none reported	Improving work environment perceptions for nurses employed in a rural setting	Convenience sample of Nurses in rural Kentucky Hospital Baseline data collection occurred in 2003 with a sample of 31 nurses and was repeated following implementation of study interventions in December 2004 Demographic data not reported	Quantitative Non-Randomized	4 strategies chosen for implementation, based on nursing staff feedback (1) a shared decision making or governance model; (2) enhanced role of licensed practical nurses (LPNs) within the organization; (3) augmentation of administrative support on night and weekend shifts (4) utilization of wireless communication devices	Individual workload perception scale, 21-item Likert Scale tool with five dimensions: Nurse Manager Support, Team Support, Workload, Organizational Resources and Intent to Stay.	Statistically significant improved perceptions of workload and intent to stay were reported for registered nurses and individuals with more than 11 years in nursing Most improvement in night shift nurses	recruitment primarily driven by nursing leadership, sample was small and within 1 institution, convenience sampling differences in the individual nurses participating in the pre- and post-assessment No controls	A proactive strategy encouraging a healthy work environment for experienced rural nurses seems like a far more logical choice than enduring the lengthy delays in recruiting new nurses to fill vacancies.	Not reported The respondents in the baseline and follow-up surveys were not all the same individuals	Reliability: Weak Risk of Bias: High Overall Quality: Weak Instrument used: EPHPP
Thornton 2022 Funding: None reported	Implementing Code Lavender in a Rural Acute Care Hospital: A Quality Improvement Project	convenience sample of 18 nurses from a rural Louisiana hospital Demographic data not reported	Quantitative descriptive Pre- and post-intervention surveys	Code Lavender is a multidisciplinary team-based intervention developed by the Cleveland Clinic to support nursing staff members during periods of personal stress, mental or emotional duress, or unresolved internal moral conflicts	Brief Resilience Scale (BRS) and the Professional Quality of Life (ProQoL) to measure resilience and compassion fatigue	A statistically significant increase in nurse resilience was noted after 11 weeks of the utilization of the Code Lavender program. There was not a statistically significant decrease in compassion fatigue over the same time.	For the sample to have statistical reliability, the sample required 34 participants. This project had 18 participants. Thesis report Not peer reviewed	The involvement of nurse scientists, DNP and hospital administration is essential for developing nurse-tailored residence building programs	Not reported	Reliability: Weak Risk of Bias: High Overall Quality: Weak Instrument used: EPHPP
Vesel et al 2015 Funding: Bill & Melinda Gates Foundation	Psychosocial support and resilience building among health workers in Sierra Leone: interrelations between coping skills, stress levels, and interpersonal relationships.	271 community health workers And nurses from rural districts in Sierra Leone 157 control health workers Demographic data not reported	Mixed Methods Quantitative non-randomized and Qualitative	Individual intake counseling assessment Ten group counseling sessions (gender segregated) Training on three core skill sets – stress management, self-care and client-care additional individual counseling for high-risk individuals	Perceived stress on a four-point Likert scale Coping skills communication (four variables), self-care (five variables) and social connectedness (six variables) on a 4-point Likert scale Relationships between health workers and their supervisors, on a 4-point Likert scale Interviews and Focus Group Discussions	Statistically significant differences in pre-post coping skills as compared to controls but less improvement in stress level and relationships with supervisors and patients Main themes from qualitative analysis: perspective-taking, empathy, mutual accountability, and work-life balance.	Baseline data not collected, a retrospective assessment in the intervention area carried out after the intervention was delivered Non-randomized allocation Matching variables not described	Improving social relationships with supervisors and patients can result in reduction of stress	Not reported	Reliability: Weak Risk of Bias: High Overall Quality: 3/5 Instrument used: MMAT
White et al 2022 Funding: None reported	The implementation of a stress management program for health care workers through a rural occupational health clinic	33 HCW’s (from 213 screened through elevated scores on the Perceived Stress Survey) from the occupational health center of a rural healthcare organization in Illinois. 26 completed the intervention Demographic data not reported	Quantitative descriptive assessment of non-randomized allocation to a support program No control group	Stress interventions: relaxation room, yoga classes, educational sessions	Program Evaluation Forms (Not described)	Effectiveness scores improved from 33% baseline to 64% at the project end,	Measurement Instruments not described Small sample size Predominantly Caucasian participants	Shared decision making can assist HCWs in developing a stress management plan the worker felt they understood and could use to manage or reduce their stress.	21 % 26 out of 33 HCW’s completed the intervention	Reliability: Weak Risk of Bias Overall: High Quality: Moderate Instrument used: EPHPP
Wright et al 2012 Funding: Department of Human Services (Victoria)	Supporting international medical graduates in rural Australia: a mixed methods evaluation.	17 international medical graduates working in rural hospitals in Australia mean age :35 (27 to 51 years) Participants came from 12 countries	Mixed Methods Qualitative and non-randomized quantitative Implementation of a supportive educational program	5 educational sessions over 3 months. Simulation-based training to address clinical knowledge, attitudes and skills Audiovisual review of performance Website offering	Participants’ evaluation of the program pre- and post-program 15 item multisource feedback	15 (88%) completed the program overall MSF scores showed no statistically significant changes Participants self- reported increased knowledge, skills and professionalism	Sample is small and highly selective Non-standardized self-assessment Limited generalizability	IMG’s working in rural health settings face unique challenges, connecting regional clinicians and the medical school can be an effective intervention	11 % 15 out of 17 participants completed the program	Reliability: Weak Risk of bias: High Overall:2/5 Quality Instrument used: MMAT
Zimmerman, et al 2015 Funding: Government of Nepal and the Nick Simons Institute	A staff support program for rural hospitals in Nepal.	7 hospitals in rural Nepal selected 20 family doctors through an entrance exam and offered them compulsory service scholarships for 3 years Penalty for leaving early 34 control hospitals Demographic data not reported	Quantitative Descriptive Evaluation of Program implementation	Program doctors were paid three times the government salary personal, professional and management support	changes in hospital use as a proxy indicator for community satisfaction and quality of care.	5-year follow up: Mean annual admissions and outpatient visits per hospital almost doubled (1.5 times in control hospitals) Modest increase in deliveries and C-sections as compared to the control hospitals 5 of 20 doctors paid penalty to leave early	The 34 control hospitals with complete data also showed increased use between 2006 and 2013 High cost on implementation US$ 66 387 per hospital per year Self-selected group	Financial incentives for compulsory service can attract most talented healthcare workers to rural practice	25 % 5 of 20 doctors paid penalty to leave early	Reliability: Weak Risk of Bias: Highs Overall Quality: Moderate Instrument used: EPHPP

There was a wide geographic variability, 7 studies were conducted in the US, 8 in Australia, 3 in Canada and the remainder in Finland, Nepal, Kenya, South Africa, East Africa. Kenya and Taiwan. None of the studies utilized randomized allocation of participants or blinding. Only 6 studies had control groups. 14 studies used quantitative measures, 5 used mixed methods approaches that combined qualitative and quantitative data. The rest of the studies (6) utilized qualitative approach. The characteristics of included studies are summarized in Table 3.

Table 3:

Characteristics of Included Studies

This table identifies and summarizes key characteristics of the included studies.

Types of Studies	Quantitative	14 (56%)
	Qualitative	6 (24%)
	Mixed Methods	5 (20%)
	Peer Reviewed Articles	20 (80%)
	Thesis Reports	4 (16%)
	Website Reports	1 (4%)
Geographic Location	USA	7 (28%)
	Australia	8 (32%)
	Canada	3 (12%)
	Finland	1 (4%)
	Nepal	1 (4%)
	East Africa	1 (4%)
	South Africa	1 (4%)
	Sierra Leone	1 (4%)
	Kenya	1 (4%)
	Taiwan	1 (4%)
Settings	Hospital Setting	15 (60%)
	Community Setting	10 (40%)
Study participants	Physicians	7 (28%)
	Nurses	7 (28%)
	Physicians, nurses and allied health professionals	7 (28%)
	Community Health Workers	4 (16%)
Types of Interventions	Emotional Support	11 (44%)
	Improving Work Environment	4 (16%)
	Professional Development	3 (12%)
	Work Life Balance	2 (8%)
	Second Victim Programs	3 (12%)
	Effective Supervision	1 (4%)
	Financial Support	1 (4%)
Outcomes used	Psychological Measures	15 (60%)
	Attitude towards work	5 (20%)
	Retention	5 (20%)
Dropout rates	Not reported	11 (44%)
	0–30% 10	10 (40%)
	30–60 % 4	4 (16%)
Follow up period	Not Reported	5 (20%)
	0–6 months	4 (16%)
	6 months- 1 year	5 (20%)
	1 to 2 years	7 (28%)
	2 to 4 years	3 (12%)
	5- year or more	1 (4%)
Funding Source	None	10 (40%)
	Locally Funded by the hospital or Health Authorities	6 (24%)
	Philanthropic Foundation Grants	5 (20%)
	National Research Funding Grants	4 (16%)

Study design and quality:

The overall quality of included studies was evaluated as weak to moderate. The outcome measures were heterogeneous and generally low in reliability and validity. There were high attrition rates and variable follow-up lengths. None of the studies used randomized allocation or blinded rating. A wide range of supportive services were used ranging from peer support to financial support. Similarly, the outcome measures include a variety of outcomes including burnout, psychiatric symptoms, job retention and satisfaction, perceived resilience and support.

The 14 quantitative studies were rated using EPHPP. Out of these, 6 studies were rated as weak (two or more weak components). Eight were rated as moderate (one weak component). The 5 mixed methods studies were rated using MMAT, four of the studies scored 2 out of 5 and one of the studies scored 3 out of 5. The 6 qualitative studies were rated using JBI and the scores ranged from 3 to 7 out of 10.

Included interventions:

A wide range of interventions were used: Education interventions, including knowledge and skills development, time off to attend professional activities, personal skills training, peer support groups, stress management, cognitive behavioral therapy groups, meditation, financial incentives, supportive supervision, second victim support programs. The most common intervention was peer support.

Reliability and validity of outcome measures:

Fifteen out of the twenty-five studies had low reliability of outcome measures. Only ten studies used validated outcome measures. Five studies only used retention as an outcome. One study did not analyze the data due to small sample size (Olcon et al., 2022) Most of the studies used self-reported measures. One study only used “sticky notes” from the participants as the outcome (Adam et al., 2021)

Heterogeneity of outcome measures:

Ten studies used a validated outcome measure of well-being and health (Maslach Burnout Inventory, Patient Health Questionnaire-9, General Health Questionnaire-12, Global Assessment of Functioning Scale, Professional Quality of Life Scale version 5, Quality of Work Life Scale, Depression Anxiety Stress Scales, WHO Well-being Index, Brief Symptom Rating Scale-5, Rural Doctor Distress Scale, Brief Resilience Scale, Motivational Outcome Scale, Individual workload perception scale). The other fifteen studies used subjective self-report measures, two of which used the same measures. Only one study used structed questionnaire for job involvement (3 items from Kanungo’s Job Involvement Questionnaire).

Study design:

Six out of the twenty-five studies used a control group. Most (80%) of the studies reported pre and post-data for the participants and included some follow up. Follow-up periods varied from 1 month to 5 years. Follow-up rates also varied as workforce members changed over the follow-up period.

Attrition rates:

Eleven out of the twenty-five (44%) studies did not report enough information to calculate attrition rates. 10 studies (40%) reported attrition rates less than 30 % and 4 studies (16%) reported attrition rates between 30 and 60%.

Participants and Demographics:

The number of participants ranged from 9 to 1689, with the median of 33. Seven studies (28%) offered the intervention to physicians; seven studies (28%) included only nurses and seven studies (28%) included nurses, doctors and other healthcare workers. Of these 2 studies offered interventions to HCW referred through employee assistance programs and one study focused on HCW referred by an occupational health clinic, two studies specifically focused on second victim support programs^. Four studies (16%) focused exclusively on Community Health Workers. None of the studies described the interventions in sufficient detail to allow replication.

Demographic data was not consistently reported. 10 studies included none or very little demographic information about the participants. The remaining studies included some demographic data such as age, sex distribution and professional roles of the participants. Studies involving physician participants reported information about their practice specialty and years of experience, while the studies involving nursing participants reported on their practice setting (e.g. inpatient or outpatient) and managerial roles, however this information was not consistently reported across studies. Only two studies reported racial breakdown of participants. One of these was a narrative inquiry into the practices of healthcare workers’ wellness program in Australia and included 21 % non-white participants (Olcoń et al., 2022). The other study reported on a self-management training intervention in South Africa and all of the participants identified as being “colored” (Majee et al., 2019).

Study Settings:

A total of 15 out of 25 (60%) studies were performed in rural hospitals, 4 of which were described as teaching hospitals and one as a tertiary hospital. The rest of the studies (40%) were done in a variety of rural community settings.

Effectiveness of interventions:

All interventions were deemed by their authors to be at least partly effective. Seven studies reported statistically significant improvement in perception of workload (Individual Workload Perception Scale), attitude towards work (Kanungo’s Job Involvement Questionnaire), distress (Rural Doctor Distress Scale), Depressive symptoms (Patient Health Questionaire-9, coping skills (Self-rated Likert scale), psychological discomfort (Brief Symptom Rating Scale-5) and resilience (Brief Resilience Scale). Other studies reported a percentage (pre and post) intervention that indicated some improvement. Qualitative studies reported participants’ positive perception about the interventions. The methodology, findings and quality appraisal of the studies is summarized in Table 4. Due to the heterogeneity of types of study and measures used, comparisons were not made between the studies.

Funding Source: 10 studies (40%) did not report any funding support. 6 studies (24%) were funded locally by the hospital or health authorities. Philanthropic foundation grants funded 5 studies (20%). Only 4 studies (16%) were funded by the national research funding grants.

Discussion:

To our knowledge this is the first systematic review focusing specifically on well-being interventions for the rural healthcare workforce. We attempted to capture the complexity of successfully implementing support interventions for healthcare workers in the rural context. Accordingly, we did not exclude any studies based on quality and examined the full range of available evidence related to support programs for the rural healthcare workforce. Our results suggest that any intervention strategy for rural healthcare workforce should consider several factors related to the proposed interventions. These include their impact, cost-effectiveness, acceptability, and context. These findings suggest that Cochrane-style reviews alone may not be sufficient to synthesize the evidence that includes both qualitative and quantitative research.

Most of the existing literature on rural healthcare workforce focuses on recruitment and retention. A previous systematic review by Russell and colleagues of 34 studies did not find a strong association of retention with educational, regulatory, or financial interventions. They found that preferential selection of students who grew up in a rural area and undertaking substantial lengths of rural training had some association with increased rural retention (Russell, et al., 2021) However, this study did not examine other well-being related outcomes and only included two studies employing emotional support interventions. Previous studies have linked burnout and stress among rural HCW’s to increased staff turnover (Snadden & Kunzli, 2017) (Irving et al., 2017) Increased staff turnover, in turn, has been associated with higher costs, poor quality of care and increased patient mortality in rural areas (Russell et al., 2021) (Zhao et al., 2019)

HCW wellbeing is a multi-dimensional concept, encompassing individual, social and organizational dimensions. Our results suggest that there are not enough support services for rural HCW as compared to the urban settings. For example, a recent scoping review of interventions for HCW well-being identified 12 systematic reviews comprising 162 studies, nearly all carried out in urban areas. (Fadel et al., 2022) Similarly, we found that second-victim support is almost non-existent in rural areas. Second-victim refers to the healthcare providers harmed by an adverse event during the course of their work (Wu, 2000). We found only two second-victim support programs, neither of which was fully implemented. By contrast, a recent systematic review by Busch et al. found 12 second victim support resources, all of which were located in urban areas (Busch et al., 2021). However, these findings should be interpreted with the care, as it is possible that additional support program exist for rural healthcare workers that we have not been able to identify. Despite the relative lack of support interventions for rural HCW, our review supports the feasibility of these interventions in rural settings. In the included studies, interventions appear to be well-received by HCW. Some studies suggest that even a relatively brief workplace intervention can contribute to reductions in staff burnout and may represent a feasible approach to improving resilience and well-being. For example, Slayter et al reported significant reductions in burnout and depressed mood upon completion of one day course on mindful self-care and resiliency and these improvements persisted at 6 months follow up in the control group (Slatyer et al., 2018). We also found that support interventions can be effectively delivered remotely to rural health care employees through virtual communication, thus reducing implementation barriers.

A common theme among successful programs was shared decision-making involving different stakeholders, particularly the organizational leadership. Developing a proactive culture with participation of organizational administration and external stakeholders appears to be a promising strategy. Strategies to improve clinical knowledge base and skills training could also be effective ways to combat demoralization of healthcare workers. Supportive supervision training and improving social relationships with supervisors appeared to contribute to improved job satisfaction and motivation. This is consistent with previous evidence that suggests that managerial training which focuses on enhancing manager confidence and teaching them new skills in having mental health focused conversations generates improved rates of supportive managerial behaviors in a range of organizational settings (Gayed et al., 2018)

This systematic review suggests that highlighting positive aspects of rural practice, fostering self-management skills, and facilitating HCW individual coping can improve well-being and retention in rural settings. This is supported by previous research showing that a healthcare worker’s decision to locate to, stay or leave a rural community is influenced by a host of factors, including access to practice supports, opportunities for professional growth, organizational commitment to supporting the practitioner and understanding the context of rural practice (Cosgrave et al., 2019) Indeed, successful rural clinicians feel deeply connected to their patients and their communities. Rural physicians cite strong doctor-patient relationships as a primary motivator to practice in a rural setting, in addition to lower cost of living and slower pace of life (Arredondo, 2023)

In line with previous research, this review suggests that direct financial or regulatory interventions can be important but insufficient for retaining rural workforce. Ultimately it is important to create meaningful organizational change to retain rural healthcare workers. Second victim support programs are important in this regard. Cultural change in which an error is looked upon as an opportunity for improvement rather than blame or shame is necessary for retaining rural healthcare workforce. This will require leadership involvement and reevaluation of how power is perceived and used in rural organizations.

This study also identified several barriers to implementing HCW support programs in rural settings including cost, technical support, induction and training. One study that included both urban and rural HCW suggested that the rural HCW may be less likely to participate in supportive interventions than their urban counterparts. We noted a relative lack of external funding in the included studies. In our review, two-thirds of the studies were either unfunded or funded internally by the local hospital. Only one study was funded by a national healthcare research agency (Australian National Health and Medical Research Council). None of the funding came from US National Institutes of Health, Canadian Institute of Heath Research, or other federal research agencies. However, this does not imply that such funding may not exist. Another concerning finding is that most of the support interventions were in rural hospitals and directed at doctors and nurses rather than the rural community-based settings. More research is needed to investigate support interventions for rural health workers particularly those practicing in community settings.

Unfortunately, the included studies did not report detailed demographic data on study participants such as race, ethnicity or sexual orientation. Only two studies provided the racial breakdown of study participants. Both of these studies were qualitative enquiries, one of these studies did not evaluate outcomes formally and the other study did not find any difference based on race. One study in Australia specifically targeted international medical graduates and indicated that supportive educational programs may enhance self-reported professional wellbeing. Because of the relative social isolation in the rural communities, minority HCW may face additional challenges. Understanding these needs is essential for building and retaining a diverse and inclusive rural healthcare workforce and to improve health equity.

Strengths and Limitations

Our review methodology had several strengths. First, we conducted a comprehensive search to identify all eligible studies involving support interventions for rural healthcare workers. Second, we performed quality appraisals with three different tools according to the types of studies and combined the findings from quantitative, qualitative and mixed methods studies through convergent synthesis design.

Our study had several limitations. The first was methodological limitations of the included studies. None of the studies used random allocation, most of the studies lacked a control group and generally had high attrition rates. Inconsistent reporting of participant demographics and outcome measures was also a significant limitation. Due to the heterogeneity of outcome measures, it was not possible to compare studies or aggregate quantitative findings. Second, the reporting quality of included studies limited the ability to conduct risk of bias assessments. Finally, we combined the findings from different countries, including the US, Australia, and Canada. These are very different healthcare systems and our findings are dependent upon how these systems interact with rural healthcare services in those countries. Due to the significant limitations of the included studies, results should be interpreted with caution.

Conclusion:

This systematic review suggests that there is a lack of supportive interventions for rural healthcare workers, but they appear to be feasible and well-accepted. The limited evidence, not consisting of any randomized trials, suggests that these interventions may effectively improve a variety of outcomes for healthcare workers and for rural healthcare services and patients. However, there are significant gaps in our understanding. Future research should address methodological limitations in existing studies, lack of experimental design and heterogeneity of outcome measures. Studies should also target a broader range of HCW including allied health professionals and community settings. Policy-makers should prioritize research funding in this area to identify potential interventions to improve well-being, recruitment and retention of rural healthcare workforce.

Availability of data and materials:

The datasets used and/or analyzed during the current study available from the corresponding author on reasonable request.