Literature searches and eligibility assessment occurred in June and July of 2021 using Preferred Reporting Items for Systematic reviews and Meta-Analysis guidelines [PRISMA-Equity; search activities were completed by author MR (27)]. The search strategy was conducted within PsycINFO and PubMed/Medline databases. The second step in the search protocol involved all combinations of our population and content keywords. Population keywords included words used to describe and identify Indigenous populations. There were more than 30 content keywords including reference to dissemination and implementation frameworks and approaches. Combinations of each population term and all content keywords were administered via advanced search options with selections made for peer-reviewed articles written in English.

Search keywords were either combined with “AND” or in one plain search phrase (e.g., American Indian and Alaska Native; Māori “AND” dissemination and implementation science; implementation strategies; Explore, Preparation, Implementation and Sustainment; health equity implementation framework; Indigenous Implementation Framework). For a complete list, please refer to the Supplementary materials. The third step in the research strategy was to take the references identified by all combinations of search terms from each database extraction and save them as Research Information Systems (RIS) text files that were then uploaded into COVIDENCE, a systematic review management software (28). Duplicates were identified and removed within the software system.

KH (descendent of the Eastern Shoshone Tribe/White, mixed European ancestry), MR (citizen of the Haliwa-Saponi Tribe), and KB (descendant of Filipino ancestry and immigrants) reviewed research articles for eligibility in COVIDENCE based on the title and the abstract. Study eligibility criteria was determined by the research team using knowledge of existing literature (e.g., anticipated state of the science), consideration of limitations of the research team (e.g., language capacity), and specific interest in SUD treatment implementation (29). Inclusion criteria were: (a) peer-reviewed, (b) written in English, (c) conducted with an Indigenous, adult (18 and older) population, (d) comprised of aims and outcomes related to substance misuse; (e) delivering an evidence-based intervention or practice based upon Western scientific criteria (e.g., efficacy or effectiveness randomized controlled trials), and (f) described one or more implementation processes or strategies. Exclusion criteria were: (a) studies completed entirely with non-Indigenous populations, (b) interventions intended for youth (younger than 18 years-old), (c) research that did not include outcomes related to substance use, and (d) research that did not discuss an evidence-based intervention (e.g., efficacy or effectiveness randomized controlled trials) for substance use. There was no exclusion based on publication date.

For articles which met initial eligibility criteria based on title and abstract, a full text review was conducted by KH and MR. Data extraction was completed by KH, MN (Turtle Mountain Band of Chippewa Indians), MR, KB, and MB (Spokane Tribe of Indians). Articles were initially reviewed independently and then underwent a second review by an alternate author to ensure validity and reliability of the extraction. None of the reviewers were individuals who had authored one of the articles. Consensus was determined and resolved by KH, MR, KB and MB. Final data review was carried out by KH and MR.

Once the final sample of articles was confirmed, the adapted RE-AIM framework was utilized for coding by two independent coders. While none of the studies were explicit D&I outcome studies, we coded the articles based upon what applied to the RE-AIM framework. This included articles that were clinical trials, methods papers, qualitative research, and case studies. Operational definitions based on the RE-AIM framework were identified and developed by the first author along with the review protocol (see Supplementary materials adapted from D’Lima et al. (30) and Shelton et al. (21)). The health equity and long-term sustainability extension codes were embedded within each existing domain. Codes were used to complete the data extraction within COVIDENCE, with final consensus completed in Excel.

For Reach, two questions were coded around the intended audience and who participated, as well as ways to better reach and engage the intended audience. Effectiveness was coded with three items related to whether the intervention was effective, if there had been any unintended consequences, and if intervention effectiveness was assessed over time. Adoption was assessed by three questions at the staff and setting level. This included application of the intervention and by whom, which staff were invited to participate versus excluded, and how staff were supported in delivering the intervention or could have been better supported. Implementation was assessed by six items ranging from efforts to culturally adapt the intervention, fidelity and delivery of the intervention, and costs. Maintenance was reviewed through the lens of sustainability of the intervention beyond 1 year and if multi-level contextual determinants were discussed that might impact sustainment.

The initial search produced 44,301 articles. After the removal of 19,379 duplicates, an additional 24,689 were deemed ineligible through title and abstract screening and 234 articles underwent full-text review. After the full text-review, 20 articles were identified for inclusion (see Figure 1). Just over one-third were methods or protocol papers and about half of the studies included a randomization component. Of the studies identified, 65% included AI/AN adults, 20% Aboriginal Australian adults, 10% First Nations adults, and 5% Native Hawaiian adults. The evidence-based SUD programs being implemented (sometimes in combination) included: nicotine replacement therapy (n = 8), motivational interviewing (n = 5), community reinforcement approach (n = 4), medications for opioid use disorder (n = 3), contingency management (n = 3), community reinforcement approach and family training (n = 2), and recovery housing (n = 1). Primary substance use outcomes included smoking cessation (n = 7), alcohol use (n = 6), more than one substance or not specified (n = 5), and opioid use (n = 3). All the interventions focused on the individual, with a little over half also having a family, community, or organizational component (n = 12). Only two of the studies intentionally included implementation strategies and existing dissemination and implementation frameworks, but this was integrated retrospectively (Table 1).

Reach, Adoption, and Implementation were discussed by 60% of the studies. This was followed by Effectiveness (50% n = 10), with five studies discussing Maintenance/sustainability (25%, Table 2). Frequency and distribution of each of the domains coded in the literature are also presented (Figure 2). Below we provide descriptions of how the RE-AIM domain was characterized and include examples from the literature that highlight each domain. Definitions of domains and abbreviated summaries of examples can be found in Table 3.

Sixty percent of the studies reported Reach [n = 12; (33, 35–38, 40, 42, 45–48, 50)] with information mainly focused on inclusion and exclusion criteria for participants along with recruitment strategies (33, 35, 36, 40, 42, 45, 46, 48, 50). For example, Campbell and colleagues (35) noted that they recruited 58.8% (N = 40) out of the 68 clients that were eligible to participate in their study. In another study, the involvement of treatment providers from the community was also critical to Reach, with one study describing their recruitment efforts as including clinician referrals, radio newspaper, and digital (i.e., Facebook) ads, outreach at community events, and word-of-mouth (33). Reach was expanded in a pharmacist-led initiative through the Indian Health Service to increase access to medications for opioid use disorder [MOUD; (33)]. This was achieved by increasing the number of individuals initiated onto medications, poisoning awareness and procurement of naloxone kits that were distributed nationally. Other strategies around Reach included naloxone/drug take-back initiatives and engagement through culturally responsive educational tools (37).

Within the Reach domain, though not always referencing CBPR specifically, articles highlighted the importance of community buy-in and engaging collaborators on the project that were a part of the community and had knowledge of cultural protocols [e.g., how to appropriately greet an Elder; (50)]. Another common implementation strategy across studies was convening an advisory board (e.g., Community Advisory Board) or collaborative board to provide oversight on research activities, assist with recruitment and facilitate connections across organizations in the community. In one study, for example, the Community Advisory Board assisted in developing the culturally appropriate intervention title and how to brand the project in each community, identified and adapted measures, and guided focus groups to increase cultural acceptability of the intervention (40).

Trust was also emphasized in a few studies as another reason Reach was successful (36, 37). A trusting relationship between the participants and the research team was key. In these studies, it was argued that participants reported that they participated because they knew that the research team would manage their data with care, which speaks to data sovereignty and privacy. A unique aspect of Reach and trust was illustrated by a study completed in Australia (45). The authors described how “jealousy” played a significant role in recruitment. As reported by the authors, the spouses of participants were concerned that participation in research would provide opportunities for romantic infidelity. Research staff responded by developing strategies to make the intervention more welcoming and inclusive by providing opportunities to non-participating spouses. While specific to this community, this study demonstrates the importance of an engaged and flexible staff that can identify barriers and then develop solutions.

The need for flexibility in participant engagement strategies, and the ability to pivot based on community-generated solutions for health equity in substance use accessibility and capacity was, in fact, described across studies. For example, Orr and colleagues (47) initially intended to recruit AI/AN college students for their tobacco cessation program. However, recruitment efforts led to only 9 college students enrolling. The researchers then turned their recruitment strategy to Quitlines in states that had higher AI/AN rural populations (e.g., Alaska, New Mexico) and were able to complete their recruitment efforts. Another study identified recruitment challenges due to staff not having the time to recruit at the site, a short recruitment window, and eligibility criteria that were too stringent (38). One study team discussed needing to make changes to recruitment strategies when participants were screened eligible but declined to participate. Of note, the reasons for declining varied but were both practical and directly related to the research fit in the community. For example, participants reported not being comfortable with randomization, lack of interest in research, a preference for another type of treatment, and study location that was too far away (50).

Out of the 20 studies reviewed, ten (50%) of them reported outcomes related to intervention effectiveness (1, 34, 36, 38, 40, 41, 45, 48–50). There were several studies where the primary outcome was tobacco cessation [n = 7; (32, 36, 38, 45, 47–49)]. In a tobacco cessation program for pregnant Aboriginal women, self-reported 12-week 7-day point-prevalence abstinence was 13.6% (38). In another smoking cessation study integrated across two geographically and culturally distinct Aboriginal primary care settings, no statistically significant difference was observed in quit rates, but the study was also statistically underpowered (45). Although the quit rate varied, up to 43% of patients were tobacco free for 3 months across a healthcare system with five locations serving Native Hawaiian people (49).

Interventions addressing alcohol and other substance use also resulted in favorable outcomes. In a study assessing the effectiveness of Motivational Interviewing and Community Reinforcement Approach among a Tribal community in the Southwest United States, at 8 months, percent days abstinent had increased for both alcohol and cannabis, as well as other substances [excluding tobacco; (47)]. In the primary outcome clinical trial of Motivational Interviewing and Community Reinforcement Approach, participants in both the treatment, and treatment as usual (TAU), improved in percent days abstinent and substance use severity [MICRA PDA =72.63%, TAU = 73.62%; (1)]. At the three-month follow up in another study, Aboriginal participants in the Community Reinforcement Approach intervention had significantly reduced their alcohol use. There were also reductions in frequency of days of alcohol consumed and number of drinks consumed per drinking day based on pre- and post- self-report (34).

Sixty percent of the studies described processes and outcomes related to Adoption [n = 12; (1, 32, 36–38, 40–42, 44, 47–49)]. At the staff level, authors highlighted strategies, for and barriers to, Adoption. Including enhancements to Adoption focused on the staff members themselves. While some studies had specific inclusion and exclusion criteria for staff (32, 38, 48), Indigenous staff and level of staff buy-in were highlighted as the key to increasing Adoption (37, 41, 44). The duration of the intervention also impacted Adoption. While shorter interventions are generally considered easier to Adopt, some of the studies highlighted that shorter project duration among Indigenous communities actually inhibited long-term Adoption and more time was needed to build relationships (36, 38, 40). Recommendations to increase Adoption included hiring qualified study staff who were trusted community members cross-trained to facilitate coverage (38). Also, the quality of the Tribal-organization and university partnership and the need for relationship-building at all levels was of noted importance, for managing potential distrust of research by the community, as well as to facilitate approval processes from multiple entities (1, 33, 41, 46).

At the organizational level, Adoption was discussed in the context of recruitment success, with higher recruitment translating to better adoption (36, 41, 45). One study focused on tobacco cessation located at a healthcare system with different locations across the Hawaiian Islands. Differences in adoption by site were explained by barriers such as lower level of administrative or clinical support, staff turnover, and lower organizational readiness to adopt a new protocol or service (49). Although the authors initially assumed locations would be similar because they were within the same health system, each of the five locations evidenced variation in community needs, representation on the governing board, use of their medical record system, or attitudes toward medications for tobacco cessation (49).

Facilitators of Implementation were described at both the internal (i.e., organizational) and external levels, with 60% of studies touching upon this [n = 12; (1, 33–35, 37, 38, 40–43, 46, 47, 49)]. Organizations face an array of internal challenges (e.g., shifting priorities and reduced funding) along with external factors that impact what types of interventions may be supported or can be reimbursed. Authors discussed strategies related to the need for capacity building within and outside of the organization, having champions at multiple levels within (e.g., direct service, supervisor levels) and outside of the organization [e.g., Tribal leadership; (1, 40, 41, 49, 50)] and holding information meetings about the intervention for service providers outside of the organization (i.e., community members and leaders). Within the organization, it was recommended that there be on-going face-to-face communication between research staff and site staff and sharing of information and lessons from Implementation including monthly ongoing technical assistance/facilitation (1, 35, 38, 40, 41, 45).

Modifications to the intervention to enhance acceptability, integration, and implementation included cultural adaptations tailored to each setting. Making cultural adaptations to evidence-based substance use disorder treatment was described as a more holistic approach that sits within the cultural framework of the partnering community. This can take the form of integrating appropriate representation (e.g., art, study staff from the community) knowledge, and Indigenous worldviews into the evidence-based treatment which may also increase engagement and adoption (33, 42, 43, 50) (Table 1). Seventy-five percent of the studies (n = 15) mentioned culturally adapting the intervention, but only 15.0% (n = 3) described a specific framework guiding the adaptations made. While 90% of the studies mentioned a process of community engagement, only one-third of the studies explicitly stated the use of CBPR. Of those, there was no description or definition of how CBPR was being interpreted. None of the studies explicitly identified an Indigenous framework used as a part of the program implementation or intervention adaptation.

Examples of the commonly described cultural adaptation processes and methods included: focus groups, having members of the community lead the intervention, and culturally adapting materials (e.g., including Indigenous languages or pictures that represent the community). One study was initiated when a Tribal elected official contacted the researchers following the development of a community sober living house to assist with evaluation of feasibility, acceptability, and effectiveness (42). The study was unique in that the residents culturally adapted the model and the implementation processes in real time. The communal nature of the sober living house aligned with Tribal culture, and participants were able to include cultural activities (e.g., sweat lodge, talking circles with the inclusion of sacred objects like feathers, and drumming circles). Further modifications included the house being run by the residents who decided all activities, house rules and procedures. The house was also inclusive of non-Tribal members, so that anyone would have a home if they needed one, reflecting the Tribe’s values of providing social support and resources to those most in need (42).

Although funding is a core aspect of Implementation, it was not frequently discussed. Many of the studies did not directly specify the costs of delivering the evidence-based treatment, and, when funding was discussed, studies highlighted that it had enhanced Implementation efforts. For example, one study described how the total budget was dispersed across partnering communities during the duration of the research project and that this had been one of the major strengths of the collaboration [e.g., $1.8 million budget over 6 years for contingency management; (48)]. The authors of a multi-component intervention for smoking cessation among pregnant Aboriginal women did not list the entire cost of the intervention and delivery but noted that each site received $6,000 to support delivery of the intervention and staff contribution (38). In another study assessing recovery housing, Oxford Houses were argued to be self-sustaining and a lower cost option for post-treatment or recovery maintenance because each resident contributed to the cost of running the home. Professional staff were not employed (Oxford houses are operated by the residents) and residents’ rent typically amounted to about $100 per week (42).

Maintenance, or Sustainability, was the least discussed across studies with 20% of the articles including some discussion around this domain [n = 4; (39, 40, 44, 49)]. Multi-level factors related to sustainment suggested by a few authors included societal and policy forces (e.g., racism, historical trauma, poverty, and discrimination) but these components were not directly assessed within the studies (1, 41, 44, 45, 50). The studies that did highlight Sustainability also speculated on activities that may support or hinder the long-term delivery of the intervention. Authors also described future plans for program sustainment (39).

One qualitative study that examined a program delivering medications for opioid use disorder emphasized the need for community education and buy-in around harm reduction strategies to effectively sustain program activities. The need for education and training related to these medications was especially important to inform family members about how the medications worked to support treatment and how they were different from other types of opioids and illicit drugs (e.g., some family members were unsupportive of methadone maintenance treatment). Among those interviewed, there was a belief that the long-term impact of the medication for opioid use disorder program had been to increase cleanliness/safety (e.g., reduction in discarded needles) and reduce crime (e.g., burglary, vandalism) in the community. The program had already been on-going for 5 years by the time of publication, so institutionalization of the program was achieved, but the specifics of how this was accomplished were not discussed in detail (44).