need help

Vol:.(1234567890)

Journal of Racial and Ethnic Health Disparities (2024) 11:2766–2785 https://doi.org/10.1007/s40615-023-01739-7

1 3

Income‑Related Inequities in Primary and Specialist Care Among First Nations Peoples Living Off‑Reserve in Canada

Mohammad Hajizadeh1  · Benjamin L. Keefe2 · Yukiko Asada3 · Amy Bombay4 · Debbie Martin5

Received: 8 March 2023 / Revised: 26 July 2023 / Accepted: 27 July 2023 / Published online: 10 August 2023 © W. Montague Cobb-NMA Health Institute 2023

Abstract Background Improving equity in healthcare is a primary goal of health policy in Canada. Although the investigation of equity in healthcare utilization is common in the general population, little research has been conducted to assess equity in healthcare utilization within First Nations peoples living in Canada. Objective To examine income-related inequities in primary care (family doctor/general practitioner and nurse practitioner care) and specialist care within status and non-status First Nations adults living off-reserve. Methods Using the 2017 Aboriginal Peoples Survey (APS), a nationally representative survey of Indigenous peoples living off-reserve in Canada, we analyzed income-related inequities in healthcare among Indigenous adults (>18 years) who self- identified as a member of any First Nations group in Canada. Logistic regression analysis was performed to identify factors associated with the utilization of primary and specialist care. The Horizontal Inequity index (HI), which measures unequal healthcare use by income for equal need, was used to quantify and decompose income-related inequities for primary and specialist care for status and non-status, and total First Nations groups. Results The regression results revealed higher primary and specialist care use among females, high socioeconomic status (high income and more educated) and status First Nations peoples in Canada. The positive values of the HI suggested a higher concentration of primary care and specialist care utilization among higher income First Nations peoples after adjusting for healthcare need. These pro-rich inequities persisted for the total First Nations populations, and for those in each status group individually. The decomposition results suggested observed inequities in both primary and specialist care among First Nations peoples can be predominantly attributed to the unequal distribution of education and income. Conclusion Although primary and specialist services in Canada are free at the point of the provision, we found pro-rich inequities in healthcare use among First Nations adults living off-reserve in Canada. These results warrant policies and initiatives to address barriers to healthcare use within and outside health system among low-income First Nations peoples living off-reserve.

Keywords Equity · Healthcare utilization · First Nations peoples · Canada

Introduction

Equity is a concept that underlies many health-related policy objectives for governments around the world [1, 2]. The con- cept of equity is applicable not only for health outcomes of individuals, but also in many areas within the healthcare system [3–5]. Although different countries place emphasis on different aspects of equity within their health systems, a shared perspective about the general meaning of the term “equity in healthcare” is that healthcare should be financed according to ability to pay and used based on healthcare need [6]. In Canada, improvements in the financing and delivery of healthcare have been guided by the Canada

* Mohammad Hajizadeh [email protected]

1 School of Health Administration, Dalhousie University, Halifax, Canada

2 Faculty of Medicine, Dalhousie University, Halifax, Canada 3 Department of Community Health and Epidemiology,

Dalhousie University, Halifax, Canada 4 Department of Psychiatry and School of Nursing, Dalhousie

University, Halifax, Canada 5 School of Health and Human Performance, Dalhousie

University, Halifax, Canada

2767Journal of Racial and Ethnic Health Disparities (2024) 11:2766–2785

1 3

Health Act (CHA, 1984) with the primary objective of facilitating adequate access to healthcare without financial or other barriers [7].

Although there is a long-standing interest in reducing inequities in healthcare utilization [8], inequity is very much present in Canada’s healthcare system. In a 2017 compari- son of healthcare systems for 11 Organization for Economic Cooperation and Development (OECD) countries, Canada placed ninth overall and ninth in healthcare system equity, which included measurements of inequity in 11 aspects of the healthcare system, from areas of both access (timeli- ness, affordability, etc.) and care process (patient engage- ment, preventative care, etc.). Canada’s cumulative score in this area placed them ahead of only France and the USA [9]. Indigenous populations in Canada (i.e., Inuit, Métis and First Nations), in particular, face several barriers to access healthcare services. For example, studies have shown that Indigenous populations in Canada have a relatively lower level of access to primary care in the form of general prac- titioners (GPs)/family physicians [10].

There are some studies (e.g., [11, 12]) that specifically examine the inequalities in healthcare utilization experi- enced by Indigenous populations in Canada. These studies, conducted in various Canadian provinces, have shown some differences in utilization patterns between Indigenous and non-Indigenous populations. This research tends to group Indigenous populations and generally reports inequality rather than inequity (unfair inequality), as need for health- care is not assessed. A study of Ontario’s First Nations popu- lation reported a higher rate of utilization for ambulatory care, as well as a lower rate of utilization for referral care (usually specialist care) when compared to the general popu- lation [13]. A systematic review focused on patients with arthritis showed that compared to the general population in Canada, Indigenous populations in Canada had comparable or (in some cases) greater use of primary care as well as comparable rates of hospitalization for arthritis related care [14].

While some studies have provided insight into inequi- ties in healthcare utilization between Indigenous and non- Indigenous individuals, the available evidence regarding inequities in healthcare use among Indigenous peoples is scant. It is crucial to understand and address the existing inequities in healthcare utilization within Indigenous popu- lations, as this remains an important concern [15]. Since financing of healthcare varies among First Nations peoples in Canada based on eligibility for federal health coverage provided through the Non-Insured Health Benefits pro- gram (NIHB), which is contingent on government-defined status, investigating inequities among First Nations can shed light on improving healthcare use within this popula- tion. Additionally, given the expressed interest from stake- holders, including the Assembly of First Nations (AFN),

investigating inequities in healthcare use in this population becomes paramount [16].

Current literature has established the importance of income as a determinant of healthcare utilization, and recent evidence [17, 18] points to higher healthcare use by more affluent individuals after adjusting for healthcare need (pro- rich inequities) in general population in Canada. The role of income can potentially be different among Indigenous populations given that their income is persistently lower than non-Indigenous populations: 2016 Canadian Census reports a gap of 25% in 2015 with little improvement since 2005 [19]. Differences in healthcare coverage and the con- text of colonialism can also lead to differences in income- related inequities in healthcare utilization among Indigenous populations. This study, for the first time, aimed to measure and explain income-related horizontal inequities (unequal healthcare use by income for equal need) in primary and specialist care use among First Nations adults living off- reserve (36% of the total Indigenous population [20, 21]) in Canada, looking at inequities within status (registered) First Nations and non-status (non-registered) groups. The find- ings of this study can contribute to a better understanding of inequities in healthcare use among First Nations peoples and provide policymakers with valuable insights to address these inequities.

Indigenous Populations in Canada

In Canada, there are three distinct Indigenous groups recog- nized by the Canadian Constitution: Inuit, Métis and First Nations [22]. First Nations are the largest Indigenous group and comprise almost 60% (977,235) of the population of Indigenous peoples in Canada [23]. There are over 600 First Nations across Canada each with distinct history and cul- ture. First Nations peoples in Canada can be further qualified based on whether or not they live in their home communi- ties and their “Indian status”. Not having “status” means that this individual has not met the qualifications laid out by the federal government under the Indian Act. These quali- fications are made based primarily on lineage, but there are other ways in which Indigenous individuals have historically been stripped of their status [24]. If an individual does meet these qualifications, they are registered in the Indian roll (a list of status individuals). The terms “status” and “regis- tered” are used interchangeably. First Nations peoples may also be members of a band. The legal definition of “a band is a group of Indians for whom land has been set aside (a reserve) or who have been declared a band by the Governor General” [24]. In total, the population of Indigenous peoples living in Canada is growing and, in 2016, was almost 1.7 million (5% of the Canadian population) [23]. In contrast to the general population, the population of Indigenous people

2768 Journal of Racial and Ethnic Health Disparities (2024) 11:2766–2785

1 3

is young, with an average age of 32.1 years, compared to the general population’s average age of 40.9 years [23]. The relatively younger Indigenous population living in Canada can be attributed to their higher birth rates [25] and lower life expectancy [25] compared to the non-Indigenous Cana- dian population.

The impact of colonization has had serious effects on the Indigenous populations in Canada, including a negative impact on health generally as well as their interactions with the health care system [26]. The loss of cultural traditions, being confined to reserves, being forced to attend Indian Residential Schools (IRSs), among other historical acts, have taken a toll on both physical and mental aspects of health for this population. Some of these differences in health have been attributed directly to these historical events [27]. In addition, current differences in the social determinants of health (lower quality of education, reduced employment opportunities, increased levels of poverty among others) contribute to poorer health status when compared to the non-Indigenous population in Canada [28]. This may mani- fest in many ways including decreased life expectancy [29], increased prevalence and incidence of chronic disease [30] and increased risk factors for those diseases [31].

Indigenous Healthcare in Canada

The evolution of the Canadian healthcare system began with separation of legislative power between federal and provin- cial governments in the Constitution Act, 1867 and culmi- nated with the CHA, 1984. Improvements in the financing and delivery of healthcare in Canada have been guided by the single primary objective outlined in the CHA, which is to “facilitate reasonable access to health services without financial or other barriers” [7]. In continuing to adapt and make improvements to the healthcare system in Canada, the satisfaction of this objective is paramount. Regulation of healthcare is primarily the responsibility of provincial and territorial governments in Canada, however, the federal government contributes financially to “medically neces- sary” health services. The CHA is the basis for the financial contribution from the federal government, which provides funding to the provinces and territories given they meet the criteria presented in the Act. This transfer of funds is called the Canada Health Transfer and is based on five criteria namely, universality, public administration, comprehensive- ness, portability, and accessibility. Although equity is not explicitly written in this piece of legislation, the criteria of universality and accessibility provide the basis for equity as a fundamental part of the system [3, 32, 33].

For most Canadians, healthcare is primarily publicly financed with approximately 70% of funds being contrib- uted by federal and provincial/territorial governments.

This figure has remained relatively constant for the past 20 years [34], with the other 30% being contributed privately through either private insurance or direct out-of-pocket (OOP) payments by the healthcare user. Publicly funded healthcare (referred to as Medicare) provides the insured access to “necessary” medical care. This varies by province and territory, but generally includes access to physician and hospital services and may additionally provide complete or partial coverage in several other areas, such as vision care, prescription drug coverage, ambulance services, and long- term/palliative care. Private insurance in Canada has varying degrees of coverage and is offered by several different pro- viders. This may include coverage for prescription medica- tion, dental care, rehabilitation, and various other types of healthcare provided outside of hospitals. This insurance is often provided by employers or purchased on an individual level and covers certain services that are not deemed “medi- cally necessary” and therefore not covered through provin- cial governments. Private insurance is more commonly acquired by those of higher socioeconomic status (SES), meaning that others may have to pay OOP for services that are left uncovered. The OOP payments for healthcare has led to non-adherence for uncovered treatments and services [35]. This disproportionately affects those of low SES in the population and may contribute to inequity in use of health- care [3, 35, 36]. Additionally, there are other factors which may influence inequities in utilization in this country. As healthcare is the responsibility of the provincial and territo- rial governments some variation in coverage is inevitable, with different jurisdictions placing value in different areas.

Although health coverage in Canada is provided through provincial/territorial governments, there are cer- tain federal supports for specific groups within Canada. First Nations peoples living on-reserve may receive some direct healthcare services through federal programs funded by Indigenous Services Canada (ISC). These programs, like the National Native Alcohol and Drug Abuse Program (NNADAP), provide certain healthcare services directly to First Nations reserves [37]. Additionally, most status First Nations and recognized Inuit may qualify for the NIHB program. This program provides additional coverage for certain services not covered under Canada’s Medicare such as dental and vision care, mental health services, medical transport, drug and pharmacy products, and medical equip- ment [38]. Although these programs aim to address issues related to healthcare use and access, they are situated in an imperfect system. Issues of institutional, anti-indigenous racism are apparent in Canada [39] and may contribute further to potential inequities in health care. Additionally, it is important to note that this program does not cover all Iindigenous populations as non-status First Nations, some Inuit populations and Métis are excluded. For those indi- viduals who are not covered by the NIHB, primary care

2769Journal of Racial and Ethnic Health Disparities (2024) 11:2766–2785

1 3

services are covered as a part of the provincial/territorial Medicare programs and therefore do not require OOP pay- ments. For this reason, income should not act as a barrier to access in principle, although there are ways in which it may indirectly have an impact. Importantly, there may still be other potential barriers for Indigenous populations. Specialist care, which includes any other type of physi- cian visit (surgeon, oncologist, psychiatrist, dermatolo- gist, etc.), may or may not be covered without a private insurance plan [40]. This distinction comes primarily from location of healthcare service (hospital or profit-making clinics) and means that income/location may act as a bar- rier to service utilization for this population. As well, even those who qualify for the NIHB may experience lack of access to timely service due to discriminatory billing prac- tices on the part of healthcare service providers [41]. This study aims to assess income-related inequities in primary care and specialist care within status and non-status First Nations adults living off-reserve in Canada.

Methods

Data & Study Design

This research was performed as a secondary analysis of the 2017 Aboriginal Peoples Survey (APS). As a cross-sec- tional survey data, the APS collects general information on topics related to health, income, education, housing and employment from Indigenous peoples aged 15 years and older. The target population for the APS is those who report “aboriginal identity” or “aboriginal ancestry,” live in private dwellings and are not living on a First Nations’ reserve. The sampling frame for the APS comes from those who have responded “yes” to belonging to an Abo- riginal population, or to having Aboriginal ancestry on the 2016 Census of Population. The 2017 APS had a 76.0% response rate, with the total of 24,220 Indigenous respond- ents. Statistics Canada used a stepwise method to calculate weights of the participants, where initial weights for each respondent were calculated based on weights in the Census and are then adjusted for non/partial response. The weights were subsequently readjusted using post-stratification to fit known totals for the population and the sigma-gap method was employed to reduce excessively large weights in cer- tain stratum [42]. The analytical sample for this research was single-identity First Nations adults aged 19 and older living off-reserve in Canada. There were a total of 8298 respondents a with a weighted representation of 430,741 First Nations individuals.

Variables

Assessing income-related inequities in healthcare utiliza- tion for a population requires four general types of variables [3, 43], which were collected in the 2017 APS: healthcare utilization, income, and variables indicating need for health- care, as well as variables that are known to be associated with healthcare utilization beyond need for healthcare, often termed as non-need factors.

Healthcare Utilization The utilization variable represents whether a certain type of healthcare was used, or how much was used, in a given time frame. In this project the vari- able was a binary measure (have or have not used) over the past 12 months and was categorized based on the type of healthcare: primary care and specialist care. Since there is evidence to suggest that nurse practitioners (NPs), especially in rural areas, may act as effective primary care reaching a wider range of those that may need it [44], primary care was defined as the use of a family doctor/GP or NP. The APS reports utilization of different type of healthcare as a binary variable (have or have not used) over the past 12 months. Specifically, participants were asked if they have “seen or talked to any of the following health professionals about their physical, emotional or mental health, in the past 12 months?”

Income Income, consumption, and wealth are the main indicators of individual economic resources. However, none of these alone offers an accurate, universal measure due to potential errors in resource evaluation and inherent measure- ment noise. Income is the most frequently used indicator of economic resources in the studies on inequities in healthcare utilization. Considering the availability of information in the APS, total personal annual income for the past calendar year was used in the study. In the APS, the value of personal income includes any source, such as employment or self- employment, government income, pensions and annuities and other sources.

Need Need is the primary distinction between inequity and inequality in a population. Need is a legitimate reason that one should access care, and need for healthcare is gener- ally measured by two types of variables, demographic and health status. The demographic variables used were age and sex, which have shown to be linked to the health of Indig- enous populations [45]. The APS reports age as a categorical variable and sex as a dichotomous variable (male/female). Health status variables used for this research include self- rated health status (SRH) and number of chronic diseases. SRH is a well validated measure of health status and has been shown to be predictive of other health variables includ- ing mortality [46–48].

2770 Journal of Racial and Ethnic Health Disparities (2024) 11:2766–2785

1 3

Non‑need A non-need variable is any variable that is known to be associated with healthcare utilization but does not reflect a need for healthcare. To assess income-related ineq- uities in healthcare utilization, based on the relevant existing studies [49, 50], the following non-need variables other than income were included in the analysis: education, employ- ment status, marital status, household size, rurality, and First Nations status (in non-stratified analyses). Each of these non- need variables was collected in a categorical fashion (more information in the results section). The non-need variables were used in this research only as a correction for factors that may influence the individual’s propensity to seek medical care [50, 51]. These non-need variables (if not included) may confound the relationship by influencing both the income and utilization variables. Each of these non-need variables may be associated with health inequalities in Indigenous popula- tions in Canada based on literature related to deficits when compared to a general Canadian population [28]. Table 1 reports detailed definition of variables used in the study.

Empirical Approach

First, logistic regression models were used to demonstrate the existence of inequities in primary and specialist use among the entire population and subsequently stratified by First Nations status. Then, the concentration index (C) based horizontal inequity index (HI) approach was used to quan- tify and decompose income-related inequity in healthcare utilization.

Regression Analysis Multivariable logistic regression model was used to identify factors associated with healthcare uti- lization among First Nations peoples living off-reserve in Canada. Each of the regression models were run with the entire analytical sample and subsequently stratified by First Nations status.

Measuring and Explaining Horizontal Inequity The HI approach was used to measure income-related inequitites

in healthcare utilization using the indirect standardization method. The HI is defined as the C of actual healthcare use minus the C of predicted healthcare use. The C for actual healthcare use (predicted, i.e., expected to use based on healthcare need) is based on the Concentration curve (CC) for actual (predicted) healthcare. The CC was created by plotting cumulative use of actual (predicted) healthcare used against the population ranked by income. On the CC plot, a 45-degree diagonal represent perfect equality in actual (pre- dicted) healthcare utilization. The C for actual (predicted) healthcare use can be derived as twice the area between the CC and the line of perfect equality. The C can be estimated using the “convenient regression” methods as follows:

where the �2

R represents the variance of the fractional rank,

and Ri is the fractional rank of the individual within the population (i = 1 and n for the lowest income and highest income individuals, respectively), calculated as Ri = i/n. An ordinary least square estimate of δ is direct approximation to the C. The C value ranges from −1 to 1, where positive values indicate higher income individuals using more health- care and vice versa for negative values.

Suppose that we have a non-linear logit model of the rela- tionship between healthcare use variable need and non-need control variables in term of general functional form, G:

where Yi is the utilization of healthcare for individual i, the xs represent the non-need variables (income, education, employment status, household size, marital status, rural- ity, and indigenous status) and the zs represent each vari- able used to assess need for healthcare (sex, age, SRH and chronic conditions). α, β, and γ represent regression param- eters, and εi represents an error term. The need-predicted use for indivdual i,Ŷi,can be calculated as follows:

(1)2�2

R

(

YIS i

Y

)

= � + �Ri + �i,

(2)Yi = G (

� + ∑

j �jxji +

∑

k �kzki

)

+ �i,

(3)Ŷi = G (

�̂� + ∑

j 𝛽jxjp +

∑

k rkzki

)

+ 1

n

n ∑

i=1

G (

�̂� + ∑

j 𝛽jxjp +

∑

k rkzki

)

,

where Ŷi is the healthcare amount an individual would receive if they were treated the same as others with the similar level of healthcare need. xjp is the population aver- ages. In order to allow for measurement of the ideal level of healthcare use an individual requires, all non-need indica- tors are set equal to average. Using Ŷi and Yi, the indirect

need-standardized healthcare use for each individual ( Y IS

i )

can be estimated as follows:

where Y represents the population averages for healthcare utilization.

(4)Y IS

i = Y − Ŷi + Y ,

2771Journal of Racial and Ethnic Health Disparities (2024) 11:2766–2785

1 3

Table 1 Definition of variables used in the study

Variables Description

Outcome variables Primary care =1 if the respondent uses either a GP or nurse in the past 12 months, 0 otherwise Specialist care =1 if the respondent use a specialist (any other medical doctor or specialist such as a sur-

geon, allergist, orthopedist, [urologist/gynecologist] or psychiatrist) services in the past 12 months, 0 otherwise

Need variables Sex (male: ref.) Female =1 if the respondent is female, 0 otherwise Age (19–24: ref.) 25–34 =1 if the respondent’s age falls within the range of 25 to 34, 0 otherwise 35–44 =1 if the respondent’s age falls within the range of 35 to 44, 0 otherwise 45–54 =1 if the respondent’s age falls within the range of 45 to 54, 0 otherwise 55+ =1 if the respondent’s age is 55+, and 0 otherwise Number of chronic conditions (0 conditions: ref.) 1 condition =1 if the respondent has one chronic condition, 0 otherwise 2 conditions =1 if the respondent has two chronic conditions, 0 otherwise ≥ 3 conditions =1 if the respondent has ≥ 3 chronic conditions, 0 otherwise Self-rated health (excellent: ref.) Very good =1 if the respondent self-reports their health status as “very good,” 0 otherwise Good =1 if the respondent self-reports their health status as “good,” 0 otherwise Fair =1 if the respondent self-reports their health status as “fair,” 0 otherwise Poor =1 if the respondent self-reports their health status as “poor,” and 0 otherwise Non-need variables Individual income (<$5,000: ref.) $,000 – $9999 =1 if the respondent's personal annual individual income (derived from sources such as

employment or self-employment, government income, pensions and annuities, and other sources [e.g., child support, spousal support, scholarships, etc.]) falls within the range of $5000 to $9999, 0 otherwise

$10,000 – $19,999 =1 if the respondent's individual income falls within the range of $10,000 to $19,999, 0 otherwise

$20,000 – $29,999 =1 if the respondent's individual income falls within the range of $20,000 to $29,999, 0 otherwise

$30,000 – $39,999 =1 if the respondent's individual income falls within the range of $30,000 to $39,999, 0 otherwise

$40,000 – $49,999 =1 if the respondent's individual income falls within the range of $40,000 to $49,999, 0 otherwise

$50,000–$69,999 =1 if the respondent's individual income falls within the range of $50,000 to $69,999, 0 otherwise

≥$70,000 =1 if the respondent's individual income is ≥$70,000, and 0 otherwise Education level (grade 8 and less: ref.) Some secondary =1 if the respondent has obtained some secondary education (meaning s/he has completed

some schooling after grade 8 but have not graduated high school), 0 otherwise Secondary diploma/Equivalent = if the respondent has obtained a Secondary diploma/Equivalent (including academic or

vocational high school diplomas, certificates from graduating secondary school, passing a high school equivalency test like the General Educational Development test, or obtaining an Adult Basic Education certificate), and 0 otherwise.

Post-secondary diploma (less than bachelors) = 1 if the respondent has obtained a post-secondary diploma (less than bachelor’s degree), 0 otherwise

Bachelor’s degree or above = 1 if the respondent has obtained bachelor’s degree or above, 0 otherwise Employment status (employed: ref.) Unemployed = 1 if the respondent is unemployed in the reference week, 0 otherwise Not in the labour force = 1 if the respondent is not in the labour force in the reference week, 0 otherwise

2772 Journal of Racial and Ethnic Health Disparities (2024) 11:2766–2785

1 3

The HI can be estimated by measuring the C value for Y IS

i

using the “convenient regression” method. The HI ranges from −1 to 1. A positive (negative) value for the HI implies higher (lower) income individuals used more healthcare than predicted amount after adjusting for their health need. The larger magnitude of the index implies higher levels of income-related inequity [52, 53]. Given that both primary care and specialist care, the outcome variables in our study, are binary variables, we applied the Wagstaff [54] and Errey- gers [55] correction (WC and EC, respectively) methods to adjust the C for actual use and HI values. The WC and EC are calculated by multiplying the final value of both indexes by 1∕

(

1 − Y )

and 4 × Y , respectively. We decomposed the HI to quantify and compare the

extent to which observed non-need variables contributed to inequities in healthcare use among First Nations peoples living off-reserve in Canada. If we have a linear regression model linking healthcare use, Y, to a set of need and non- need explanatory factors, Wagstaff and colleagues [56] demonstrated that the C of Y can be decomposed into the contribution of factors that explain inequality in healthcare use as follows:

Based on Eq. 5, the C value for healthcare use equal to a weighted sum of the C of the explanatory variables, Cj, and

(5)C = ∑

j

(

�jxj

Y

)

Cj + ∑

k

(

�kzk

Y

)

Ck + GC�

Y .

Ck. The weights assigned to the C values of need and non- need explanatory variables are determined by the elasticities of the outcome variable Y with respect to these variables, denoted as ( �kzk

Y ) and

(

�jxj

Y

)

. Here, xj and xk represent the population averages for the relevant variables and βs are the average of the marginal effects (AMEs) for non-need and need variables obtained from the logistic regression model. The GCε is the generalized C for the error term defined as GC� =

2

n

∑n

i=1 �iRi . Determinants that possess a significant

elasticity and exhibit an unequal distribution by income will contribute to income-related inequality in healthcare utiliza- tion. A positive (negative) contribution of a certain explana- tory variable to the C of healthcare use indicates that the distribution of the variable across income groups and its relationship with healthcare utilization result in higher (lower) healthcare use among the high (low) income group [56].

In Eq. 5, the first part represents the contribution of other non-need variables, the second part represents the contribu- tion of need variables, and the third part represents the contri- bution of the residual term, which represents the unexplained component [53]. The second component in Eq. 5 represents the justifiable or legitimate aspect of inequality in healthcare utilization, as it aligns with the individuals’ actual needs [57]. The first part of Eq. 5 addresses the unjustifiable or avoidable portion of inequality, which arises from non-need factors such as income, education, employment status, and other relevant variables [58]. By subtracting the contributions of need factors

Ref. denotes reference category in regression analysis

Table 1 (continued)

Variables Description

Household size (1 person: ref.) 2 people = 1 if the respondent’s household size is 2, 0 otherwise 3 people = 1 if the respondent’s household size is 3, 0 otherwise 4 people = 1 if the respondent’s household size is 4, 0 otherwise ≥5 people = 1 if the respondent’s household size is ≥5, 0 otherwise Marital status (Separated/divorced/widowed: ref.) Never married 1= if the respondent never married, 0 otherwise Married 1= if the respondent's legal marital status is married, 0 otherwise Rurality (Census metropolitan area [at least

100,000 with 50,000 or more in the core]: ref.) Other population center 1= if the respondent lives in an area with a population of at least 1,000 and a density of 400

or more people per square kilometer, 0 otherwise, Other – rural =1 the respondent lives in all areas outside population centers, which are collectively defined

as rural area, 0 otherwise Indigenous status (non-status: ref.) Status =1 if the respondent is a “Status Indian,” which includes both Registered Indians (those

registered under the Indian Act of Canada) and Treaty Indians (those belonging to a First Nation or Indian band that signed a treaty with the Crown), 0 otherwise

2773Journal of Racial and Ethnic Health Disparities (2024) 11:2766–2785

1 3

from the C value for actual healthcare utilization, we can obtain the values for HI, as follows:1

In the measurement and decomposition of the HI using Eqs. 5 and 6, we employed the WC [54] and EC [55] meth- ods to appropriately handle the binary nature of the outcome variables. All the statistical models in the study are esti- mated using survey weights. Robust standard errors were employed in all regression calculations to account for the potentially inaccurate variance estimations. A p-value < 0.05 was used to determine statistical significance.

Results

Descriptive Results

Table 2 shows descriptive statistics results for the entire analytical sample as well as for stratified samples by First Nations status. The proportion of those who have used primary care (GP or NP) and specialist care in the past 12 months was 70.6% and 35.1%, respectively, for the total population. Additionally, there were statistically sig- nificant differences between the status groups, with non- status First Nations using more than status First Nations (p-value<0.001). The majority of sample (53.8%) are younger than 44 years old, with the status First Nation population tending to be slightly younger than the non-sta- tus population. A total of 7.3% of the population reported perceiving their health as poor. There was no statistically significant difference in the SRH (p-value=0.609) between status and non-status First Nations groups. A majority of the sample (59%) reported having at least one chronic condition. There was a statistically significant difference in the number of chronic conditions with status populations reporting less chronic conditions (p-value=0.001). Furthermore, 5.2% of the population reported an annual individual income of more than $70,000, with non-status First Nations peoples report- ing slightly higher income than their status counterparts. In terms of education, 11.8% of the total population obtained bachelor and above degree (non-status group: 14.5% and status group: 10.3%). Approximately 8% of the total popula- tion had an unemployed status, with 7.2% for the non-status group and 8.6% for the status population. A total 53.3%

(6)HI = C − ∑

k

(

�kzk

Y

)

Ck.

of respondents resided in larger census metropolitan areas (non-status group: 59.1% and status group: 50.2%).

Regression Results

Tables 3 and 4 report univariate and multivariable logis- tic regression results for primary and specialist care for the whole population as well as for stratified samples by First Nations status. Across both utilization types and popula- tions, females have greater odds of service utilization in all adjusted models, when compared to males. The health variables tend to show a strong association with healthcare use. Number of chronic conditions appears to be associ- ated with use quite well with all those who have more than zero chronic conditions having a statistically significantly increased odds of using primary care. Similarly, SRH appears to be associated with healthcare among Indigenous groups. The income variable shows a relatively weak asso- ciation with use of healthcare in the non-status popula- tion. In the status group, a statistically significantly higher specialist care use was seen in high-income individuals (≥$70,000) compared with low-income adults (adjusted odds ratio [aOR], 1.705). Higher education was statistically significantly associated with primary and specialist care use among First Nations peoples. Rurality appears to have a rela- tionship with specialist care use, where those living in more rural areas tend to use less care. Indigenous status shows a strong association with healthcare use; status as compared to non-status First Nations peoples use fewer primary (aOR, 0.823) and specialist care (aOR, 0.816) services.

Horizontal Inequities in Healthcare Use and their Determinants

The CCs for actual and need-standardized primary care and specialist use are depicted in Figs. 1 and 2, respectively. The CCs for actual use suggest an equal distribution of pri- mary and specialist care by income among the entire sam- ple. Additionally, when stratified by First Nations status, the result shows an equal distribution of these types of utiliza- tion. By looking at the need-standardized use CCs we can see that the curves are below the line of perfect equality in each case, indicating pro-rich inequities in primary and specialist care use among First Nations peoples.

Tables 5 and 6 report the magnitude of the C and HI for primary and speciliast care. As reported in the table, the C for actual healthcare use show no statistical significance in any population indicating actual use of primary and special- ist care are relatively equally spread across income groups. The statistically significantly positive values of the HI sug- gested pro-rich inequities in both healthcare services.

Tables 5 and 6 report the results of the decomposition analysis of the HI for primary and specialist care use among

1 The value of the HI  can be obtained by measuring the C value for Y��

i using the “convenient regression” method or the

formula presented in Eq. 6.

2774 Journal of Racial and Ethnic Health Disparities (2024) 11:2766–2785

1 3

Table 2 Descriptive statistics for individuals included in the analysis

Total Non-status Status *Pearson chi2 value

Variables Proportion Proportion Proportion P-value Outcome variables Use of primary care in the past 12 months 0.706 0.743 0.686 <0.001 Use of specialist care in the past 12 months 0.351 0.395 0.327 <0.001 Need variables Sociodemographic variables Sex Female 0.552 0.537 0.560 0.003 Age (years) 19–24 0.147 0.151 0.144 25–34 0.210 0.200 0.215 35-44 0.181 0.175 0.185 45–54 0.215 0.209 0.218 55+ 0.247 0.264 0.238 0.058 Health Variables Number of chronic conditions 0 conditions 0.409 0.373 0.428 1 condition 0.235 0.233 0.236 2 conditions 0.171 0.184 0.165 ≥3 conditions 0.185 0.210 0.172 0.001 Self-rated health Excellent 0.165 0.162 0.167 Very good 0.284 0.293 0.28 Good 0.321 0.306 0.329 Fair 0.155 0.157 0.154 Poor 0.073 0.081 0.069 0.609 Non-need variables Socioeconomic variables Individual income <$5000 0.078 0.059 0.088 $5000 – $9999 0.068 0.068 0.067 $10,000 – $19,999 0.229 0.226 0.231 $20,000 – $29,999 0.148 0.144 0.150 $30,000 – $39,999 0.115 0.115 0.116 $40,000 – $49,999 0.088 0.089 0.087 $50,000–$69,999 0.122 0.135 0.116 ≥$70,000 0.152 0.164 0.145 0.001 Education level Grade 8 and less 0.051 0.041 0.056 Some secondary 0.132 0.112 0.143 Secondary diploma/equivalent 0.348 0.328 0.359 Post-secondary diploma (less than bachelors) 0.351 0.374 0.339 Bachelor’s degree or above 0.118 0.145 0.103 <0.001 Employment status Employed 0.589 0.606 0.579 Unemployed 0.081 0.072 0.086 Not in the labor force 0.330 0.322 0.335 <0.001 Other non-need variables Household size

2775Journal of Racial and Ethnic Health Disparities (2024) 11:2766–2785

1 3

First Nation peoples. In cases where the distribution of healthcare use related to income is fair and justifiable, fac- tors unrelated to need would not contribute to inequities. As presented in Tables 5 and 6 and depicted in Fig. 3, income and education were the two primary drivers of pro-rich ineq- uities in both primary care and specialist care utilization. Income accounted for approximately 66–75% and 56–76% of the inequities observed in primary care and specialist care utilization among First Nation peoples, respectively. If income inequality were eliminated or if income had no impact on primary care and specialist care use, the HI value would decrease by the percentage contributions attributable to income. Our findings also indicate that education among First Nation peoples explained approximately 55–61% and 31–34% of the overall contribution towards these inequities.

Discussion

The primary objective of this study was to assess inequities in primary and specialist care among First Nations adults living off-reserve in Canada. This objective was met through the description of these populations, identification of fac- tors associated with primary and specialist care use using multivariable regression and then measuring and decompos- ing inequities through the HI. This study was an important evaluation of inequity within First Nations adults living off- reserve in Canada, where all the existing studies focus on comparing to the general Canadian population.

Our descriptive results showed that primary care was used at least once within 12 months by 70.6% of the total sample and specialist care was used by 35.1% in the same

time frame. There was a statistically significant difference in healthcare utilization between status groups, with non-status individual use statistically significantly more care.

In assessing the associations with utilization for both primary and specialist care, across all univariate and mul- tivariable models, the odds of utilization for females were statistically significantly higher than males. This result is supported by previous research in other populations [59] and is logical based on physiological between the sexes and gendered health seeking behaviors. Rurality is found to be negatively associated with specialist care use among First Nations adults living off-reserve. Specifically, those living in more rural areas tend to use less specialist care. This could be partially attributed to the limited accessibility of these services in rural regions compared to densely populated cen- sus metropolitan areas. The regression analyses also sug- gested a strong positive association between SES variables, such as education level, and healthcare use.

The HI suggested pro-rich inequities in primary and spe- cilist care among First Nations adults living off-reserve. This result is similar in the general population of Canada and else- where, where use of GP services and specialist services tend to be pro-rich [3, 18, 60–63]. A study in Australia showed a pro-poor distribution of GP services among Indigenous peoples; however, the magnitude is lesser, and the results were statistically significant in very few subpopulations [64]. It is important to note that these are not perfect comparisons primarily because of the type of use (GP versus “primary care”) and also because of varying methods for the HI index calculation in these studies. The HIs for each group and in total suggest pro-rich income-related inequity in specialist care use among these populations. This is similar to what is

*The Pearson chi2 value shows differences between status and non-status populations on each of the vari- ables in question

Table 2 (continued) Total Non-status Status *Pearson chi2 value

1 person 0.152 0.157 0.149

2 people 0.302 0.323 0.290

3 people 0.212 0.206 0.216

4 people 0.182 0.184 0.180 ≥5 people 0.152 0.129 0.165 <0.001 Marital status (legal) Never married 0.488 0.436 0.516 Married 0.325 0.357 0.308 Separated/divorced/widowed 0.187 0.207 0.176 0.001 Rurality Census metropolitan area 0.533 0.591 0.502 Other population centre 0.275 0.211 0.310 Other – rural 0.192 0.198 0.188 <0.001 Sample size (n) 8298 2891 5407

2776 Journal of Racial and Ethnic Health Disparities (2024) 11:2766–2785

1 3

Table 3 Univariate and multivariable logistic regression models (stratified and total) for use of primary care

Non-status Status Total

Univariate Multivariable Univariate Multivariable Univariate Multivariable

Variables OR aOR OR aOR OR aOR Sex (male: ref.) Female 1.875*** 1.749*** 1.775*** 1.524*** 1.791*** 1.605*** Age (19–24: ref.) 25–34 1.260 0.969 0.936 0.847 1.028 0.888 35–44 1.893** 1.350 1.308 0.949 1.467** 1.068 45–54 2.074*** 1.465 1.674*** 1.005 1.791*** 1.147 55+ 1.979*** 1.213 1.642*** 0.805 1.757*** 0.949 Number of chronic conditions (0 conditions: ref.) 1 condition 2.085*** 1.932*** 1.883*** 1.678*** 1.955*** 1.740*** 2 conditions 2.569*** 2.135*** 2.759*** 2.325*** 2.721*** 2.274*** ≥ 3conditions 3.354*** 2.300*** 4.293*** 3.090*** 3.970*** 2.713*** Self-rated health (excellent: ref.) Very good 1.501* 1.364 1.553** 1.343* 1.540*** 1.341** Good 1.788*** 1.579* 1.724*** 1.381* 1.735*** 1.426** Fair 2.888*** 2.181** 3.203*** 2.122*** 3.094*** 2.130*** Poor 4.129*** 2.457* 3.796*** 2.011** 3.936*** 2.168*** Individual income (<$5,000: ref.) $5000 – $9999 1.532 1.425 1.154 1.074 1.278 1.159 $10,000 – $19,999 1.347 1.278 1.141 1.011 1.215 1.067 $20,000 – $29,999 1.245 1.497 0.982 0.973 1.069 1.106 $30,000 – $39,999 0.975 1.277 1.035 1.127 1.022 1.145 $40,000 – $49,999 1.083 1.313 1.056 1.229 1.075 1.244 $50,000–$69,999 1.345 1.776 0.902 1.163 1.054 1.314 ≥$70,000 1.38 1.763 1.065 1.368 1.182 1.454* Education level (grade 8 and less: ref.) Some secondary 1.170 1.466 0.511** 0.607* 0.658* 0.805 Secondary diploma/equivalent 1.577 2.437* 0.599* 0.871 0.812 1.209 Post-secondary diploma (less than

bachelors) 1.909* 2.711** 0.908 1.219 1.154 1.564*

Bachelor’s degree or above 2.626** 3.946*** 0.933 1.323 1.342 1.904** Employment status (unemployed: ref.) Employed 1.446 1.368 1.149 1.002 1.248 1.085 Not in the labour force 2.230*** 2.084** 1.586* 1.185 1.771*** 1.387* Household size (1 person: ref.) 2 people 0.789 0.734 0.933 0.982 0.887 0.889 3 people 0.633* 0.625 0.884 0.972 0.789 0.837 4 people 0.66 0.693 0.844 0.993 0.778 0.895 ≥5 people 0.689 0.73 0.764 0.887 0.724* 0.831 Marital status (separated/divorced/widowed: ref.) Never married 0.802 1.420 0.572*** 0.844 0.637*** 1.008 Married 1.216 1.581* 0.770* 0.945 0.912 1.121 Rurality (census metropolitan area: ref.) Other population centre 1.008 0.921 0.911 0.949 0.910 0.957 Other – rural 1.008 1.017 0.863 0.88 0.905 0.928 Indigenous status (non-status: ref.) Status – – – – 0.754*** 0.823* Constant – 0.155** – 0.97 – 0.623

2777Journal of Racial and Ethnic Health Disparities (2024) 11:2766–2785

1 3

observed among Indigenous populations elsewhere, where specialist services tend to be more pro-rich [64].

Similar to previous international and Canadian studies [17, 50, 62, 63, 65], our decomposition analysis confirmed that unequal income distribution and its significant associa- tion with healthcare use were the main drivers of pro-rich inequities in both primary and specialist care use among First Nations peoples living off-reserve. Education also played a significant role in the observed inequities, as it posi- tively correlated with both primary and specialist care use among First Nations peoples, and high education attainment was more concentrated among high income adults. This find- ing is consistent with prior research conducted in Canadian [17] and European countries [66, 67] and is also supported by a study focusing on the Australian Indigenous peoples [68]. Higher education levels facilitated better interaction with healthcare providers and improved health literacy, resulting in higher healthcare use for high-income Cana- dians [69].

The results of this research and recent relevant studies [17, 18] indicate a different pattern of income-related ineq- uities in healthcare utilization in Canada as compared to several other high income countries. Most countries with universal health systems demonstrates pro-rich inequity in specialist care use whereas GP care use was either pro-poor or equitable [63, 70–74]. In contrast, this research and pre- vious studies in the general population of Canada [17, 18] revealed pro-rich inequity in GP (primary care) and special- ist care. The pro-rich inequities in GP and primary care use are concerning because these services include a substantial amount of health promotion and preventative services and GP and primary care are the entry point into the Canada’s healthcare system [75, 76].

Since the Canadian health system provides free primary and specialist care for all Canadians, including First Nations peoples living, pro-rich inequities in these services are not related to the direct healthcare costs associated with the use of these services. Healthcare use barriers within and outside of health system may have contributed to the lower healthcare use among the low-income First Nations adults living off-reserve in Canada. Research has demonstrated that Indigenous peoples often have very difficult and different experiences in seeking care, when compared to their non- Indigenous counterparts [77, 78]. These experiences, which have involved racism and improper treatment, have resulted

in limited use of many types of care and, in some cases, have led individuals to avoiding care altogether [79, 80]. A 2011 qualitative study of individuals’ experiences in accessing emergency room care showed that Indigenous participants felt that they may have decreased chances of receiving help if identified as Indigenous or poor [77], evidence that institu- tional racism is present in the delivery of healthcare and that there is still much work to be done in this area. Low-income, compared to the high-income, Indigenous peoples may face disproportionately high levels of improper treatments in Canada. This, in turn leads to lower utilization of healthcare among the poor Indigenous peoples living in Canada. Barri- ers outside healthcare system such as costs associated with transportation to use healthcare [81]) may have also con- tributed to the observed pro-rich inequities in primary and specialist care use. In this case, policymakers may consider covering indirect costs such as travel cost associated with healthcare use among First Nations adults in Canada. Other indirect costs associated with the use of healthcare such as time away from work and childcare [81] may also contribute to decreased utilization of health services among the poor. The pro-rich inequities in the utilization of both primary and specialist care among First Nations people living in Canada are likely the result of a combination of all these factors, thus further studies are required to provide further insight on this issue in Canada.

Our results have important policy implications for First Nations peoples. In Canada, where many policies are gener- ated around First Nations status, our stratified analysis based on status group may provide a useful piece of evidence for decision makers. Our analyses show important differences according to government defined “Indian status,” where status First Nations peoples tend to use lower primary and specialist healthcare compared to non-status. Other results such as the pro-rich inequity in healthcare utilization could provide evidence for programs providing healthcare directed towards specific Indigenous populations who may require more care.

This research is subject to some limitations. First, this study is a descriptive and secondary analysis of cross-sec- tional data, which means that no causal inferences should be drawn from the results. Second, the APS stems from a sampling method which is based on the Canadian Cen- sus of Population. There are multiple reasons for which underrepresentation may be present. The first is that the

*p-value <0.05, **p-value  <0.01, ***p-value <0.001

Table 3 (continued)

Non-status Status Total

Univariate Multivariable Univariate Multivariable Univariate Multivariable

Sample size (n) 2659 4810 7469

2778 Journal of Racial and Ethnic Health Disparities (2024) 11:2766–2785

1 3

Table 4 Univariate and multivariable logistic regression models (stratified and total) for use of specialist care

Non-status Status Total

Univariate Multivariable Univariate Multivariable Univariate Multivariable

Variables OR aOR OR aOR OR aOR Sex (male: ref.) Female 1.505*** 1.400** 1.408*** 1.309** 1.432*** 1.330*** Age (19–24: ref.) 25–34 1.346 1.056 0.973 0.903 1.09 0.943 35–44 1.580** 1.108 1.365* 1.017 1.435** 1.051 45–54 1.691** 1.045 1.169 0.777 1.333** 0.882 55+ 1.877*** 1.053 1.283* 0.844 1.487*** 0.934 Number of chronic conditions (0 conditions: ref.) 1 condition 1.648*** 1.415* 1.812*** 1.759*** 1.761*** 1.610*** 2 conditions 2.029*** 1.597** 2.518*** 2.447*** 2.350*** 2.043*** ≥ 3 conditions 2.987*** 1.871** 3.177*** 2.749*** 3.160*** 2.325*** Self-rated health (excellent: ref.) Very good 1.231 1.124 1.340* 1.386* 1.302* 1.174 Good 1.866*** 1.683** 1.669*** 2.176*** 1.730*** 1.501*** Fair 3.232*** 2.693*** 2.785*** 2.525*** 2.935*** 2.361*** Poor 3.408*** 2.464** 3.350*** 1.386* 3.396*** 2.470*** Individual income (<$5000: ref.) $5000 – $9999 1.216 0.998 1.270 1.157 1.28 1.155 $10,000 – $19,999 1.140 0.982 1.016 0.894 1.085 0.941 $20,000 – $29,999 1.139 1.202 0.915 0.896 1.015 1.027 $30,000 – $39,999 0.773 0.859 1.178 1.320 1.039 1.175 $40,000 – $49,999 1.269 1.453 0.887 1.058 1.047 1.232 $50,000-$69,999 1.174 1.437 1.121 1.438 1.182 1.469* ≥$70,000 0.990 1.192 1.280 1.705** 1.196 1.497* Education level (grade 8 and less: ref.) Some secondary 0.828 0.900 0.782 0.812 0.802 0.868 Secondary diploma/equivalent 0.871 1.125 0.959 1.207 0.945 1.213 Post-secondary diploma (less than

bachelors) 0.971 1.152 1.213 1.466 1.152 1.375

Bachelor’s degree or above 1.249 1.698 1.538* 1.883* 1.485* 1.843** Employment status (unemployed: ref.) Employed 1.574* 1.547 1.307 1.204 1.405* 1.311 Not in the labour force 2.254*** 1.892* 1.525* 1.310 1.755*** 1.486** Household size (1 person: ref.) 2 people 0.813 0.813 0.853 0.923 0.843 0.879 3 people 0.813 0.893 0.903 1.005 0.861 0.956 4 people 0.763 0.923 1.025 1.228 0.915 1.095 ≥5 people 0.591* 0.630 1.013 1.224 0.832 0.978 Marital status (separated/divorced/widowed: ref.) Never married 0.668** 0.909 0.801 1.062 0.735** 0.987 Married 0.905 1.119 0.894 0.950 0.898 1.001 Rurality (census metropolitan area: ref.) Other population centre 0.925 0.872 0.741** 0.770* 0.769** 0.800* Other – rural 0.699* 0.693* 0.766* 0.814 0.736*** 0.762** Indigenous status (non-status: ref.) Status – – – – 0.747*** 0.816** Constant – 0.150** – 0.120*** – 0.145***

2779Journal of Racial and Ethnic Health Disparities (2024) 11:2766–2785

1 3

Indigenous populations may have lower response rate to the census [82] limiting the sampling frame for the APS.

Secondly, the census collects information on households in Canada and therefore does not account for homeless

*p-value <0.05, **p-value <0.01, ***p-value <0.001

Table 4 (continued)

Non-status Status Total

Univariate Multivariable Univariate Multivariable Univariate Multivariable

Sample size (n) 2657 4807 7464 0

.2 .4

.6 .8

1

C u

m u

la ti

v e

sh ar

e o

f p

ri m

ar y

ca re

v is

it

0 .2 .4 .6 .8 1

Cumulative share of individuals (poorest to richest)

Actual use

Need standardised use

Line of equality

A: Non-status First Nations

0 .2

.4 .6

.8 1

C u

m u

la ti

v e

sh ar

e o

f p

ri m

ar y

ca re

v is

it

0 .2 .4 .6 .8 1

Cumulative share of individuals (poorest to richest)

Actual use

Need standardised use

Line of equality

B: Status First Nations

0 .2

.4 .6

.8 1

C u

m u la

ti v

e sh

ar e

o f

p ri

m ar

y ca

re v

is it

0 .2 .4 .6 .8 1

Cumulative share of individuals (poorest to richest)

Actual use

Need standardised use

line of equality

C: Total First Nations

Fig. 1 Concentration curves for primary care use for First Nations populations (non-status, status, and unstratified)

2780 Journal of Racial and Ethnic Health Disparities (2024) 11:2766–2785

1 3

populations [82]. Indigenous peoples who are homeless will not have been included in the APS for that reason. It is important to consider these limitations as they may have an impact on our measurement of inequity, especially if those First Nations peoples with very low income are being excluded. Third, several of the variables used for the analyses are self-reported and, in some cases, are collected retrospectively. This means that the variables may be sub- ject to certain measurement biases, specifically reporting and recall bias, among others. There is no reason to believe

that bias would be dependent on any individual value used in this study and was therefore deemed uniform throughout the population. Fourth, the income variable used in this study represents “personal sources” of income, making it an assessment of individual income rather than house- hold income. Generally, household income is more desir- able for inequity studies as it better represents the total income available to each individual in the home. Fifth, this paper investigates inequities in healthcare use among First Nations individuals living off-reserves in Canada.

0 .2

.4 .6

.8 1

C u

m u

la ti

v e

sh ar

e sp

ec ia

li st

ca re

v is

it

0 .2 .4 .6 .8 1

Cumulative share of individuals (poorest to richest)

Actual use

Need standardised use

Line of equality

A: Non-status First Nations

0 .2

.4 .6

.8 1

C u

m u

la ti

v e

sh ar

e o

f sp

ec ia

li st

ca re

v is

it

0 .2 .4 .6 .8 1

Cumulative share of individuals (poorest to richest)

Actual use

Need standardised use

Line of equality

B: Status First Nations

0 .2

.4 .6

.8 1

C u

m u

la ti

v e

sh ar

e o

f sp

ec ia

li st

ca re

v is

it

0 .2 .4 .6 .8 1

Cumulative share of individuals (poorest to richest)

Actual use

Need standardised use

line of equality

C: Total First Nations

Fig. 2 Concentration curves for specialist care use for First Nations populations (non-status, status, and unstratified)

2781Journal of Racial and Ethnic Health Disparities (2024) 11:2766–2785

1 3

Given the distinct healthcare accessibility experienced by Indigenous peoples living on-reserves, further analysis is required to assess inequities in healthcare use within on- reserve First Nations populations. Similarly, to conduct a thorough assessment of inequities in healthcare use among Indigenous populations in Canada, it is essential for future

research to examine inequities in healthcare among Métis and Inuit peoples. Furthermore, given the long-lasting effects of Coronavirus disease 2019 (COVID-19) on both health and healthcare systems as well as broader social and economic systems, it is crucial to evaluate how it has exacerbated existing inequities.

Table 5 The concentration and horizontal inequity indices for primary care use in each of the stratified populations and factors contributing to observed inequities among First Nations populations

The WC and EC represent the utilization of the Wagstaff and Erreygers methods, respectively, for correcting the indices; The CHI represents the contribution to the Horizontal Inequity Index (HI); the total contributions of all CHI add up to the HI values reported in each column

Total Non-status Status

WC EC WC EC WC EC

Concentration index (C for actual use) 0.0034 0.0028 0.0154 0.0117 −0.0082 −0.0071 P-value 0.820 0.557 0.646 Horizontal inequity index (HI) 0.0828 0.0688 0.0894 0.0683 0.0652 0.0562 P-value <0.001 <0.001 <0.001 Decomposition of the HI

CHI CHI CHI CHI CHI CHI Individual income $5000 – $9999 −0.0070 −0.0058 −0.0182 −0.0139 −0.0033 −0.0029 $10,000 – $19,999 −0.0065 −0.0054 −0.0264 −0.0201 −0.0011 −0.0009 $20,000 – $29,999 −0.0014 −0.0012 −0.0080 −0.0061 0.0003 0.0002 $30,000 – $39,999 0.0024 0.0020 0.0028 0.0022 0.0023 0.0020 $40,000 – $49,999 0.0064 0.0053 0.0070 0.0053 0.0063 0.0054 $50,000–$69,999 0.0172 0.0143 0.0376 0.0287 0.0094 0.0081 ≥$70,000 0.0440 0.0365 0.0708 0.0541 0.0349 0.0301 Education level Some secondary 0.0064 0.0053 −0.0096 −0.0073 0.0157 0.0135 Secondary diploma/Equivalent −0.0072 −0.0060 −0.0415 −0.0317 0.0045 0.0039 Post-secondary diploma (less than bachelors) 0.0164 0.0136 0.0284 0.0217 0.0082 0.0070 Bachelor’s degree or above 0.0300 0.0249 0.0738 0.0563 0.0117 0.0101 Employment status Unemployed 0.0017 0.0014 0.0045 0.0034 0.0000 0.0000 Not in the labor force −0.0253 −0.0210 −0.0419 −0.0320 −0.0180 −0.0155 Household size 2 people −0.0022 −0.0018 −0.0044 −0.0033 −0.0004 −0.0004 3 people 0.0012 0.0010 0.0016 0.0013 0.0003 0.0003 4 people 0.0009 0.0008 0.0037 0.0028 −0.0179 −0.0155 ≥5 people 0.0032 0.0027 0.0049 0.0037 0.0021 0.0018 Marital status (legal) Married 0.0053 0.0044 0.0062 0.0047 0.0049 0.0042 Separated/divorced/widowed −0.0001 0.0000 0.0012 0.0009 0.0022 0.0019 Rurality Other population center 0.0003 0.0002 0.0008 0.0006 0.0002 0.0001 Other – rural 0.0001 0.0001 −0.0001 −0.0001 −0.0003 −0.0002 Status 0.0026 0.0022 Residual −0.0056 −0.0047 −0.0038 −0.0029 0.0035 0.0030

2782 Journal of Racial and Ethnic Health Disparities (2024) 11:2766–2785

1 3

Conclusion

Although reducing inequities in healthcare is a primary objective of health policy in Canada, this study suggests pro-rich inequities in primary and specialist care among First Nations peoples living off-reserve in Canada. The main drivers of pro-rich inequity in specialist visits were

inequalities in income and educational attainments. These findings warrant urgent policy response towards improv- ing equity in healthcare use for this population in Canada. Both within and outside health system, discriminatory and accessibility barriers can potentially explain lower healthcare use among the low-income First Nations popu- lation living off-reserve.

Table 6 The concentration and horizontal inequity indices for specialist care use in each of the stratified populations and factors contributing to observed inequities among First Nations populations

The WC and EC represent the utilization of the Wagstaff and Erreygers methods, respectively, for correcting the indices; The CHI represents the contribution to the Horizontal Inequity Index (HI); the total contributions of all CHI add up to the HI values reported in each column

Total Non-status Status

WC EC WC EC WC EC

Concentration index (C for actual use) 0.0175 0.0160 −0.0050 −0.0048 0.0233 0.0205 P-value 0.223 0.834 0.200 Horizontal inequity index (HI) 0.1221 0.1112 0.0957 0.0915 0.1290 0.1136 P-value <0.001 <0.001 <0.001 Decomposition of the HI

CHI CHI CHI CHI CHI CHI Individual income $5000 – $9999 −0.0071 −0.0065 0.0001 0.0001 −0.0069 −0.0061 $10,000 – $19,999 0.0066 0.0060 0.0020 0.0019 0.0109 0.0096 $20,000 – $29,999 −0.0004 −0.0003 −0.0037 −0.0036 0.0012 0.0010 $30,000 – $39,999 0.0029 0.0026 −0.0018 −0.0017 0.0057 0.0050 $40,000 – $49,999 0.0063 0.0057 0.0098 0.0093 0.0018 0.0016 $50,000–$69,999 0.0255 0.0232 0.0248 0.0237 0.0233 0.0205 ≥$70,000 0.0488 0.0445 0.0224 0.0214 0.0614 0.0541 Education level Some secondary 0.0044 0.0040 0.0027 0.0026 0.0068 0.0059 Secondary diploma/equivalent −0.0075 −0.0069 −0.0057 −0.0054 −0.0064 −0.0056 Post-secondary diploma (less than bachelors) 0.0120 0.0109 0.0042 0.0040 0.0159 0.0140 Bachelor’s degree or above 0.0294 0.0268 0.0293 0.0280 0.0272 0.0239 Employment status Unemployed 0.0062 0.0057 0.0064 0.0062 0.0049 0.0043 Not in the labor force −0.0134 −0.0122 −0.0210 −0.0201 −0.0089 −0.0078 Household size 2 people −0.0025 −0.0022 −0.0030 −0.0029 −0.0016 −0.0014 3 people 0.0003 0.0003 0.0004 0.0004 0.0000 0.0000 4 people −0.0008 −0.0007 0.0008 0.0008 −0.0016 −0.0014 ≥5 people 0.0005 0.0004 0.0074 0.0070 −0.0037 −0.0032 Marital status (legal) Married 0.0006 0.0005 0.0127 0.0122 −0.0051 −0.0045 Separated/divorced/widowed 0.0001 0.0001 −0.0003 −0.0003 −0.0008 −0.0007 Rurality Other population center 0.0016 0.0015 0.0014 0.0013 0.0009 0.0008 Other – rural 0.0005 0.0004 0.0034 0.0032 −0.0004 −0.0004 Status 0.0028 0.0025 Residual 0.0054 0.0049 0.0036 0.0034 0.0044 0.0039

2783Journal of Racial and Ethnic Health Disparities (2024) 11:2766–2785

1 3

Acknowledgements We would like to thank Thunderbird Partnership Foundation for reviewing the manuscript.

Code Availability The code can be obtained from the corresponding author upon a reasonable request.

Author Contributions All authors contributed to the conception and design of the study. MH and BK performed the statistical analyses, and all authors interpreted the results. MH drafted the manuscript, and BK, YA, AB, and DM helped with drafting and revisions. All authors read and approved the final version of the manuscript.

Funding The authors acknowledge funding for this research provided by the Research Nova Scotia Establishment Grant program (grant #: 1017). MH is supported by a Tier II Canada Research Chair in Health Economics through the Canada Research Chairs (CRC) Program (grant # CRC-2020-00219; https:// www. chairs- chair es. gc. ca/ home- accue il- eng. aspx). DM is supported by a Tier II Canada Research Chair in Indigenous Peoples’ Health and Well-Being through the CRC Program (grant # CRC-2021-00436).

Data Availability We used the 2017 Aboriginal Peoples Survey (APS) Public Use Microdata File (PUMF) dataset, which is available in a pub- lic, open-access repository. The data can be accessed for a subscription fee through Statistics Canada’s the PUMF Collection.

Declarations

Ethical Approval The data used for this study is publicly available and therefore required no ethics approval. This data was processed at Statistics Canada to ensure that no information is identifiable. Addi- tionally, this project was conducted as part of a successful research grant entitled: “The dynamics of health inequalities faced by Indig- enous populations in Canada: What factors account for the inequality”, which has been granted ethics approval by Dalhousie University (REB No: 2017-4295).

Competing Interests The authors declare no competing interests.

References

1. Wagstaff A, van Doorslaer E. Equity in the finance of health care: some international comparisons. J Health Econ. 1992;11:361–87.

2. Wagstaff A, Van DE, Paq P. Equity in the finance and delivery of health care: some tentative cross-country comparisons. Oxf Rev Econ Policy. 1989;5:89–112.

3. Allin S. Does equity in healthcare use vary across Canadian prov- inces? Healthcare policy/Politiques de santé. 2008;3:83–99.

Fig. 3 Contributing factors to horizontal inequities in primary and specialist care use among First Nations populations (non- status, status, and unstratified). Note: The WC and EC represent the utilization of the Wag- staff and Erreygers methods, respectively, for correcting the horizontal inequity index (HI)

-0.06 -0.04 -0.02 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14

Total_WC

Total_EC

Non-status_WC

Non-status_EC

Status_WC

Status_EC

A: Primary Care

Income Education Employment status Rurality Other

-0.04 -0.02 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16

Total_WC

Total_EC

Non-status_WC

Non-status_EC

Status_WC

Status_EC

B: Specialist Care

Income Education Employment status Rurality Other

2784 Journal of Racial and Ethnic Health Disparities (2024) 11:2766–2785

1 3

4. Chan J, Polo A, Zubizarreta E, Bourque JM, Hanna TP, Gaudet M, et al. Access to radiotherapy and its association with cancer outcomes in a high-income country: addressing the inequity in Canada. Radiother Oncol. 2019;141:48–55.

5. Corscadden L, Allin S, Wolfson M, Grignon M. Publicly financed healthcare and income inequality in Canada. Healthc Q. 2014;17:7–10.

6. van Doorslaer E, Wagstaff A. Equity in the delivery of health care: some international comparisons. J Health Econ. 1992;11:389–411.

7. Government of Canada. Health Act: R.S.C., 1985, c. C-6. , Ottawa; 1984.

8. Stankiewicz A, Herel M, Desmeules MM. Canada and other United Nations (UN) Member States endorsed the ’ ’Rio Politi- cal Declaration on Social Determinants of Health. Public Health Agency of Canada; 2012.

9. Schneider E, Sarnak D, Squires D, Shah A, Doty M. Mirror, Mir- ror 2017. International comparison reflects flaws and opportunities for better U.S. health care; 2017.

10. Cameron BL, Carmargo Plazas MDP, Salas AS, Bourque Bearskin RL, Hungler K. Understanding inequalities in access to health care services for aboriginal people: a call for nursing action. ANS Adv Nurs Sci. 2014;37:1–16.

11. Lafond G, Haver CRA, McLeod V, Clarke S, Horsburgh B, McLeod KM. Characteristics and residence of First Nations patients and their use of health care services in Saskatchewan, Canada: informing First Nations and Métis health services. J Eval Clin Pract. 2017;23:294–300.

12. Chung H, Ye M, Hanson C, Oladokun O, Campbell MJ, Kramer G, et al. Disparities in healthcare utilisation rates for Aboriginal and non-Aboriginal Albertan residents, 1997-2006: a population database study. PLoS One. 2012;7:e48355.

13. Shah BR, Gunraj N, Hux JE. Markers of access to and quality of primary care for aboriginal people in Ontario, Canada. Am J Public Health. 2003;93:798–802.

14. Loyola-Sanchez A, Hurd K, Barnabe C. Healthcare utilization for arthritis by indigenous populations of Australia, Canada, New Zealand, and the United States: a systematic review ☆. Semin Arthritis Rheum. 2017;46:665–74.

15. Mitrou F, Cooke M, Lawrence D, Povah D, Mobilia E, Guimond E, et al. Gaps in Indigenous disadvantage not closing: a census cohort study of social determinants of health in Australia, Canada, and New Zealand from 1981-2006. BMC Public Health. 2014;14:201.

16. Assembly of First Nations. The First Nations health transfor- mation agenda. Assembly of First Nations (AFN); 2017. [cited 2023 August 1]. Available from: https:// scoinc. mb. ca/ wp- conte nt/ uploa ds/ 2021/ 06/ FNHTA- AFN- wcag. pdf

17. Pulok MH, Hajizadeh M. Equity in the use of physician services in Canada’s universal health system: A longitudinal analysis of older adults. Soc Sci Med. 2022;307:115186.

18. Hirello L, Pulok M, Hajizadeh M. Equity in healthcare utiliza- tion in Canada’s publicly funded health system: 2000–2014. Eur J Health Econ. 2022;1:1–15. https:// doi. org/ 10. 1007/ s10198- 022- 01441-1.

19. Mangalore R, Knapp M. Income-related inequalities in com- mon mental disorders among ethnic minorities in England. Soc Psychiatry Psychiatr Epidemiol. 2012;47:351–9.

20. Statistics Canada and the Assembly of First Nations. A snap- shot. Status First Nations people in Canada [Internet]; 2021. [cited 2022 Nov 2]. Available from: https:// www150. statc an. gc. ca/ n1/ pub/ 41- 20- 0002/ 41200 00220 21001- eng. htm

21. Statistics Canada. Census Program [Internet]. Census of Popu- lation; 2022. [cited 2022 Nov 14]. Available from: https:// www12. statc an. gc. ca/ census- recen sement/ index- eng. cfm

22. Department of Justice. A consolidation of the constitution acts 1867 to 1982. Ottawa, Canada; 2012. [cited 2023 August 1].

Available from: https:// publi catio ns. gc. ca/ colle ctions/ colle ction_ 2013/ lois- statu tes/ YX1-1- 2012- eng. pdf

23. Statistics Canada: Census Program [Internet]. [cited 2021 Oct 16]. Available from: https:// www12. statc an. gc. ca/ census- recen sement/ index- eng. cfm

24. Vowel C, Got Status? Indian Status in Canada. Indigenous Writes. A guide to First Nations, Metis, and Inuit issues in Canada; 2016. 30–9.

25. Morency J-D, Caron-Malenfant É, Daignault D. Fertility of Aboriginal people in Canada: an overview of trends at the turn of the 21st century, vol. 7. Aboriginal Policy Studies; 2018.

26. Nelson SE, Wilson K. The mental health of Indigenous peo- ples in Canada: a critical review of research. Soc Sci Med. 2017;176:93–112.

27. Bombay A, Matheson K, Anisman H. The intergenerational effects of Indian residential schools: implications for the con- cept of historical trauma. Transcult Psychiatry. 2014;51:320–38.

28. Ford JD, Berrang-Ford L, King M, Furgal C. Vulnerability of Aboriginal health systems in Canada to climate change. Glob Environ Change. 2010;20(4):668–80.

29. Adelson N. The embodiment of inequity: health disparities in aboriginal Canada. Can J Public Health. 2005;96(Suppl 2):S45–61. Available from: http:// www. ncbi. nlm. nih. gov/ pub- med/ 16078 555

30. Frohlich KL, Ross N, Richmond C. Health disparities in Canada today: some evidence and a theoretical framework. Health Policy. 2006;79:132–43.

31. Gracey M, King M. Indigenous health part 1: determinants and disease patterns. The Lancet. 2009;374:65–75. Available from: http:// www. scien cedir ect. com/ scien ce/ artic le/ pii/ S0140 67360 96091 44

32. Romanow R. Building on values: the future of health care in Canada. CRNCC: Canadian Research Network for Care in the Community; 2002.

33. McGrail K. Medicare financing and redistribution in British Columbia, 1992 and 2002. Healthc Policy. 2007;2:123–37.

34. Canadian Institute for Health Information. National health expend- iture trends, 1975 to 2015. Canadian Institute for Health Informa- tion; 2015.

35. Law MR, Cheng L, Dhalla IA, Heard D, Morgan SG. The effect of cost on adherence to prescription medications in Canada. Cmaj. 2012;184:297–302.

36. Lee A, Morgan S. Cost-related nonadherence to prescribed medi- cines among older Canadians in 2014: a cross-sectional analysis of a telephone survey. CMAJ Open. 2017;5:E40–4.

37. Government of Canada. National Native Alcohol and Drug Abuse Program [Internet]. 2019 [cited 2023 Aug 1]. Available from: https:// www. sac- isc. gc. ca/ eng/ 15760 89851 792/ 15760 89910 366

38. Government of Canada. Benefits and services under the Non- Insured Health Benefits Program [Internet]. 2019 [cited 2019 Jun 12]. Available from: https:// www. canada. ca/ en/ indig enous- servi ces- canada/ servi ces/ non- insur ed- health- benefi ts- first- natio ns- inuit/ benefi ts- servi ces- under- non- insur ed- health- benefi ts- progr am. html

39. Allan B, Smylie J. First Peoples , second class treatment: the role of racism in the health and well-being of Indigenous peoples in Canada. Wellesley Institute; 2015.

40. Marchildon GP. Canada: health system review, vol. 15. European Observatory on Health Systems and Policies; 2013. p. 179.

41. Hadskis M, Hutt L, McNally M. Dental law in Canada. Toronto, Canada: LexisNexis; 2019.

42. Vongdara B, Léger D, Latendresse E, Budinski R. Aboriginal Peo- ples Survey, 2017. Ottawa: concepts and methods guide [Internet]; 2018. Available from: https:// www150. statc an. gc. ca/ n1/ pub/ 89- 653-x/ 89- 653- x2018 001- eng. htm

2785Journal of Racial and Ethnic Health Disparities (2024) 11:2766–2785

1 3

43. Aday LA, Andersen RM. Equity of access to medical care: a con- ceptual and empirical overview. Med Care. 1981;19:4–27.

44. Mitchell GK, Senior HE, Bibo MP, Makoni B, Young SN, Rosen- berg JP, et al. Evaluation of a pilot of nurse practitioner led, GP supported rural palliative care provision. BMC Palliative Care. 2016;15(1):1–11.

45. Bruce SG, Riediger ND, Lix LM. Chronic disease and chronic disease risk factors among First Nations, Inuit and Métis popula- tions of northern Canada. Chronic Dis Inj Can. 2014;34:210–7.

46. Pu C, Bai YM, Chou YJ. The impact of self-rated health on medi- cal care utilization for older people with depressive symptoms. Int J Geriatr Psychiatry. 2013;28:479–86.

47. Pu C, Tang G-J, Fang Y-T, Chou Y-J. Which domain of self-rated health best predicts medical care utilization among Taiwanese adults? J epidem. 2012;22(5):417–24.

48. Cullati S, Mukhopadhyay S, Sieber S, Chakraborty A, Burton- Jeangros C. Is the single self-rated health item reliable in India? A construct validity study. BMJ Glob Health. 2018;3:856.

49. Li C, Dou L, Wang H, Jing S, Yin A. Horizontal inequity in health care utilization among the middle-aged and elderly in China. Int J Environ Res Public Health. 2017;14:1–14.

50. Van Doorslaer E, Clarke P, Savage E, Hall J. Horizontal inequi- ties in Australia’s mixed public/private health care system. Health Policy (New York). 2008;86:97–108.

51. Gravelle H. Measuring income related inequality in health: standardisation and the partial concentration index. Health Econ. 2003;12:803–19.

52. Pulok MH, van Gool K, Hajizadeh M, Allin S, Hall J. Meas- uring horizontal inequity in healthcare utilisation : a review of methodological developments and debates. Eur J Health Econ. 2020;21:171–80.

53. O’Donnell O, van Doorslaer E, Wagstaff A, Lindelow M. Analyz- ing health equity using household survey data. a guide to tech- niques and their implementation. Washington DC: The World Bank; 2008.

54. Kjellsson G, Gerdtham UG. On correcting the concentration index for binary variables. J Health Econ. 2013;32:659–70.

55. Erreygers G. Correcting the concentration index. J Health Econ. 2009;28:504–15.

56. Wagstaff A, Evan D, Watanabe N. On decomposing the causes of health sector inequalities with an application to malnutrition inequalities in Vietnam. J Econom. 2003;112:207.

57. O’Donnell O, van Doorslaer E, Wagstaff A, Lindelow M. Ana- lyzing health equity using household survey data a guide to tech- niques and their implementation analyzing health equity using household survey data. World Bank Publications; 2008.

58. Fleurbaey M, Schokkaert E. Unfair inequalities in health and health care. J Health Econ. 2009;28:73–90.

59. Zangiabadi S, Costanian C, Tamim H. Dental care use in Ontario: The Canadian community health survey (CCHS). BMC Oral Health. 2017;17:1–8.

60. Asada Y, Kephart G. Equity in health services use and intensity of use in Canada. BMC Health Serv Res. 2007;7:41.

61. OECD. Health for Everyone? Paris: Social Inequalities in Health and Health Systems | en | OECD; 2019.

62. Allin S, Hurley J. Inequity in publicly funded physician care: what is the role of private prescription drug insurance? Health Econ. 2009;18:1218–32.

63. Pulok MH, van Gool K, Hall J. Inequity in physician visits: the case of the unregulated fee market in Australia. Soc Sci Med. 2020;255:113004.

64. Pulok MH. Horizontal equity in the Australian healthcare system: exploring the unknowns and updating the knowns. University of Technology Sydney (UTS); 2019.

65. van Doorslaer E, Koolman X, Jones AM. Explaining income- related inequalities in doctor utilisation in Europe. Health Econ. 2004;13:629–47.

66. Tavares LP, Zantomio F. Inequity in healthcare use among older people after 2008: the case of southern European countries. Health Policy (New York). 2017;121:1063–71.

67. Terraneo M. Inequities in health care utilization by people aged 50+: evidence from 12 European countries. Soc Sci Med. 2015;126:154–63.

68. Pulok M, van Gool K, Hall J. Inequity in healthcare use among the indigenous population living in non-remote areas of Australia. Public Health. 2020;186:35–43.

69. Allin S, Masseria C, Mossialos E. Equity in health care use among older people in the UK: an analysis of panel data. Appl Econ. 2011;43:2229–39.

70. Devaux M. Income-related inequalities and inequities in health care services utilisation in 18 selected OECD countries. Eur J Health Econ. 2015;16:21–33.

71. OECD. Health for Everyone?: social inequalities in health and health systems. Paris: OECD Publishing; 2019.

72. Hajizadeh M, Connelly LB, Butler JRG. Health policy and hori- zontal inequities of health-care utilization in Australia: 1983– 2005. Appl Econ Lett. 2012;19:1765–75.

73. Pulok MH, van Gool K, Hall J. Horizontal inequity in the utilisa- tion of healthcare services in Australia. Health Policy (New York). 2020;124:1263–71.

74. Devaux M, De LM. Income-related inequalities in health service utilisation in 19 OECD Countries, 2008-2009. In: OECD Health Working Papers. Paris, France: OECD Publishing; 2012. Report No.: 58.

75. World Health Organization. Primary health care (Now more than ever). Geneva, Switzerland: World Health Organization; 2008.

76. Romanow RJ. Commission on the future of health care in Canada. Building on values: the future of health care in Canada; 2002.

77. Browne AJ, Smye VL, Rodney P, Tang SY, Mussell B, O’Neil J. Access to primary care from the perspective of aboriginal patients at an urban emergency department. Qual Health Res. 2011;21:333–48.

78. Puxley C. Brian Sinclair inquest to look at hospital backlogs; man died after 34-hour ER wait. The Canadian Press; 2014.

79. Kurtz DLM, Nyberg JC, Van Den Tillaart S, Mills B, (OUAHRC) TOUAHR. Silencing of voice: an act of structural violence urban aboriginal women speak out about their experiences with health care. Int J Indig Health. 2013;4:53.

80. Tang SY, Browne AJ. “Race” matters: Racialization and egalitar- ian discourses involving Aboriginal people in the Canadian health care context. Ethn Health. 2008;13:109–27.

81. Levesque JF, Harris MF, Russell G. Patient-centred access to health care: conceptualising access at the interface of health sys- tems and populations. Int J Equity Health. 2013;12:1–9.

82. Rotondi MA, O’Campo P, O’Brien K, Firestone M, Wolfe SH, Bourgeois C, et al. Our Health Counts Toronto: Using respondent- driven sampling to unmask census undercounts of an urban indig- enous population in Toronto, Canada. BMJ Open. 2017;7:1–8.

Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.