Every 3 minutes, someone in Canada is diagnosed with diabetes, while 11 million are living with diabetes or prediabetes [1]. About 6.1% of Canadian adults aged 20-79 years have prediabetes, putting them at high risk of developing type 2 diabetes [1]. The prevalence of diabetes is expected to increase to 5 million (12.1%) by 2025 [1]. Ontario has more people living with diabetes than anywhere else in the country. Rates of type 1 and type 2 diabetes have increased by 42% since 2009 [1]. There are now an estimated 4,424,000 people with diabetes or prediabetes in Ontario [1]. Diabetes Canada states that the costs of treating diabetes have surpassed CAN $30 billion [2] (a currency exchange rate of US $0.70). Goeree et al [3] have estimated that the excess attributable cost per incident case per person of diabetes in Ontario is CAN ~$2930 in the first year after diagnosis and CAN $1240 in the following years [3]. In Canada, diabetes disproportionately impacts First Nations and Métis people, and people of African, East Asian, and South Asian ethnic backgrounds, who experience higher rates of type 2 diabetes compared to the general population [4]. Inequities in the social determinants of health (eg, income, education, and housing), resulting from the impacts of systemic racism, intergenerational trauma, and colonization, are associated with higher rates of type 2 and gestational diabetes in priority populations [4].

Health equity is created when individuals have a fair opportunity to reach their fullest health potential. It is directly correlated with the distribution of resources within any setting [5]. Low socioeconomic groups with multiple health conditions and limited access to resources have decreased access to achieving health equity [5]. Intersectionality (Figure 1) is a theoretical framework in which consideration of heterogeneity across different intersections of social positions is critical to the conceptualization of health and social experiences [6]. The framework was published by Crenshaw [7] and developed within Black feminist theory to better explicate the situation of Black women in the United States [8]. It encompasses a wide range of intersections of ethnoracial groups, gender, socioeconomic status, sexual orientation, and other social identities and positions [7]. The framework notes that social positions exist on a hierarchy of social power and are not independent [7] but rather that they shape human experience jointly. As social positions intersect at the individual level (eg, income, age, and rurality), experiences at those intersections are influenced by larger interpersonal and structural systems of oppression, such as racism and sexism [9]. The synergies of oppression are created through the intersections of three domains: (1) socioeconomic status (ie, income), (2) identity-isms (eg, race), and (3) geography (segregation). Intersectionality in this conceptual framing (Figure 1) is an intersectional categorical axis where social determinants of health, systems of oppression, and environmental factors are integrated and three categories or framings within the context of the synergy of oppressions, specifically the outcomes of increased marginalization through oppressions and intersections of social determinants of health. This synergy is further compounded with access to health equity [7-9].

Diabetic retinopathy (DR) is the primary vision complication caused by diabetes [10] and is the leading cause of new cases of blindness in adults aged 20-65 years [11], yet it is often asymptomatic in the initial stages [12]. The prevalence of DR in Canada ranges from 20% to 30% [12]. In 2016, more than a million Ontarians were affected by DR [6]. The prevalence of vision loss in Canada is expected to increase by nearly 30% in the next decade [13]. Lower income and type 2 diabetes are associated with increased odds of visual impairment [14]. In 2007, the cost of vision loss in Canada was estimated at CAN $15.8 billion and is projected to grow to CAN $30.3 billion by 2032 [12]. In addition to these costs, the Canadian National Institute for Blindness estimated the cost of associated complications of vision loss: falls CAN $25.8 million, depression CAN $175.2 million, hip fractures CAN $101.7 million, and nursing home admissions CAN $713.6 million [12]. Early detection ensures that treatment or intervention can decrease the incidence of vision impairment or blindness. Evidence notes that one-third of adult patients with diabetes did not receive an eye examination for DR within 2 years [6], and more specifically, 25.3% of people with diabetes over 60 years had not seen an eye care provider in the last year [12]. Tele-retina is one of the DR screening modalities that can be used to bring the testing to community settings. It is focused on reducing eye care disparities that lead to avoidable vision loss. Tele-retina is a branch of telemedicine that delivers eye care remotely. Retinal images and data are collected and transferred via telecommunication technology to eye specialists [15]. Previous research has noted that it is a more cost-effective alternative than the standard screening in detecting DR among lower-income individuals with limited access to eye care [15]. Specifically, the cost per case correctly detected was CAN $281.10 with tele-retina and CAN $982 with the standard of care (SOC) screening (note, the standard of care screening for DR entails walking into an optometrist or ophthalmologist office to get the eyes checked), and the cost per case correctly diagnosed was CAN $82.21 and CAN $314.14, respectively. For both pilot and pan-Ontarian sensitivity analyses, tele-retina remained the dominant strategy (ICER <0). Unfortunately, screening rates among at-risk populations remain below 60% [12,15].

For DR screening, we know that socioeconomic status is the primary driver of risk for DR screening and that systemic societal barriers, for example, ageism, racism, and sexism [7,8], have had a detrimental impact on eye health and vision. Lack of education or information, social role and identity, competing priorities (getting time off work or study to attend appointments), language barriers, and lack of family and clinical support [9,10] are noted by patients as barriers associated with access to DR screening.

Differential access to retinopathy screening was evident when we evaluated the designed tele-retina model of DR screening. We found that only 50% of low-income individuals living with diabetes with limited access to vision care would get screened through SOC screening (walking into an optometrist or ophthalmologist office to get their eyes checked, which may be even more challenging in remote and rural areas), whereas 80% of the same population would get screened with the tele-retina program [11]. Screening percentage was the major driver of tele-retina’s economic dominance. This further suggests the need to elucidate social factors associated with the inequity of screening, as ignorance of these remains a major factor in perpetuating lack of access to equitable care.

In collaboration with patient partners, the primary objective of this project is to assess the equity impacts of tele-retina screening in populations that experience systemic social and economic disadvantages with limited access to conventional eye care.

More specifically, we will:

Understanding the equity impact of the tele-retina program will contribute to the creation of socially and culturally acceptable economic milieu in line with the multicultural Canadian context, with the ultimate purpose of improving the screening rate for DR, augmenting the quality and access to DR care, and decreasing the incidence of severe vision loss ultimately leading to blindness.

This study will build on our current intersectionality work for DR screening access among those who identify as women of low socioeconomic status [16]. Meanwhile, combining the advantages of multiple designs and enabling a comprehensive understanding of social complexities related to the implementation context, we propose the deductive theoretical drive sequential multimethod approach. Multimethod design entails conducting two or more research methods rigorously and completely in one project, where the major research question drives the study, but the approach consists of 2 interrelated studies. The deductive theory-driven approach indicates that we begin with an (intersectionality) theory, develop a hypothesis from that theory, and then collect and analyze data to test our hypotheses. The sequential approach indicates quanàQUAN, encompassing two quantitative methods used sequentially, one of which is dominant (in capital letters) [17,18]. The two components of this study are (1) modified Delphi and (2) DCEA study. Once the first project (modified Delphi) is completed, we will use the results to provide details for the DCEA (Figure 2).

In the modified Delphi, we are seeking the opinion of participants as to whether social constructs may dictate one’s health (seeking) behavior. The Delphi will guide us in developing the set of social construct variables to be integrated into the tele-retina screening economic model. We anticipate 35-50 participants (patient representatives: n=20-25; field experts, including in health technology assessment [HTA], and clinicians: n=5-10; decision and policy makers: n=10-15) who will be purposely selected based on experience, training, and specialized areas of knowledge (Table 1). Please note that there is no standard size of the Delphi panel members, and it varies from 10 to 1000 (typically between 10-100) in published studies. However, due to data management difficulties and logistic issues (rounds of the survey), a panel with a 3-digit sample size is unusual [1,2]. Generally, a double-digit number close to 30-50 is considered optimal in concluding rounds for a homogenous Delphi [3,4]. Appropriate size depends on the complexity of the problem, homogeneity (or heterogeneity) of the panel, and availability of the resources. Apart from panel members with knowledge, some studies recruit members from diverse academic and practice backgrounds or involve end users in the process [5]. Delphi participants are all members of the Diabetes Action Canada (DAC) Network. Potential participants, who will be invited via Network email to participate [19], will consist of patient partners, researchers, diabetes specialists, primary care practitioners, nurses, pharmacists, data specialists, and health policy experts committed to improving the lives of persons with diabetes. The team has successfully partnered and collaborated with DAC to plan, execute, and evaluate research projects to improve patient outcomes and experiences [20]. Before joining the Delphi panel, each participant will be approached by a qualitative researcher and study lead to sign a consent form.

Below is a summary of the modified Delphi, as it is a well-established approach to answering a research question by identifying a consensus view across subject experts.

Once Delphi participants have selected the most relevant social constructs or variables to be included in the distributional cost-effectiveness model, we will rank them and select the 5 social constructs with the highest ranking to be considered for inclusion in the model.

In the following, the modified Love-Koh et al [19] and Asaria et al [23,24] approaches (stages A and B as described in detail below), directed by a senior health economist, will evaluate which DR screening intervention has the best “social value” based on selected social constructs in the modified Delphi, as these are deemed important to the participants in health care resource allocation. We will investigate how these fit with the implicit value judgments inherent in the original cost-per-quality-adjusted life years model formulation of the tele-retina screening program. Note that quality-adjusted life years are a measure of disease burden, including both the quality and the quantity of life lived [25]. The case study is based on the Toronto tele-retina screening program offered in partnership with 7 primary care organizations with a population focus on those located in low-income urban and rural communities with a high prevalence of diabetes and low DR screening rates, with more details described in Figure 3 [15]. The Toronto tele-retina model was selected because it is a validated economic model but does not provide information on equity in the distribution of costs and effects. Please note that the case study population includes individuals residing in Ontario who are unable to obtain government insurance coverage (non–Ontario Health Insurance Plan [OHIP]–insured patients), including those residing illegally, those new to Canada awaiting coverage, and those so marginalized that they may struggle to obtain the identification and home address required to receive OHIP.

The study is under review (24-6035.0) by the University Health Network (UHN) Research Ethics Board (REB). Please note that in accordance with the UHN REB will report any protocol modifications (eg, changes to eligibility criteria, outcomes, analyses) to relevant parties (eg, investigators, REB, trial participants journals, and regulators). As indicated in the earlier section of the manuscript, before joining the Delphi panel, each participant will be approached by a qualitative researcher and study lead to sign an informed consent. Data in the Delphi study will be deidentified and individuals will be identified using a unique study ID. In the compensation process of patient partners, we follow the regulations and protocols of DAC.

This protocol outlines a study designed to provide information on the inequality impacts of tele-retina screening programs in at-risk patient populations. Understanding the equity impact of the tele-retina program will inform decision makers and program administrators of what may be necessary for creating a contextually and culturally acceptable screening program in line with social context, with the ultimate purpose of improving the screening rate for DR and thus decreasing the incidence of severe vision loss for the at-risk populations. This will be the first study in collaboration with patient partners to mainstream how we assess the economic impact of screening programs and interventions with respect to inequality and access to care for women, people from lower socioeconomic groups, or people from certain cultures or racial backgrounds while remaining focused on economic analysis and the best use of care resources. The study has substantial external validity as the social construct-guided economic model (DCEA) can be applied across disciplines and domains and a range of care settings when evaluating large-scale public health programs that have an explicit goal of tackling health inequality. The “de-novo” integration of social constructs into economic models may be transferred to other screening programs (breast, colon, etc), care settings (primary, secondary, and tertiary), and models of care. We plan to expand this approach to different stroke initiatives.