Economic evaluations of eye care services for Indigenous populations in high-income countries: a scoping review

Abstract

Background

Indigenous people in high-income countries have worse eye health outcomes when compared to non-Indigenous people, contributing to ongoing socioeconomic disadvantage. Although services have been designed to address these disparities, it is unclear if they have undergone comprehensive economic evaluation. Our scoping review aimed to identify the number, type, quality, and main findings of such evaluations.

Methods

MEDLINE, Embase, Web of Science, Cochrane Library Database, the National Health Service Economic Evaluation Database, EconLit, and relevant grey literature were systematically searched as per our pre-registered protocol. All economic evaluations of real or model services designed to meet the eye care needs of Indigenous populations in high-income countries were included. Two reviewers independently screened studies, extracted data, and assessed quality using the Quality of Health Economic Studies instrument.

Results

We identified 20 studies evaluating services for Indigenous populations in Australia (n = 9), Canada (n = 7), and the United States of America (n = 4). Common services included diabetic retinopathy (DR) screening through fundus photographs acquired in local primary health care clinics (n = 7) or by mobile teams (n = 6), and general eye care through teleophthalmology (n = 2), outreach ophthalmology (n = 2) or an Indigenous health care clinic optometrist (n = 1). These services were economically favourable in 85% of comparisons with conventional alternatives, mainly through reduced costs of travel, in-person consults, and vision loss. Only four studies assessed the benefits of increased patient uptake. Only five included patient evaluations, but none integrated these into their quantitative analysis. Methodological issues included no stated economic perspective (n = 10), no sensitivity analysis (n = 12), no discounting (n = 9), inappropriate measurement of costs (n = 13) or outcomes (n = 5), and unjustified assumptions (n = 15).

Conclusion

Several Indigenous eye care services are cost-effective, particularly remote DR screening. Other services are promising but require evaluation, with attention to avoid common methodological pitfalls. Well-designed evaluations can guide the allocation of scarce resources to services with demonstrated effectiveness and sustainability.

Trial registration

Our scoping review protocol was pre-registered (Open Science Framework DOI: https://doiorg.publicaciones.saludcastillayleon.es/10.17605/OSF.IO/YQKWN).

Background

Despite making up over 6% of the global population, Indigenous people experience political, economic, social, and health injustices contributing to their marginalisation within societies [1, 2]. While the majority live in low- to middle-income regions, Indigenous people in high-income countries also experience gross disparities in most indicators of well-being, including lower employment, income, education, and health status [1,2,3]. For example, the life-expectancy gaps between Indigenous and non-Indigenous populations are among the highest in the world in Canada (-12.5 years) and Australia (-10.0 years) [3].

Poor eye health is a well-described disparity experienced by many Indigenous populations in high-income countries. In Australia, extensive research, including two national surveys, have found higher rates of eye problems among Indigenous people when compared to the general population, including a three- to four-fold higher prevalence of vision loss in adulthood [4]. Studies in New Zealand [5], Canada [6, 7], and the United States of America (USA) [8], have similarly demonstrated eye health disparities among Indigenous communities. Many of these are attributed to reduced access to services, caused by geographical, financial, cultural, and other barriers [9, 10]. Accordingly, the Lancet Global Health Commission has identified the development of services that effectively prioritise and reach marginalised groups, such as Indigenous people, as a key priority in global eye health [11]. Such services, designed to improve access to eye care for Indigenous populations, will hereafter be referred to as Indigenous eye care services.

Within each country, limited resources are available to address a variety of competing health issues. To justify the use of scarce resources for Indigenous eye care services, it is vital that these have demonstrable value to individuals and society. Dunt et al. argues that such value can be demonstrated using three approaches: (1) health needs assessment based on disease epidemiology (e.g., prevalence); (2) economic evaluation; and (3) assessing the ability of a service to meet health performance benchmarks [12]. Notably, economic evaluations are increasingly recognised as essential tools for designing and implementing effective services which produce objective health benefits in a sustainable manner [13].

While Indigenous eye care services have value from a health needs and performance benchmark perspective, it is unclear whether they have undergone comprehensive economic evaluation [12]. We conducted the first known scoping review of this topic, aiming to identify the number and types of evaluations performed to date, and to summarise the reported economic impacts of specific services. This can guide policymakers and clinicians in making evidence-based decisions to support cost-effective services. We also appraised the methodological quality of these evaluations and identified knowledge gaps to inform future research.

Methods

This scoping review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for scoping reviews (PRIMSA-ScR) [14] and the Joanna Briggs Institute (JBI) Guidance for Conducting Systematic Scoping Reviews [15]. No deviations were made from our pre-registered protocol (Open Science Framework DOI: https://doiorg.publicaciones.saludcastillayleon.es/https://doiorg.publicaciones.saludcastillayleon.es/10.17605/OSF.IO/YQKWN). Two reviewers (M.M.N. and A.S.) independently screened reports for eligibility (title and abstract screening followed by full-text reviews), extracted data, and performed quality assessments. Discrepancies were resolved by consensus or with involvement of a third reviewer (H.R.).

Eligibility criteria

Eligibility was defined using the Population, Intervention, Comparator, and Outcome (PICO) framework [14]:

• Population: wholly or partially Indigenous populations, as defined by the United Nations Permanent Forum on Indigenous Issues [1], within high-income countries, defined by the World Bank (Appendix A) [16].

• Intervention: real or model diagnostic, preventative, or therapeutic eye care service. Studies were excluded if there was no indication of how the service was designed and/or implemented to meet the needs of an Indigenous population.

• Comparator: any or no alternative service.

• Outcome: an economic evaluation, defined as any measure of service costs and/or service outcomes (i.e., health, monetary, or other benefits produced by the service) reported by a cost-minimisation analysis (CMA), cost-effectiveness analysis (CEA), cost-utility analysis (CUA), and/or cost–benefit analysis (CBA) [17]. Evaluations from any economic perspective were included. This refers to the viewpoint from which costs and outcomes are analysed (e.g., from the perspective of individual patients, the healthcare system, or society as a whole).

There were no restrictions on publication status or year of publication. The following were excluded: reviews, case studies, commentaries, conference abstracts, reports with no full-text access, and reports unavailable in English.

Search strategy

A three-step strategy was conducted in consultation with a library and information scientist [15]. An initial search of MEDLINE using a preliminary strategy identified relevant reports. Index terms and keywords in the titles and abstracts were used to refine the strategy. MEDLINE (Ovid), Embase, Web of Science, Cochrane Library Database, the National Health Service Economic Evaluation Database (Ovid), and EconLit (EBSCO) were searched from inception to May 2023 using the refined strategy adapted for each database (Appendix B). No search limits or filters were applied. Database search records were imported into EndNote 20 and the deduplication function was used.

Grey literature was assessed through searches of Australian Indigenous HealthInfoNet, Vision 2020 Australia, Informit, the National Bureau of Economic Research, Canada’s Drug and Health Technology Agency, the Institute of Health Economics, the International Health Technology Assessment database, the International Agency for the Prevention of Blindness, and Google Scholar. Reference lists of all reviews and included reports were screened for additional reports.

Data collection and quality assessment

The following were collected into standardised, pre-piloted data forms: study setting (country, rural versus urban), population, design (trial, observational, model-based) and methodology (type and economic perspective of analysis, time horizon, methods of costing and evaluating), service provided, comparator/s, and findings (costs, cost-effectiveness ratios, incremental cost-effectiveness ratios, benefit–cost ratios, and sensitivity analysis). Time horizon refers to the duration over which service costs and outcomes were analysed. Patient evaluations were recorded to capture the value of services from an Indigenous perspective, which is often underestimated or ignored in traditional economic analyses [18]. Methodological quality was assessed using the Quality of Health Economic Studies (QHES) instrument, a widely used checklist with demonstrated construct validity [19]. Two reviewers (M.M.N and A.S.) assessed each study against the 16 weighted criteria in this instrument [19], scoring them as ‘Yes’, ‘No’, or ‘Not Applicable’. Discrepancies were resolved by consensus or with involvement of a professor in economics (I.L.). The scores derived from this checklist were used to categorise the quality of studies as very poor (0–24), poor (25–49), moderate (50–74), and high (75–100).

Data synthesis and analysis

Key study characteristics, findings, and quality were summarised through tabulation and narrative description. For comparability, all currencies were converted to international dollars using the purchasing power parity exchange rate for the year of pricing (or year of publication if pricing year was unreported) [20]. Prices were then inflated to 2023 using GDP implicit price deflators for the USA [21, 22]. A synthesis of issues with study quality and knowledge gaps was provided to inform future research practices and directions.

Results

Study selection and characteristics

Of the 3857 unique database records, 101 underwent full-text review and 11 studies were included (Fig. 1). An additional 79 records from grey literature and citation searching underwent full-text review, of which nine studies were included. Reports excluded after full-text review were mostly evaluating non-Indigenous services (n = 48), review articles (n = 40), or lacked an evaluation of a service’s costs and/or outcomes as per our eligibility criteria (n = 38).

Fig. 1

Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow diagram of study selection

Key characteristics and findings of the 20 included studies are described in Table 1. Studies were published between 1990–1999 (n = 3), 2000–2009 (n = 6), and 2010–2019 (n = 11). They evaluated services for Aboriginal and Torres Strait Islander people in Australia (n = 9), First Nations people in Canada (n = 7), and Native Americans in the USA (n = 4). Some services also included non-Indigenous Australians (n = 6) and Canadians (n = 2). Studies were set in rural communities (n = 14), both urban and rural communities (n = 1), or nationwide (n = 3). All studies evaluating real services (n = 10) were observational in design. Model services were based on published epidemiological, cost, and/or treatment outcome data (n = 6), or represented simulated expansions of real services in geographical scale (n = 3) or duration (n = 1).

Table 1 Characteristics and findings of economic analyses of Indigenous eye care services in high-income countries

The most common Indigenous eye care services were diabetic retinopathy (DR) screening using fundus photographs acquired by trained staff physically located within local primary health care clinics [23,24,25,26,27,28,29] or by mobile teams [30,31,32,33,34,35] and graded by ophthalmologists who were located offsite at other facilities (Fig. 2). This was followed by general eye care services provided via teleophthalmology [36, 37], outreach ophthalmology [38, 39], or an Indigenous health clinic optometrist [40]. Only six of these services explicitly involved Indigenous community members and/or health care workers in their design [26, 31,32,33, 35, 42]. Across the 20 studies, there were 27 distinct comparisons between an Indigenous eye care service and a conventional alternative [23,24,25,26,27,28,29,30,31,32,33,34,35,36,37, 40, 42]. Some studies included multiple comparisons of different varieties of Indigenous eye care services and conventional alternatives. Three studies compared different varieties of Indigenous eye care services with each other rather than with a conventional alternative [38, 39, 41]. Evaluations adopted the economic perspective of the healthcare system [23, 24, 26,27,28,29,30,31,32,33,34,35,36,37, 41], federal government [25, 40], society [37, 42], or had an unclear perspective [38, 39].

Fig. 2

Number and type of economic evaluations of Indigenous eye care services. Cost-minimisation analyses (CMA), cost-effectiveness analyses (CEA), cost-utility analyses (CUA), and cost–benefit analyses (CBA) were all used

DR screening using fundus photography in local primary health care clinics

These studies were based in both rural and urban areas of Canada [24, 26], the USA [23, 25], and Australia [27,28,29]. Four primary comparators were used: screening by local ophthalmology or optometry services with no patient or staff travel costs [23, 25, 27], screening by out-of-town ophthalmology or optometry services associated with costs of patients travelling to visit them [26, 28], screening by outreach ophthalmology or optometry services associated with costs of staff travelling to the local town [24, 28, 29], or no screening [27].

Fundus photography in local primary health care clinics was cost-saving compared to screening by either local or out-of-town ophthalmology or optometry services, with savings per patient screened of $71 [23], $10 [25], ≥ $210 [26], and $302 [28]. Through improved patient uptake, Whited et al.’s model service within Indigenous primary health clinics in the USA also reduced DR-related blindness by 13% compared to screening by local ophthalmology or optometry services [25].

Fundus photography in local primary health care clinics was cost-saving compared to screening by rural outreach services in two CEAs [24, 29], but more expensive in one CMA [28]. Maberley et al.’s model service in Canada saved $125 per patient and, through higher uptake, led to a 19% gain in QALYs from avoided DR-related blindness [24]. While Ballreich et al.’s model service in Australia saved $197 per patient, it detected 10% fewer cases of DR due to the assumptions of equal uptake but lower sensitivity than outreach optometry [29]. Kanagasingam’s real service in rural Australia costed at least $70 more per patient, but there were substantial methodological issues with their study, including omission of nursing and clerical staff costs of the comparator [28].

Ellery et al.’s model service of DR screening using fundus photography in local primary health care clinics in Australia was cost-effective at reducing DR-related vision loss compared to no screening or screening by a general practitioner using direct ophthalmoscopy [27]. While cost-saving compared to screening by local optometrists and ophthalmologists, the service was less effective due to the assumed equivalent uptake but lower sensitivity.

Mobile DR screening using fundus photography

There were four CMAs and one CEA in Canada [30, 31, 33,34,35] and one CMA in Australia [32]. The Canadian CMAs reported mobile services as cost-saving compared to patients travelling to out-of-town ophthalmologists, with savings per patient screened of $325 [30], $263 [31], $55 [33], and $1,029 [34] in the respective studies. The Canadian CEA reported that their model service saved $78 per patient while detecting 46% more DR than screening by local optometrists or ophthalmologists [35]. However, this study likely overestimated the cost-effectiveness of their service for two reasons: (1) mobile DR screening by fundus photography was assumed to be more sensitive and specific than screening by the comparator, which is inconsistent with published literature, and (2) the number of DR cases detected by the comparator was calculated assuming a screening rate of 55% (i.e., 15,675 patients screened), despite the cost being based on all 28,500 patients getting screened by an ophthalmologist. The Australian service saved $23 per patient compared to outreach ophthalmology, recouping the higher capital costs within 2.5 years [32].

General eye care services

Two studies evaluated teleophthalmology in rural Western Australia [36, 37]. Kumar et al.’s nurse-led store-and-forward teleophthalmology service was cost-saving compared to out-of-town ophthalmology consults ($189 saved per patient) and cost-neutral compared to outreach ophthalmology [36]. Razavi et al. reported that real-time teleophthalmology would save $1,137 and $191 per patient compared to out-of-town and outreach ophthalmology services, respectively [37]. Through clinical audits, they determined that 15% and 24% of out-of-town and outreach consults, respectively, could be provided by teleophthalmology, which would save the healthcare system $499,617 annually. Each out-of-town consult avoided was also estimated to allow two extra days of work and, based on the nation’s average income, this would generate $443,023 in annual productivity savings for society.

Turner et al. conducted two evaluations of nine outreach ophthalmology services across Australia [38, 39]. Services funded using fee-for-service appeared cost-saving compared to those using a fixed-salary model ($500 vs $771 per clinic or surgical attendance), although this was not statistically significant (p= 0.12) [38]. These services were more efficient, with 2.5-fold higher clinic and surgical outputs (p= 0.02 and 0.03, respectively). Services well-integrated with outreach optometry had similar costs to those with poor integration but trended towards higher clinic and surgical outputs and lower waiting times [39].

Lastly, Jaworski found that integrating optometry within an Indigenous health clinic in the USA produced positive monetary returns, generating $2.44 in billings for every $1.00 in costs [40].

Other services

Miller et al.’s CMA found that astigmatism screening of a preschool native American population by autokeratometry or autorefraction would begin resulting in cost savings after a minimum of 400 and 985 children, respectively, compared to screening by visual acuity [41]. These savings were attributed to a reduced number of false positive patients requiring follow-up eye exams, with each false positive exam costing $77 [41].

Lastly, a comprehensive CBA by PricewaterhouseCoopers (PWC) modelled the nationwide expansion of services to eliminate avoidable vision loss from cataract, refractive error, DR, and trachoma in Indigenous Australians [42]. Compared to preexisting services, the expansion produced a net incremental benefit of $298 million to society and $18 million to the government, with $2.55 and $1.09 in benefits, respectively, for every $1.00 spent expanding. Societal savings included the productivity gains from increased employment, at the national average income, of patients who would no longer have vision loss and the carers of such patients.

Patient evaluations of services

Five of the 14 studies on real or modelled expansions of real services underwent patient evaluations (Table 2) [26, 28, 31, 35, 36]. These evaluations were all conducted through unvalidated questionnaires designed by study authors and had variable response rates. Over 90% of respondents were very satisfied or satisfied with the services [26, 28, 35, 36], would reuse them [26, 31, 36], and/or recommend them to others [31]. Reported benefits included convenience, particularly in relation to avoided travel and rapid access [26, 28, 31, 35, 36], increased awareness of eye health [26, 35], and the use of trusted local staff [26]. A minority using a general teleophthalmology service were concerned that it was less comprehensive and provided delayed advice compared to an in-person service [36].

Table 2 Patient evaluations of Indigenous eye care services

Quality of studies

Figure 3 illustrates the proportion of studies meeting each item on the QHES checklist, with individual study scores detailed in Supplementary Table 1. Five studies were high quality [24, 25, 27, 37, 42], eleven were moderate [26, 29, 32,33,34,35,36, 38,39,40,41], three were poor [23, 30, 31], and one was very poor [28]. Most studies presented their objectives, methods, and findings clearly (QHES items 1, 10, and 12) and used appropriate sources for costs (real expenditures or local data) and health outcomes (randomised controlled trials) (items 3 and 7).

Fig. 3

Proportion of studies scoring for each item on the Quality of Health Economic Studies (QHES) checklist

Ten studies failed to explicitly state their economic perspective (item 2) and 12 did not conduct a sensitivity analysis (item 5). Eight annuitized capital cost without discounting [28,29,30,31, 33, 34, 36, 41], one did not discount any costs beyond one year [23], and two used a limited timeframe potentially missing some health outcomes of their service [25, 42] (item 8). Supplementary Table 2 outlines the types of costs and outcomes considered in each study. Thirteen studies had inappropriate or unclear methods for calculating costs (item 9). Specifically, three omitted capital costs completely [23, 26] or partially [42], five were unclear about the components of capital costs and whether these were annuitized [28, 33, 34, 38, 39], three were unclear about components of their operating costs [31, 38, 39], four omitted costs of some or all staff [24, 26, 30, 41] or treatment [24], and one included costs for patients not using their service [35]. Five of the seven non-CMA studies used invalid methods for assessing outcomes (item 11), including overestimating blindness avoided from DR screening [24, 27], overestimating service income [40], and calculating the cost per case of DR detected without including the savings associated with reduced vision loss once these cases are treated [29, 35].

The main assumptions and/or choice of economic model were unjustified in 12 out of 13 CMAs (item 13). Specifically, two omitted all capital costs [23, 26], four used a comparator which was more comprehensive than their service [31, 33, 36, 41], and two could not determine if their findings were confounded by other variables [38, 39]. Six of the eight CMAs of DR screening by fundus photography omitted the costs of consults for unreadable and abnormal photos [28, 30,31,32,33,34]. Conversely, three of the seven non-CMA studies had major assumptions that were unjustified [27, 29, 35]. Ellery et al. [27] and Ballreich et al. [29] assumed that uptake of their model DR screening services would be equal to conventional screening, despite increased availability being the main purpose of their services. This led to their services detecting fewer cases of DR. Stanimirovic et al. made assumptions likely leading to overestimation of the cost-effectiveness of their service (Supplementary Table 2) [35].

Discussion

Across all studies, we identified 27 comparisons between a service for Indigenous populations in Australia, Canada, or the USA and a conventional alternative. Indigenous eye care services were economically favourable in 23 (85%) of these comparisons, despite most omitting key benefits of culturally tailored care including increased patient uptake and value from an Indigenous perspective. Several common methodological pitfalls were identified, which should be avoided in future evaluations.

The primary economic value of services identified in our review arose from reduced costs of travel and in-person consults. Up to two thirds of Indigenous people in Australia, Canada, and the USA live outside major cities [47]. Traditionally, accessing eye care either requires patient travel to major cities or outreach services, both of which have high logistical costs borne by the health care system and patient [48,49,50]. Nine studies found that DR screening through fundus photographs, acquired by local primary health care clinics or mobile teams and graded offsite, led to health care savings by avoiding costs of patient travel [26, 28, 30, 31, 33, 34] or outreach services [24, 29, 32]. While the remaining four studies on DR did not consider travel costs, they found that these screening methods were cost-saving through avoiding expensive in-person optometry or ophthalmology consults [23, 25, 27, 35]. General teleophthalmology services in rural Australia also led to health care savings through reduced patient travel [36, 37] and outreach service expenses [37]. Many studies may have underestimated the savings from local Indigenous eye care services by not accounting for indirect costs associated with patient travel, such as the travel costs of companions (included in only two studies [26, 33]) and the productivity losses for patients when travelling (included in one study [37]). For example, the inclusion of productivity losses from patient travel led to an 89% increase in savings in the single study that analysed this [37].

The services identified adopted evidence-based strategies to improve accessibility and uptake by Indigenous populations, including local delivery of care, integration within Indigenous health clinics, and use of Indigenous health workers [9]. The positive patient evaluations of services further indicates that these strategies would improve uptake relative to conventional care. Despite this, only four studies evaluated the impact of increased patient uptake [24, 25, 35, 42], with the remainder potentially underestimating the cost-effectiveness of their services. For instance, among the eleven studies comparing local DR screening [24, 26, 28,29,30,31,32,33,34] or teleophthalmology [36, 37] to out-of-town or periodic outreach services, only Maberley et al. [24] included uptake as a variable, where higher screening rates increased QALYs by 19%. Among the four studies evaluating DR screening in local Indigenous health care clinics [23, 25, 26, 29], only Whited et al. [25] explored the benefit of integrated care on patient uptake. In their analysis, higher uptake contributed to reduced DR-related blindness and associated savings on health care and social welfare costs and increased income tax revenue. Lastly, among six services recruiting Indigenous health workers [29, 32, 33, 38, 39, 42], only PWC’s [42] evaluation explored the economic benefit of this strategy, whereby service coordination by Aboriginal Health Workers was modelled to increase patient uptake and reduce drop-out. This contributed to reduced vision loss with associated savings in health care and social welfare costs and increased societal income and tax revenue. To fully capture the economic benefits of Indigenous health programs, future studies must include patient uptake within their evaluations.

All studies adopted traditional methods of economic evaluation, which may underestimate the value of services from an Indigenous perspective [18]. While traditional evaluations focus on value derived from the health gain of individuals, Indigenous concepts of health extend beyond the individual to include the health and empowerment of their community, connections to land, and cultural security [18, 51]. For instance, an Indigenous-designed service is valued more by Indigenous people than one which delivers equal individual health benefits in a less culturally sensitive manner [52]. Despite this, none of the six studies involving Indigenous people in service design analysed the economic value of this collaboration [26, 31,32,33, 35, 42]. The New South Wales government recently published strategies to overcome these recognised limitations of traditional evaluations, such as the use of contingent valuation methods to quantify the value of different services from the perspective of Indigenous communities [53]. Other strategies which could be adopted by future studies include Indigenous-specific discrete choice experiments and health-related quality of life measures [51, 52].

Several other methodological issues limited the quality of studies included in our review. Future studies should include the economic perspective of their evaluation, sensitivity analyses, discounting items beyond one year, and clear methods for estimating capital and operating costs including sources used, components included, and annuitization. An economic evaluation checklist, which none of the studies explicitly used, may help ensure such essential items are included and avoid other methodological pitfalls [19]. As different eye care services are unlikely to have identical outcomes, CMAs should be avoided. Lastly, any study of screening or diagnostic services should use accurate estimates of sensitivities and specificities, as these significantly impact economic outcomes, as was the case in Stanimirovic et al.’s evaluation [35].

Our review also highlights the need for evaluations of multiple services yet to be analysed. Firstly, while DR screening through offsite grading of fundus photographs has been evaluated, automated grading through artificial intelligence can provide a cheaper, timelier service while maintaining acceptable accuracy [54, 55]. This could more easily be expanded nationwide to achieve universal screening, which should lead to extensive savings particularly for Indigenous people who have less access to screening despite a higher prevalence of DR [55]. Indeed, a study published in 2024 predicts that such screening methods among Indigenous Australians would generate a net societal benefit of $509 million dollars [56]. Evaluations of real-world applications of such services should be conducted to confirm these findings. A review by Burn et al. identified 37 studies on other, non-DR related service delivery models designed to improve access to eye care for Indigenous populations in high-income countries. [9] Most have not undergone economic evaluation, including mobile general ophthalmology services in Australia [57] and Taiwan [58], integration of optometry care within Indigenous health care clinics within Australia [59, 60], integration of preoperative and/or postoperative cataract assessments within Indigenous health care clinics [61] or optometry services [62], Indigenous spectacle subsidy schemes [59, 60], trachoma control programs [63], or culturally tailored health promotion activities [63,64,65]. While many of these have demonstrated potential to improve access, evaluations are needed to identify which represent the best value for money. Future studies should particularly focus on services which target the most common causes of vision loss in Indigenous populations, such as refractive error, cataract, and diabetic retinopathy [4,5,6,7,8]. Lastly, evaluations of Indigenous populations in other countries with documented disparities in eye health should be considered, such as those living in New Zealand, Taiwan, and Greenland [66].

Limitations

The lack of a meta-analysis prevented a statistical evaluation of the cost-effectiveness of services. However, given the limited number of heterogenous studies, a meta-analysis is unlikely to provide meaningful results [67]. Nine of the 20 studies included were conducted prior to 2010 and may be less relevant to modern times given changes to the costs of providing services. We minimised the impact of this through providing inflation adjusted results. Excluding the four poor and very poor-quality studies may have improved the relevance of findings summarised in our review. However, as a scoping review, we aimed to provide a comprehensive overview of all evaluations performed and highlight common issues with study quality that should be addressed in future research. Lastly, during the systematic search, we identified five studies which analysed the cost of real Indigenous eye care services without including outcomes or a comparator (Appendix C). Including cost-only studies may have provided useful data about these additional services but was beyond the scope of our review. Furthermore, such partial economic evaluations have limited value in decision-making, as they provide no indication of the value for money of a service [17].

Conclusions

Our review identified a variety of cost-saving and/or cost-effective DR screening, general ophthalmology, and optometry services for Indigenous populations in Australia, Canada, and the USA. Services that improve access to DR screening were particularly well explored and could substantially reduce avoidable vision loss among Indigenous populations. Future evaluations should include the economic impact of improved uptake and Indigenous concepts of health, while avoiding common methodological pitfalls, particularly those related to the assessment of costs and outcomes.