Background

It is well established that Indigenous people experience poorer health outcomes than non-Indigenous people globally [1]. Australian Aboriginal and Torres Strait Islander people, like other Indigenous populations in Canada, New Zealand and the United States, endure ongoing health inequities such as a high burden of chronic disease and difficulty accessing culturally safe health care [2, 3]. Chronic diseases with strong environmental and behavioural etiology, such as cardiovascular disease and Type Two Diabetes Mellitus (T2DM), contribute to approximately 80% of the mortality gap between Australian Aboriginal and non-Aboriginal people between 35 and 74 years of age [4]. Aboriginal people residing in more geographically remote areas experience further disadvantage and a higher burden of chronic disease [5]. For example, the proportion of Aboriginal people with Diabetes Mellitus in Very Remote areas of Australia is approximately twice that of Aboriginal people in Major Cities [6]. The lack of affordable fresh fruit and vegetables in these areas is one contributing factor [7].

Targeted chronic disease prevention and management programs delivered in the primary health care setting are imperative to alleviating the burden of disease and improving health outcomes for Indigenous people [3, 8]. In Australia, little progress has been made in improving health outcomes, the distribution of chronic disease, and risk factors for developing disease for Aboriginal people [2, 9]. This is despite numerous funded Aboriginal chronic disease programs implemented at a national level (e.g., Aboriginal Chronic Disease Package, 2008) and initiatives under the ‘Closing the Gap’ policy [2, 9]. The ineffectiveness of health programs has been attributed to multiple factors, including short government funding cycles, a lack of community ownership and consultation, and a ‘one size fits all’ approach to program design and implementation [10, 11]. Furthermore, only a small proportion of Aboriginal health programs have been evaluated (8%), with only 6% of program evaluations applying rigorous evaluation methodologies to measure program effectiveness [12, 13]. The paucity of Aboriginal health program evaluations has resulted in little opportunity to improve or modify existing programs in response to program outcomes, contributing to the cycle of program ineffectiveness.

The need for Aboriginal community-driven programs and governance of primary health care services, as supported by the international right to self-determination for Indigenous people, is becoming increasingly recognised as a key strategy to alleviating the burden of chronic disease [14,15,16]. Strong evidence supports the role of Aboriginal Community-Controlled Health Organisations (ACCHOs) and other Aboriginal organisations in improving the accessibility, appropriateness and effectiveness of primary health care services through the provision of culturally appropriate care which respects the cultural values and beliefs of Aboriginal people [8]. Therefore, the involvement of ACCHOs and other Aboriginal organisations in the design and implementation of chronic disease programs is imperative [15, 17,18,19]. Moreover, a community-based approach to program design, implementation, evaluation, sustainability and transferability acknowledges the diversity of Aboriginal culture, language and customs [10]. This ensures that chronic disease programs are tailored to local needs and evaluated in partnership with community, recognises the strengths and resilience of Aboriginal people, and empowers Aboriginal communities to promote their own health and wellbeing [20].

Although there has been a rhetorical shift from government initiated health programs to community-developed health programs for Aboriginal people in Australia [21], it is not known whether the distribution of chronic disease programs has been proportionate to the population distribution of Aboriginal people, or to the burden of chronic disease. Furthermore, it is not known geographically where evaluated chronic disease programs for Aboriginal people in the primary health care setting have been implemented. Therefore, the purpose of this scoping review was to identify where evaluated chronic disease programs have been implemented in the primary health care setting [22]. Specifically, this review sought to determine whether this distribution was proportionate to the Aboriginal population distribution, and burden of disease across all geographical areas of Australia and by doing so, highlight geographical gaps in the literature to identify priority areas for the implementation of chronic disease programs. Secondary objectives included scoping for evidence of partnerships with ACCHOs and other Aboriginal organisations, in addition to the use of ethical guidelines or protocols in the reporting of programs.

Methods

This study provides a systematic scoping review geographically examining the distribution of evaluated chronic disease prevention and management programs implemented for Australian Aboriginal people in the primary health care setting which includes community-health settings, general practice clinics and ACCHOs [22]. This review was undertaken in accordance with the methodology for conducting scoping reviews as outlined in the Joanna Briggs Institute Reviewers’ Manual 2017: Methodology for JBI Scoping Reviews [23]. Search terms were designed in a PCC (Population, Concept, Context) format by the research team and in collaboration with a health librarian. The premise and methods for this review have been published elsewhere in greater detail [22]. The Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines were adhered to in the reporting of this review (Additional file 1).

Search strategy

A preliminary search was conducted in MEDLINE and CINAHL using keywords to develop a tailored search strategy for each information source. A combination of Boolean operators, truncations and Medical Subject Headings (MeSH) were used to develop database search strategies (Additional file 2). The following databases were searched: Ovid MEDLINE, CINAHL (EBSCOhost), Scopus, Embase (Elsevier), Cochrane Database of Systematic Reviews, ISI Web of Science, SocINDEX (EBSCO-host), Sociological Abstracts (ProQuest), PubMed Central and PsycINFO (OVID).

Keywords were used to search the following information sources for unpublished studies, grey literature, and trials in order to avoid publication bias: Lowitja Institute, Indigenous Healthinfonet, National Aboriginal Community Controlled Health Organisation (NACCHO), Department of Health (Australian Government), informIT, Google, Cochrane Central Trials Register of Controlled Trials, ANZ Clinical Trials Registry, ClinicalTrials.gov, WHO International Clinical trial Registry Platform (ICTRP), Primary Health Care Research and Information Service (PHCRIS), ProQuest Dissertations and Theses Global, Trove and OAIster.

Inclusion criteria and exclusion criteria

This review considered literature based on the following criteria (Table 1).

Table 1 Inclusion and exclusion criteria
Full size table

No restrictions were placed on the quality of evaluation or study design. As stated in the scoping review protocol, programs evaluated by any party to any level were included [22]. Literature published from 1 January 2006, was included in order to capture programs published since the launch of the ‘Closing the Gap’ campaign, which resulted in a greater focus on addressing health inequities experienced by Aboriginal people in Australia [9].

For consistency, the term ‘Aboriginal’ has been used throughout this review to refer to both Aboriginal and/or Torres Strait Islander people in Australia. This is due to brevity, and no disrespect is intended to any individual or group. The term ‘Indigenous’ has been reserved for the global context.

Study selection and data extraction

Searches for published and unpublished literature were conducted by a health librarian. Titles and abstracts retrieved from the search were screened independently by two reviewers (HB and MJB). Conflicts were resolved through discussion with a third reviewer (CK). Full text review and data extraction was then conducted independently by two reviewers (HB and MJB) on selected articles. Reasons for exclusion were provided for articles that did not meet the review criteria. The reference lists of citations requiring full text review were also screened for additional citations in order to ensure that all possible literature was included.

Extracted data were categorised under the following headings: author, year of publication, year of program implementation, location of program implementation/evaluation, evaluation methods, involvement of an ACCHO/other Aboriginal organisation and reference to Australian National Health and Medical Research Council’s (NHMRC) ‘Values and Ethics: Guidelines for Ethical Conduct in Aboriginal and Torres Strait Islander Health Research’ guideline, and other ethical protocols [24, 25]. Geographical coordinates were then assigned to included programs based on the extracted data. Where specific implementation sites were not stated, the approximate location(s) were geocoded and coordinates extracted. If this information was unavailable, the corresponding author was contacted. If studies did not specify where the program was evaluated, the institution listed for the first author was used as a proxy for place of evaluation. This assumed that first authorship implied a lead role in the evaluation.

Coordinates were then exported to ArcGIS® ArcMap™ and overlayed with the Remoteness Areas of Australia for analysis [26]. To define remoteness, the Australian Statistical Geography Standard (ASGS) was applied, which is a categorisation of the Accessibility/Remoteness Index of Australia (ARIA+) [27]. Areas are classified as: i) Major Cities of Australia, ii) Inner Regional Australia, iii) Outer Regional Australia, iv) Remote Australia, and v) Very Remote Australia. Euclidean distance between the implementation site and evaluation were also extracted in ArcGIS® ArcMap™ [26]. Summary statistics were produced to examine the distance between implementation site(s) and place of evaluation. Locations of implementation and evaluation were stratified by Remoteness Area and cross-tabulated.

The extracted data, synthesis of findings and review outcomes, were critically reviewed for culturally appropriateness by two Aboriginal researchers, as stated in the review protocol [22].

Results

Database searches yielded 14,366 citations. An additional 241 citations were retrieved from a search of the grey literature and targeted organisation websites. A total of 6894 title and abstracts were screened, with duplicates removed. The full texts of 314 citations were screened for relevance to the review criteria, identifying 74 pertinent records evaluating 50 chronic disease prevention and management programs (Fig. 1 – PRISMA Flow Diagram). One of these records included reference to three evaluated programs [28].

Fig. 1
figure 1

PRISMA diagram of the systematic review process for this review

Full size image

Reasons for excluding records were provided (Additional file 3). The most frequent reason provided for exclusion was that the record was ‘not a program evaluation’ (n = 52), followed by ‘Sub-studies met inclusion criteria but already included in search, or sub-studies did not meet inclusion criteria’ (n = 50) and ‘not a primary health care delivered program’ (n = 30). Excluded records included 20 records which focused on evaluating national screening and vaccination programs. These were excluded as findings were based on national or state-wide aggregate data which would have been difficult to include in the geographical analysis.

Finding 1: heterogeneity of included programs

Citations meeting the review criteria (n = 74) included evaluated programs (n = 50) that addressed multiple chronic diseases (n = 16), a specific chronic disease (cardiovascular disease n = 5, diabetes mellitus n = 6, chronic kidney disease n = 3, liver disease n = 2, mental illness n = 4, oral disease n = 2 and polycystic ovarian syndrome n = 1) or risk factors for developing chronic disease (drug and alcohol misuse n = 2, poor nutrition and physical inactivity n = 3 and smoking n = 6) (Table 2). Of the included programs, 74% (n = 37) aimed to prevent and/or manage chronic disease using disease-specific screening, early intervention or treatment strategies, with the remaining programs (n = 13) applying general health promotion approaches to disease prevention, such as empowering participants to implement activities to improve their health.

Table 2 Characteristics of included program evaluations
Full size table

The data collection methods of program evaluations varied, with over half of the program evaluations using a mixed methods approach (n = 26, 52%), followed by a quantitative only (n = 19, 38%) or qualitative only (n = 4, 8%) approach. The methods of evaluation were not reported for one program [28], however, a summary of outcomes were provided; hence the program was included in the review. Of the included programs, only seven were evaluated using a randomised controlled trial (RCT) study design, with only one study (an RCT) including an economic evaluation of program cost-effectiveness.

Finding 2: geographical distribution of programs

Four of the included programs were excluded from the geographical analysis as programs were implemented state-wide or nationally (Home Medicines Review Program [94, 95], Get Healthy Service Program [75], QAAMS Program [77,78,79] and COACH Program [82]). Five of the included programs were also excluded from the geographical analysis as authors did not respond to the request for additional information. Geographical coordinates for program implementation sites were available for 41 of the included programs (82% of all included programs). However, one program was omitted from the analyses as the evaluation was undertaken outside of Australia as part of a multi-site program evaluation [47]. A total of 81 implementation sites for the 40 programs (80% of all included programs) with available locations were geo-coded and geographically analysed (Table 3).

Table 3 Geographical location of included programsa
Full size table

Of the included programs in the geographical analysis (n = 40), 32.1% were implemented for Aboriginal people residing in Major Cities of Australia and 29.6% for Aboriginal people residing in Very Remote Australia. The remaining programs were implemented for Aboriginal people residing in the intermediate remoteness areas of Inner Regional Australia, Outer Regional Australia, and Remote Australia (12.3, 18.5 and 7.4% respectively).

The location of program evaluation was reported for 25 of the programs included in the geographical analysis. First author affiliation was used as a proxy for the location of program evaluation for the remaining 15 programs. Evaluation activity was predominately undertaken in Major Cities of Australia (71.6%), with the remaining studies declining in order of remoteness. Of the identifiable implementation locations (n = 81), 18 (22%) of these also had an evaluation undertaken on site. For studies with the site(s) of implementation and evaluation available, the mean distance between implementation and evaluation was 660 km (95% CI 470–850; maximum 3041; median 223). A visual representation of the distribution of included programs is provided (Fig. 2).

Fig. 2
figure 2

Locations of the implementation and evaluation of programs in relation to the Remoteness Areas of Australia [27]

Full size image

The sample size of programs retrieved for highly prevalent chronic diseases, such as cardiovascular disease and T2DM, was deemed insufficient to geographically analyse whether the distribution of evaluated programs were proportionate to the burden of chronic disease across all Remoteness areas.

Finding 3: ethical approaches to program evaluation

Of the 50 programs included in the review, 39 (78%) reported on the involvement of an ACCHO in the implementation or evaluation process (Additional file 4). Of the included programs that did not report on the involvement of an ACCHO, six of these referred to the involvement of another Aboriginal organisation (12%). Overall, 90% (n = 45) of the included programs collaborated with an Aboriginal organisation in the implementation and/or evaluation of the program.

When examining the affiliation of the first author of the included citations (n = 74), 74% (n = 55) were associated with a university or research institution, 9.5% (n = 7) with an ACCHO or other Aboriginal organisation, 9.5% (n = 7) with a non-Aboriginal health service or Non-Government Organisation (NGO) and 7% (n = 5) with both a university or research institution and an ACCHO.

Of the 74 citations retrieved, seven explicitly referred to the NHMRC’s ‘Values and Ethics: Guidelines for Ethical Conduct in Aboriginal and Torres Strait Islander Health Research’ as a guideline underpinning the evaluation design and conduct [24]. However, 70% of citations (n = 52), particularly those published in a peer-reviewed journal, included a formal statement of ethical review and approval by a Human Research Ethics Committee (HREC) affiliated with a research institution or university. Only 23 citations (31%) discussed the use of other ethical protocols or a community-based ethical review process. These citations varied broadly in their descriptions of adhering to local cultural guidelines or consulting with an appointed Aboriginal advisory group. For example, Askew et al. (2016) [32] described the formation of a research advisory group consisting of both Aboriginal community members and experienced researchers who provided research governance and oversight, whereas Treloar et al. (2018) [98] described consulting with an Aboriginal advisory group in the program development phase rather than the evaluation process. Consideration of cultural sensitivities was also discussed broadly in some papers, including processes undertaken to build rapport with collaborating Aboriginal communities prior to the conduct of an evaluation [59, 72] and the receipt of cultural guidance or support from a steering group of Aboriginal people or Elders [40, 83].

Discussion

This review highlights the paucity of Aboriginal chronic disease program evaluations conducted in the primary health care setting across all geographical regions of Australia. Previous studies have acknowledged that only a small proportion of Aboriginal health programs have been subject to an evaluation process [12, 13]. Therefore, the included programs in this review are not representative of all chronic disease programs implemented for Aboriginal people across Australia. Of those included, the majority targeted highly prevalent chronic diseases, such as cardiovascular disease and T2DM, or risk factors for developing chronic disease, such as smoking and physical inactivity [2]. Due to the heterogeneity of programs across all geographical regions and small sample size, the review was unable to ascertain whether the spread of programs was proportionate to the distribution of chronic disease across all Remoteness Areas of Australia. For example, it was difficult to conclude whether a greater focus on the management of T2DM for Aboriginal people residing in Very Remote areas is required, where the prevalence of T2DM is approximately twice of Aboriginal people residing in Major Cities [6]. However, the small proportion of evaluated social and emotional wellbeing (SEWB) programs (e.g., mental health programs) was noted across all geographical regions, supporting the need for tailored early intervention and screening SEWB programs for Aboriginal people [19]. Internationally, tailored programs for mental health prevention have been deemed particularly important for Indigenous people, particularly those including an exploration of cultural identity [102, 103].

Overall, Major Cities and Very Remote areas of Australia displayed similar levels of chronic disease program implementation activity, with less activity noted for Inner and Outer Regional (IOR) areas and Remote areas of Australia. A greater focus on chronic disease programs for Very Remote Aboriginal people when compared to IOR or Remote Aboriginal people could be informed by national data which indicates that the burden of chronic disease in Aboriginal people increases with Remoteness [6]. However, less Aboriginal people reside in Very Remote areas when compared to IOR areas (11.9% compared to 23.7 and 20.3% respectively), suggesting there is a need for the evaluation of chronic disease programs for Aboriginal people residing in IOR areas [101]. Across all geographical areas in Australia, it is anticipated that the demand for chronic disease prevention programs will increase over time, due to a higher Aboriginal population growth rate when compared to non-Aboriginal populations as indicated by 2017 national data (2.26 babies per Aboriginal woman compared to 1.75 babies per non-Aboriginal woman) [104, 105]. The demand for chronic disease programs may also increase for Indigenous people in other countries (e.g., Canada) experiencing similar population growth (2.2 babies per Aboriginal woman compared to 1.6 babies per non-Aboriginal woman in Canada) [106].

When considering evaluation activity, higher levels of evaluation were noted for Major Cities (71.6%) when compared to Very Remote areas (1.2%). This is despite the fact that national data indicates that less Aboriginal people reside in Major Cities compared to the total Australian population (37% compared to 73% respectively) [101]. Further to this, the proportion of Aboriginal people is higher in all other Remoteness Areas of Australia, relative to the total Australian population [101]. This finding suggests there is a need for more Aboriginal community-led research as supported by the broader literature [15, 19, 20]. However, caution should be applied in interpreting these findings as first author affiliation was used as a proxy for the location of program evaluation for 15 of the 40 programs included in the geographical analysis. The rationale for this assumption was that first authorship implied a lead role in the evaluation.

When examining first author affiliation for all included citations (n = 74), 74% (n = 55) of citations were associated with a university or research institution, with only 9.5% (n = 7) citations associated with an ACCHO or other Aboriginal organisation. A previous review of Aboriginal health programs in Australia also found that the majority of program evaluations (72%) were led by a research institution or university rather than an Aboriginal community organisation [107]. However, first author affiliation with a research institution or university does not necessarily mean that the evaluation did not have significant Aboriginal community input; particularly as 90% (n = 45) of included programs provided details of collaborating with an ACCHO or other Aboriginal organisation in the development or evaluation of the program. Strong support for the appropriateness of ACCHOs as a collaborating organisation for activities involving Aboriginal people is found in the literature [108]. Generally, ACCHOs are geographically accessible to Aboriginal people and valued for the provision of culturally safe primary health care [8, 109, 110].

Although the majority of programs partnered with an ACCHO or Aboriginal organisation, it is difficult to ascertain for all programs, the degree of community ownership and involvement in the evaluation process. This includes steps taken by evaluators to ensure the evaluation process was ethically and culturally appropriate for Aboriginal people [20]. As reporting the formal ethical review of a research project is a standard requirement for publication in a peer-reviewed journal, both nationally and internationally, it is not surprising that the majority of citations (70%, n = 52) provided a statement of formal review by an appointed committee (e.g. HREC). However, only 31% of the included citations (n = 23) provided some evidence of actions taken to adhere to Aboriginal community-based ethical protocols, or engagement with an Aboriginal advisory group in the design of the program or conduct of the evaluation. Indeed, a statement of formal ethical review does not provide sufficient detail describing how Aboriginal people were consulted and included in the evaluation process. Other Aboriginal program evaluation frameworks and models of Aboriginal health research should also be consulted, which are valuable in informing approaches to conducting program evaluations in partnership with Aboriginal people [20, 25, 111, 112]. Program evaluations of Aboriginal programs excluding partnerships, often lack relevance and integrity, and fail to translate to outcomes for Aboriginal people [12, 113].

Limitations

It is important to acknowledge the limitations of this review. The selection criteria of the review influenced the geographical spread of studies retrieved. National screening and vaccination programs were excluded as program evaluations used national aggregate data. Geographical findings may also have been impacted by the exclusion of other programs which met the review criteria, but were excluded from the geographical analysis due to state-wide or national program implementation (Home Medicines Review Program [94, 95], Get Healthy Service Program [75], QAAMS Program [77,78,79] and COACH Program [82]). Furthermore, authors of five programs did not respond with additional information regarding the geographical program implementation locations which may also have influenced the analysis.

It is not known what proportion of evaluated chronic disease programs or implemented chronic disease programs have been included; a limitation cited by a similar review [107]. It is also possible that evaluated programs targeting more distal risk factors for chronic disease may have been overlooked. The availability of evaluation reports may also have influenced the types of citations retrieved. A recent investigation into the evaluation of health programs implemented for Aboriginal people in Australia found that only 33% of evaluation reports were available [20]. Further to this, it is acknowledged that a substantial amount of literature pertaining to Australian Aboriginal people is published in the grey literature [114]. Although the authors have made every effort to conduct a thorough search of the grey literature, it is possible some evaluation reports may not have been captured in this scoping review.

Recommendations

A greater focus is required on evaluating chronic disease prevention and management programs for Aboriginal people across all geographical areas, particularly for Aboriginal people residing in Inner and Outer Regional areas of Australia. In addition, there is a need to focus on evaluating Social and Emotional Wellbeing (SEWB) programs developed for Aboriginal people. Programs should be implemented and evaluated in collaboration with partnering ACCHOs or other Aboriginal organisations, with an emphasis on accountability, sustainability, capacity-building, ownership and Aboriginal strengths. This includes equipping Aboriginal organizations with skills in conducting program evaluations. Evaluation reporting should be transparent in describing ethical approaches to conducting the program evaluation in partnership with Aboriginal communities. Furthermore, an evaluation process should be integrated into the design of Aboriginal health programs. Evaluation outcomes should be publicly available, ideally through the peer-reviewed literature, in order to build the evidence around the effectiveness of chronic disease programs for Indigenous peoples globally.

Conclusions

A greater focus on the implementation and evaluation of chronic disease prevention and management programs for Aboriginal people in Australia is required, particularly for Aboriginal people residing in Inner and Outer Regional Areas of Australia. There is also a need to conduct evaluations of Social and Emotional Wellbeing (SEWB) programs across all geographical regions. This review highlights the need for more ethically rigorous approaches to Aboriginal health program evaluations which engage Aboriginal people in all stages of program design, implementation, evaluation and sustainability.