Barriers, frameworks, and mitigating strategies influencing the dissemination and implementation of health promotion interventions in indigenous communities: a scoping review

The formalization of research in D&I is growing and numerous models exist to guide this process [52, 53]. Research-to-practice models are most frequently applied and are intended for use by diverse stakeholders (e.g., researchers, community-based practitioners, and funders) to systematically guide and critically assess prevention efforts [56, 57]. They also help to inform on specific D&I steps, such as community needs assessment, to identify important barriers and facilitators, and inclusion of community members’ expert knowledge in implementation planning, and assessment of community capacity [56]. The “Dissemination and Implementation Models in Health Research and Practice Webtool,” a collaboratively developed decision support tool, provides an updated database of D&I frameworks to assist researchers and practitioners to generate research questions, select, adapt, and combine D&I models for particular study contexts, and implement and evaluate D&I models [53]. Despite the utility of D&I models and availability of decision tools, their application to guide program implementation has been the exception rather than the rule [8, 9, 58, 59].

These are practical tasks (often associated with D&I models) recommended to aid the successful D&I of research findings into clinical and community practice [60]. Taxonomies of strategies to successfully facilitate the adoption, use, and maintenance of EBIs include the ERIC (Expert Recommendations for Implementing Change) and SISTER (School Implementation Strategies, Translating ERIC Resources) taxonomies [54, 55]. The ERIC taxonomy comprises 73 strategies devoted to implementation of EBIs in healthcare settings [54, 60]. The SISTER strategies are an adaptation from those in ERIC but focused on, and more compatible with, school and community-based contexts [61]. The SISTER taxonomy comprises nine domains: (1) use evaluative and iterative strategies; (2) provide interactive assistance; (3) adapt and tailor to context; (4) develop stakeholder interrelationships; (5) train and educate stakeholders; (6) support educators; (7) engage consumers; (8) use financial strategies; and (9) change infrastructure [59, 60]. Within the nine domains are 75 strategies focused on training, local technical assistance, adoption, high fidelity implementation of EBIs, and program replication in school-based settings [62, 63]. Additional previously identified strategies, seminal to use in Indigenous communities, include integration of EBIs within the cultural context [64, 65], involvement of Indigenous leaders, and ensuring sufficient resources (i.e., economic, health, and political) [9, 64, 65].

The purpose of this scoping review was to identify common barriers and effective mitigating D&I models and strategies to successfully disseminate and implement evidence-based programs in American Indian/Alaska Native (AI/AN), Native Hawaiian/Pacific Islander (NH/PI), and Canadian Indigenous communities. This review builds on a published multi-case study by Jernigan et al. (2020) to develop culturally tailored D&I strategies to enhance the ability of researchers to scale up effective interventions among Indigenous communities [8]. This scoping review may further contribute to informing and guiding future D&I initiatives aimed at reducing health disparities in this population.

Three guiding research questions for the scoping review were: (1) What are the main barriers encountered in the D&I of programs and EBIs in Indigenous communities?; (2) Which research-to-practice models have been used to promote the D&I of health promotion EBIs in Indigenous communities?; (3) What implementation strategies have been used in Indigenous communities for program and EBI adoption, implementation and/or maintenance?

Keywords and mesh terms were developed in corroboration with a research librarian experienced with scoping review protocols. Search terms focused on AI/AN and NH/PI communities, Native communities, Indigenous tribes, tribal groups, dissemination models, dissemination frameworks, implementation frameworks, EBIs, and US and Canadian territories (Table 1). Educational subject headings and Boolean operators were adopted as search tools to narrow, widen, and combine literature searches. The Rayyan platform was used to condense all studies generated from our search [68]. Three electronic databases (PubMed, EMBASE, and Medline (Ovid)), selected for their breadth and focus on psychosocial and behavioral science, were searched to identify peer-reviewed literature from primary data sources, secondary data sources, and case reports. The review of the literature databases was completed over a period of 2 months, ending in June 2020. Articles were screened for eligibility by reviewer pairs (CM and BH; RS and MP) over a period of 3 months, ending in September 2020.

The lead author (LS) extracted and summarized the data from relevant studies. Reviewer pairs (CM and BH; RS and MP) reviewed the data extraction and summary tables for accuracy. Conflicting opinions were resolved by consensus discussion. Summary tables included an evidence table describing each study’s parameters including guiding D&I models, identified barriers, and mitigating strategies. D&I models were identified using the ‘Dissemination and Implementation Models in Health Research and Practice Webtool’ previously described [53]. Barriers, contextual factors that hinder implementation at each level of the socio-ecological model (SEM) [11], were classified by the 5 levels of the (SEM) and by barrier categories based on major themes within the broader SEM framework. The SEM framework acts as a comprehensive external reference to the D&I models and strategies; therefore, it aids in the assessment of such models and strategies when applied to multiple and interacting determinants of health behaviors [11].

D&I strategies were categorized and coded according to the SISTER framework (previously described). The SISTER taxonomy was used as the referent due to its utility for school and community-based contexts [61]. Initial categorization and coding by the lead author (LS) was compared to independent categorization with reviewer pairs for inter-rater reliability in a subsample of 38% (n = 8) studies. Inter-rater reliability was conducted in two rounds with discrepancies resolved by consensus discussion. Resulting inter-rater reliability was 90% for strategy-level matching and 70% for domain-level matching (Supplemental Tables 1 & 2).

Study characteristics were tabulated for primary author, country, study type, sample size, target population, study topic area, and D&I model (Table 2). Identified barriers were tabulated by SEM level and classified to one of nine barrier categories (Personnel Challenges & High Turnover; Distrust; Funding; Lack of Integration with Cultural Values; Social Determinants of Health in Communities (physical, mental, health, social, and financial challenges); Insufficient Evaluation Skills; Technology Barriers; Limited Retention and High Attrition; Climate Conditions) (Table 3). The specific strategies were rank ordered within the SISTER domains, as well as based on importance and feasibility (Table 4).

Priority populations who were actively involved (or targeted) in implementation activities were adults (81%, n = 17) and/or children/youth (43%, n = 9) (Table 2). Adult participants included tribal members and elders (AI/AN, n = 4; NH, n = 1; First Nation, n = 1), community health workers (n = 1), women (AI/AN, n = 1; Choctaw, n = 1), mothers and caregivers (AI/AN, n = 1; First Nation, n = 1, Choctaw, n = 1); and those with chronic disease and health challenges (AI/AN with Alzheimer’s, n = 1; adults enrolled in fetal alcohol spectrum disorder services, n = 1; Indigenous victims of car accidents, n = 1; NH with cardiovascular disease and hypertension, n = 2). Key stakeholders who were crucial to planning program implementation included decision makers in healthcare, school, community, organizations, academics, and government (Table 2).

The evidence-based programs targeted a variety of health domains, including chronic disease and injury, substance misuse, wellness and illness prevention, and historical trauma (Table 2). Chronic disease and injury topics included hypertension and cardiovascular disease (n = 3), obesity (n = 1), asthma (n = 1), diabetes (n = 1), hearing loss (n = 1), Alzheimer’s (n = 1), palliative care (n = 1), and motor vehicle crashes (n = 1). Substance misuse included misuse of alcohol and other drugs (n = 5) and tobacco use (n = 1). Wellness and illness prevention topics included maternal and child health (n = 1), sexual health (n = 4), nutrition (n = 4), physical activity (n = 1), improved access to healthcare services (n = 2), breast and cervical cancer screening (n = 1), overall children’s well-being (n = 1), and reduction of environmental contaminants exposures (n = 1). One study focused on a historical approach to health through walking the Trail of Tears and 2 studies reported programs addressing multiple health topics [8, 10, 31].

Diverse tribal communities were represented in this review, including AI/AN (n = 13), Inuit (n = 2), and First Nation/Indigenous (n = 7), and Native Hawaiian (n = 2) communities (Table 2). AI/AN communities included tribes in Oklahoma, California, Alaska, Arizona, and the Pacific Northwest (Oregon, Idaho, and Washington). Inuit communities included tribes in Greenland and Northern Canada. First Nation/Indigenous and Native Hawaiian communities had representation from multiple regions in Canada and Hawaii respectively. Settings comprised Native nations, reservations and reserves, tribal agencies and associations, health agencies, academic affiliates, and schools (Table 2).

Eighty-nine barriers to implementation were reported in 17 studies (81%), representing the five levels of the socio-ecological model (SEM): Individual (n = 22), interpersonal (n = 6), organizational (n = 49), community (n = 41), and society/policy (n = 26) (Table 3). Barriers were also sorted into nine categories (Table 3) based on major themes that were established through similarity of barriers highlighted across studies at the different levels of SEM. Some barriers fit into the SEM levels, and thus generated more than one theme. For instance, Barlow et al. (2018) highlighted “socioeconomic, geographic, and structural challenges” as a barrier, affecting the individual, community, and society/policy levels of the SEM. The barrier category themes emerging from this barrier and its subsequent SEM classification included “funding,” “social determinants of health in communities,” and “climate conditions.” Most cited barriers (n = 38) sorted into the Community/Society-Policy category of “Social determinants of health in communities.” A majority of studies also cited “Personnel challenges and high turnover” (n = 29), “Funding” (n = 18); “Lack of integration with cultural values (n = 11), and “Limited retention and high attrition” (n = 9) Other barrier categories included Technology barriers (n = 7); Distrust (n = 6); Insufficient evaluation skills (n = 3); and Climate conditions (n = 2).

Sixteen studies (76%) used a specific D&I model to promote the adoption and implementation of health promotion EBIs in Indigenous communities (Table 2). Eight different unique models were cited. Community-Based Participatory Research (CBPR) was most commonly reported (n = 11). Four studies used models that focused on dissemination and/or implementation (Knowledge-to-Action Framework, Diffusion of Innovation Theory, and RE-AIM), andragogy (Adult Learning Theory), or inductive and culturally responsive processes (Culturally Grounded Models of Health Promotion). Remaining models focused on the broader implementation process inclusive of dissemination. Ten studies used a D&I model for the purpose of identifying barriers and/or facilitators to the dissemination process; seven studies highlighted the main barriers and/or facilitators that were encountered during the implementation process.

All SISTER domains were represented, and all extracted D&I strategies were matched to relevant SISTER strategies However, not all SISTER strategies were represented in the included studies. One hundred and eighty-four D&I strategies (n = 184) were identified, corresponding to 60 (80%) of the SISTER strategies. A range of three through nineteen strategies were reported in any one study. The most commonly reported SISTER strategy (identified in 86% of studies) was: “Build partnerships (i.e., coalitions) to support implementation” (#21) (Table 4). Four SISTER strategies, previously recognized as being highly important for D&I success were represented in the top 10 strategies [33]. These were “Conduct ongoing training” (#39), “Monitor the progress of the implementation effort” (#9), “Provide ongoing consultation/coaching” (#44), and “Make training dynamic” (#43). These strategies occur in the domains of “Train and educate stakeholders” and “Use evaluative and iterative strategies.” Four SISTER strategies previously described as most feasible for successful D&I were also represented in the top 10. These were: “Make training dynamic” (#43), “Distribute educational materials” (#42), “Facilitation/Problem solving” (#12), and “Capture and share local knowledge” (#22) (Table 4).

The majority of the studies (76%) used a D&I model to guide the dissemination and/or implementation of an EBI. Such studies have increased in recent years with 66% of the included studies published since 2015. This reflects the recognition of D&I to address existing and emerging health disparities and is consistent with a broader increase in D&I research. The most frequently reported model was Community-Based Participatory Research (CBPR) (n = 11), which encompasses an array of principles consistent with partnering with Indigenous minorities [70]. A recent systematic review by Julian McFarlane et al. (2021) [71] highlighted the large increase in the number of CBPR-related studies targeting a broad racial and ethnic representation in research. More than 85% of these studies saw statistically positive outcomes when applying CBPR methods, particularly community partner participation in study advisory committees, data collection, the development of interventions, and participant recruitment [65].

CBPR aims to (1) recognize the Indigenous community as a unit of identity, (2) build on the community’s strengths and resources, (3) facilitate collaborative partnerships in all phases of the research, (4) integrate local knowledge and actions that benefit all partners, (5) empower community members to address social inequalities, (6) involve a cyclical and iterative process, (7) address health from both positive and ecological perspectives, and (8) disseminate findings and knowledge gained to all partners [72]. These principles represent an important foundation to guide ethical D&I studies and are complementary with common reported strategies (described below). Yet CBPR is not without limitations and may not account for the specific array of facilitation strategies and prescriptive steps associated with many D&I models [70]. The frequency of application of D&I models other than CBPR was relatively low (n = 5). Greater research on D&I models in Indigenous communities may enhance the quality of implementation planning and evaluation in those settings, building empirical evidence for the utility of such models using traditional CBPR approaches [73, 74]. Encouraging these systematic approaches can also expand our knowledge-base on the most salient D&I models and strategies for Indigenous communities [73, 74].

This study reinforced the critical need to identify and implement D&I strategies at all levels of the socio-ecological model to address common barriers that impede implementation efforts. The social milieu in which programs are deployed in Indigenous communities can be complex and challenging. Principal among these challenges are consideration of social determinants of health, perceptions of community trust, community skill sets, and financial challenges. Social determinants of health are important considerations when attempting to reach underserved populations as they address issues related to the complex mental, health, social, physical, and socioeconomic issues of communities. They can represent major barriers to program implementation. Cited factors that can compromise program implementation in Indigenous communities include poverty, homelessness or residential instability, geographic remoteness with accompanying challenges of access to healthcare service, and greater transportation expenses. Across the literature, intentional information gathering and community involvement were critical to program success. These included “assessing for readiness and identifying barriers and facilitators,” “involving governing organizations,” “informing local opinion leaders,” and “involving students, family members, and other staff” [13‐18, 23, 24, 26‐31]. More broadly, the strategy of “changing or altering the environment” was employed where feasible, again in consort with community stakeholders.

Complicating the challenge of social determinants is the perception of trust between community members and healthcare providers, or between program participants and the entity delivering the program (i.e. organization, academic institution, governmental agency). These relate to the barrier categories of “distrust” and “lack of integration with cultural values.” Building partnerships to support implementation was the most commonly cited SISTER strategy across the included studies (86%). However, despite the importance of building partnerships in the community and sharing its local knowledge, additional strategies are indicated. Most studies (55%) reported organizational barriers related to involving the views and experiences of elders, community health workers, families, and youth as part of the implementation process [13, 15, 20‐22, 28, 30, 32]. Hearing the community voice and attending to community needs can further engender trust. The expertise of Indigenous community members, elders, and health planners, many of whom have unique skills, particularly in the fields of cultural adaptation, tailoring interventions, and appropriate implementation is highly valued and can help to alleviate community concerns [75] as well as smooth logistics involved with navigating the complex tribal internal review and research review boards necessary for collaboration with external academic and research partners [8].

The studies mentioned other D&I strategies that can promote cohesion around program implementation at the organizational level. These included recruiting and retaining families through trust-building; ensuring convenience of program offerings, forming local advisory boards and task forces, creating cultural activities, and using mass media tools (newspaper, written materials, and radio programs) to promote programs. Organizational administration included attention to data management; capacity-building efforts, prioritization of strategies, and collaboration with academic researchers and regional stakeholders [8, 10, 12‐32]. Frequently cited was the need to elicit community support through engagement of the community and Native stakeholders in the planning and implementation process [13, 15, 20‐22, 27, 28, 30, 32]. This is vital to aid in cultural learning, integration of cultural values, and inclusion of indigenous role models to optimize cultural compatibility and the potential for sustained implementation. Native stakeholders should be engaged in the planning phase to ensure that their needs and desires are fulfilled [13, 15, 20‐22, 28, 30, 32].

Staff training, personalized technical assistance, staff commitment to engage youth, and continuous evaluation of staff performance [8, 10, 12, 17, 23‐25, 29, 30] are necessary for sustained implementation of programs within Indigenous communities. These strategies can mitigate the “Personnel Challenges and High Turnover” that was cited in 65% of the studies [13‐16, 18, 20‐23, 25, 32]. High turnover rates can undermine personnel skills training due to the continuous loss of acquired talent and the need to accustom new personnel to the community and program material [13‐16, 18, 20‐23, 25, 32]. Insufficient skills needed to deliver the program material were cited as a common barrier. SISTER strategies included under the two domains—“training and educating stakeholders” and “developing stakeholder interrelationships”—could help address those common barriers.

Funding is a continuous challenge affecting sustained implementation. Funding issues were frequently reported by Native stakeholders during interviews and focus group sessions and emerged as a main theme in qualitative studies [13‐15]. This included a lack of sustained funding at the organizational level to increase research outputs [12‐18, 20, 22, 26]. This in turn led to a limited availability of resources and thus the inability to maintain programs outcomes for longer periods of time. Specific financial barriers included high cost of salaries, housing, transportation, and other mission fees needed to hire social workers, program adopters and implementers, and healthcare workers [13, 14, 20‐23, 25, 32]. Accessing new funding sources was a leading D&I strategy employed in Native communities. Continuous delivery of program resources and material is predicated on sustained financial support without which D&I efforts are hobbled [55].

Studies describing intervention implementation at the policy level cited the importance of creating and implementing new public health policies to overcome societal and economic barriers. These crosscut other socioecological levels and included the high costs of imported goods and healthy foods, inadequate funding allocations to healthcare systems, limited assistance for uninsured clients, limited resources for chronic diseases, improper management of historical oppression and trauma, infrastructure shortcomings, and high levels of poverty [13, 15, 16, 18, 25]. All nine domains encompassing multiple SISTER strategies were mentioned in the studies. Studies on the effectiveness of D&I strategies in this domain are limited [54, 61]. Future work could focus on the multilevel policies that shape social determinants of health and their impact on D&I outcomes in Indigenous settings. Holistic approaches with culturally tailored strategies are essential to overcome potential barriers.

These studies correspond highly to reported SISTER strategies previously categorized as important and feasible in non-indigenous contexts [61]. Four of five strategies rated as most important were among the top ten reported in this review. These strategies included (1) “Monitor the progress of the implementation effort” (#9); (2) “Conduct ongoing training” (#39); (3) “Make training dynamic” (#43); and (4) “Provide ongoing consultation/coaching” (#44). The 5^th strategy, “Improve implementers’ buy-in” (#51), was not represented. Four of five strategies rated as most feasible were among the top ten reported in this review. These included (1) “Capture and share local knowledge” (#22), (2) Distribute educational materials” (#42); (3) “Make training dynamic” (#43); and (4) “Facilitation/Problem solving” (#12). The 5^th strategy, “Remind school personnel” (#53), was not represented in any of the studies. Financial strategies categorized under the domain “Use financial strategies” received a low feasibility rating in Lyon et al. (2019) and were only reported in a few of our studies [61]. This may reflect the lack of funding that was identified as a barrier in 50% of the studies [61].

Findings need to be interpreted in the context of study limitations. First, despite a comprehensive search of the most relevant psychosocial databases, this review did not include tracing of reference lists in included studies, hand-searches of journals, or grey literature. Broader reviews are recommended that account for these sources. Second, the D&I field is growing rapidly, so it is possible that some relevant studies were not found due to inadvertent omission of search terms. The mesh terms included as many technical D&I keywords as possible and the collaboration of a research librarian who imposed rigor in the protocols likely mitigated this concern. Future reviews are recommended to include emerging terms from this rapidly evolving field. Third, the scope of the current review was limited. Formal assessment of the quality of the included studies was beyond scope and the inter-rater reliability, though acceptable with domain and strategy correspondence of 70% and 90% respectively, was based on assessment of only eight (38%) of the included studies. Fourth, matching the identified D&I strategies to the SISTER strategies was challenging due to the diversity of terms used to describe any given strategy. Consistency of terminology represents a challenge for any emerging field. Standardizing the nomenclature will be important to enable clear research and practice guidelines for EBI implementation. Fifth, the use of SEM to categorize barriers and contextual factors limits comparison to other D&I frameworks such as CFIR (Consolidated Framework for Implementation Research) or EPIS (Exploration, Preparation, Implementation, Sustainment). However, SEM categorization will inform the selection of multilevel implementation strategies to facilitate EBI uptake in Indigenous communities [52]. It also provides an objective assessment agnostic of any particular D&I framework [52]. Finally, the SISTER strategies were originally developed based on studies in non-Indigenous settings. Although the taxonomy is comprehensive and provides a useful comparison for non-indigenous settings, it may also miss cultural influences or D&I processes that are unique to Indigenous communities. The similarity with findings from Lyon et al. (2019) indicates some validity across cultural settings [61]. Future studies are recommended to provide guidance on which strategies to use to promote behavior and health changes in Indigenous settings. The use of existing accepted taxonomies in this study may provide guidance for future work.