This article was downloaded by: [Flinders University of South Australia] On: 10 February 2015, At: 01:06 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Ethnicity & Health Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/ceth20
Can ethnicity data collected at an organizational level be useful in addressing health and healthcare inequities? a
a
ab
Annette J. Browne , Colleen M. Varcoe , Sabrina T. Wong , a
Victoria L. Smye & Koushambhi B. Khan
a
a
School of Nursing, University of British Columbia, Vancouver, Canada b
Click for updates
Centre for Health Services and Policy Research, University of British Columbia, Vancouver, Canada Published online: 02 Aug 2013.
To cite this article: Annette J. Browne, Colleen M. Varcoe, Sabrina T. Wong, Victoria L. Smye & Koushambhi B. Khan (2014) Can ethnicity data collected at an organizational level be useful in addressing health and healthcare inequities?, Ethnicity & Health, 19:2, 240-254, DOI: 10.1080/13557858.2013.814766 To link to this article: http://dx.doi.org/10.1080/13557858.2013.814766
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &
Downloaded by [Flinders University of South Australia] at 01:06 10 February 2015
Conditions of access and use can be found at http://www.tandfonline.com/page/termsand-conditions
Ethnicity & Health, 2014 Vol. 19, No. 2, 240254, http://dx.doi.org/10.1080/13557858.2013.814766
Can ethnicity data collected at an organizational level be useful in addressing health and healthcare inequities?
Downloaded by [Flinders University of South Australia] at 01:06 10 February 2015
Annette J. Brownea*, Colleen M. Varcoea, Sabrina T. Wonga,b, Victoria L. Smyea and Koushambhi B. Khana a School of Nursing, University of British Columbia, Vancouver, Canada; bCentre for Health Services and Policy Research, University of British Columbia, Vancouver, Canada
(Received 13 June 2012; final version received 29 May 2013) Objective. Following arguments made in the USA, the UK and New Zealand regarding the importance of population-level ethnicity data in understanding health and healthcare inequities, health authorities in several Canadian provinces are considering plans to collect ethnicity data from patients at the point of care within selected healthcare organizations. The purpose of this paper is to examine the potential quality, utility and relevance of ethnicity data collected at an organizational level as a means of addressing health and healthcare inequities. Design. We draw on findings from a recent Canadian study that examined the implications of collecting ethnicity data in healthcare contexts. Using a qualitative design, data were collected in a large city, and included interviews with 104 patients, community and healthcare leaders, and healthcare workers within diverse clinical contexts. Data were analyzed using interpretive thematic analysis. Results. Our results are discussed in relation to discourses reflected in the current literature that require consideration in relation to the potential utility and relevancy of ethnicity data collected at the point of care within healthcare organizations. These discourses frame excerpts from the ethnographic data that are used as illustrative examples. Three key challenges to the potential relevance and utility of ethnicity data collected at the level of local healthcare organizations are identified: (a) issues pertaining to quality of the data, (b) the fact that data quality is most problematic for those with the greatest vulnerability to the negative effects of health inequities, and (c) the lack of data reflecting structural disadvantages or discrimination. Conclusion. The quality of ethnicity data collected within healthcare organizations is often unreliable, particularly for people from racialized or visible minority groups, who are most at risk, seriously limiting the usefulness of the data. Quality measures for collecting data reflecting ethnocultural identity in specific healthcare organizations may be warranted but only if mechanisms exist or are developed for linking ethnicity with measures of perceived discrimination, stigmatization, income level, and other known contributors to inequities. Methods for linking these kinds of data, however, remain underdeveloped or non-existent in most healthcare organizations. Keywords: ethnicity; health inequities; discrimination; healthcare organizations; data collection; Canada
*Corresponding author. Email:
[email protected] # 2013 Taylor & Francis
Ethnicity & Health
241
Downloaded by [Flinders University of South Australia] at 01:06 10 February 2015
Introduction The collection of ethnicity data at the organizational level in healthcare settings has been justified as necessary to reduce health and healthcare inequities. Evidence supporting this justification has been generated at the population level in various international contexts; however, applicability at the organizational level requires further consideration. To be clear, the need to address health and healthcare inequities is not in question. Mounting evidence reveals that health and healthcare inequities are increasing, and are significantly and positively correlated with experiences of racialization, socioeconomic inequities, and systemic and interpersonal discrimination (Krieger et al. 2005; Canadian Institute for Health Information 2008, 2010, 2012; Stuber, Meyer, and Link 2008; Williams and Mohammed 2009; Health Council of Canada 2010; Thomas, Dorling, and Smith 2010; Krieger 2011; Veenstra 2011). In Canada, for example, research continues to demonstrate that people who experience particular clusters of vulnerabilities, such as people living in poverty, elderly persons, single mothers in low income brackets, significant proportions of the Aboriginal1 population, people with stigmatizing conditions including HIV/AIDS, substance use or mental health issues, and refugees and various immigrant groups, among others, are more likely to become ill and are less likely to receive appropriate health services (Adelson 2005; Canadian Institute for Health Information 2008). Ethnicity, particularly as it intersects with multiple forms of disadvantage, such as racialization, discrimination and poverty is clearly linked to health inequities. Despite Canada’s leadership in analyses related to social determinants of health and population health, health and social inequities continue to increase with profound negative effects for individuals and the population as a whole (Health Council of Canada 2010). In response, policy- and decision-makers have become increasingly concerned with understanding, tracking, and developing strategies to reduce health and healthcare inequities. While the need to examine and address inequities is recognized as integral to improving healthcare quality, debate continues regarding the appropriate analytical categories for gathering such data (Afshari and Bhopal 2010; Harper et al. 2010). Following arguments made in the USA, the UK and New Zealand regarding the importance of population-level ethnicity data as one of several variables needed to understand patterns of health and healthcare inequities, health authorities in several Canadian provinces, and their respective healthcare organizations, have considered plans to collect ethnicity data from patients at the point of care within healthcare organizations. Until now, the collection of ethnicity data has been relegated to census and related population-level surveys, and has not been mandated in healthcare contexts. This shift toward obtaining ethnicity data as part of routine patient demographic information is framed as a seemingly pragmatic response to persistent health and access inequities. The rationale for collecting and analyzing ethnicity data in public health and population health databases is to reveal and monitor associations between ethnicity, health, access to services, and social status. However, because of the potential harms that can be incurred in the process of collecting such data, particularly from racialized, vulnerable patients (Hasnain-Wynia, Taylor-Clark, and Anise, 2011; Kandula et al. 2009; Varcoe et al. 2009), initiatives in Canada to collect ethnicity data
Downloaded by [Flinders University of South Australia] at 01:06 10 February 2015
242 A.J. Browne et al. would benefit from distinguishing between the utility of collecting and analyzing population-level data versus organizational-level data. We recognize that in research contexts, progress or impediments in addressing health disparities cannot be monitored without population-level racial or ethnic data (Krieger 2004; Krieger et al. 2005; Salway et al. 2009), and that the absence of such data can lead to a whitewashing of disparities. However, we argue that whether or not ethnicity data should be collected at the organizational level depends on (a) the effects of the data collection process on patients, (b) the quality of the data collected, and (c) the potential utility of such data in addressing health and access inequities. In a prior publication we focused on point (a) above; we analyzed the harms and benefits associated with collecting ethnicity data at the point of care within organizations, and showed that harms are greatest for people who experience racialization and who are most at risk in relation to inequities, and that such risks must be weighed against the potential benefits (Varcoe et al. 2009). We recommended that if such data are to be collected, the collection should occur at an administrative level as part of administrative data, such as health insurance data, not at the point of direct care, where it can negatively influence patients’ experiences seeking care. In this paper, we address the second and third interrelated points identified above (b and c respectively) regarding the quality and potential utility of ethnicity data. Our purpose is to contribute to the ongoing dialogue about the relevance of ethnicity data by examining the potential utility of ethnicity data collected at an organizational level (that is, by specific organizations or agencies) toward the stated aim of addressing health and healthcare inequities, which we argue depends largely on the quality of such data. In forming our arguments, we draw on findings from a recent Canadian study that examined the implications of collecting ethnicity data in healthcare contexts. We first provide an overview of the research that provides the foundation for our analysis. We then frame our results and discussion by drawing on discourses reflected in the extant literature to show how the rationale for collecting ethnicity data at an organizational level is articulated and justified. In the process, we consider issues particular to the Canadian context. Integrating illustrative examples from our empirical findings, we then present an analysis of key issues addressing: (i) the potential quality, utility and relevance of organizational-level ethnicity data, including the inherent links between the quality and potential usefulness of such data, and (ii) the need for alternate sources of data, and alternate premises and subsequent strategies to redress health and healthcare inequities. While our study was conducted in Canada, our findings and analysis have relevance in other international contexts. Methods To explore the implications of collecting ethnicity data at the organizational level in healthcare settings, we used a qualitative design to collect data in a large, ethnoculturally diverse Western Canadian city. Approval to conduct this research was granted by the health authority and a university ethics board. We collected the following sets of data: (a) three focus group interviews involving community leaders (n 18) from a range of ethno-cultural groups, who served on advisory committees for the health authority; (b) interviews with 16 healthcare workers, who were
Downloaded by [Flinders University of South Australia] at 01:06 10 February 2015
Ethnicity & Health
243
involved in either administering an ethnic identity question, or whose agencies were considering doing so as part of intake data; (c) semi-structured interviews with patients (n 60) seeking health services in either a sub-acute area (an area for patients triaged as stable and non-urgent) of a large urban emergency department, or a community health center; and (d) in-depth interviews with a purposive sample of 10 health policy decision-makers from two provinces in Western Canada, who were responsible for addressing health equity issues in their health authorities or policy units. The focus group included leaders who self-identified as ‘‘visible minorities2’’ (n 8) or as Aboriginal persons (n 10). The patient interviews included 22 people who self-identified as a member of an Aboriginal group, and 17 who self-identified as English-speaking Euro-Canadians. The remaining 21 patients self-identified as members of various other ethno-cultural groups, who would be defined as visible minorities. The patient interviews focused on their thoughts regarding people being asked to identify their ethnicity in healthcare settings, their past experiences being asked to identify their ethnicity, the potential benefits of collecting ethnicity data, and their possible concerns. A comprehensive discussion of the research methods, the approach to interpretive thematic analysis, and the findings from the community leaders’ and patients’ perspectives are reported elsewhere (Varcoe et al. 2009). In the current paper, our analysis focuses on the key issues to consider regarding the potential usefulness of ethnicity data collected at an organizational level in addressing health or healthcare inequities. We base our analysis on a synthesis of findings, drawing on data excerpts to illustrate several key points. Consistent with interpretative thematic analysis, relevant literature is integrated throughout to frame the discussion of findings. Our aim is not to provide an exhaustive report of the research findings per se, but to reconsider the potential effectiveness and relevancy of ethnicity data in addressing issues of equity in healthcare contexts as the most frequently cited reason for collecting such data. Results and discussion: analysis of key issues requiring consideration We begin by highlighting key discourses regarding the potential utility and relevancy of ethnicity data collected at the point of care within healthcare organizations. These discourses frame excerpts from our qualitative data, which we present as illustrative examples. Three main challenges to the potential relevance and utility of ethnicity data collected at the organizational level are identified: (a) issues pertaining to data quality, (b) the fact that data quality is most problematic for those with the greatest vulnerability to the negative effects of health inequities, and (c) the lack of data reflecting structural disadvantages or discrimination. Taking up population-level ideas about ethnicity data at the organizational level At the population level, arguments in favor of linking ethnicity data to health and healthcare outcomes are well established (e.g., Nazroo 2003; Krieger et al. 2005; Institute of Medicine 2009; Afshari and Bhopal 2010). Population-based and public health surveillance research shows that ethnicity and ‘‘race’’ can be analyzed as influential variables provided that critical analyses of their relevance in relation to improving health or healthcare are proposed. For example, social epidemiologists
Downloaded by [Flinders University of South Australia] at 01:06 10 February 2015
244 A.J. Browne et al. argue that collecting and analyzing population-level data are essential to monitoring progress toward addressing health inequities (Institute of Medicine 2009; Krieger 2011). Although such data can represent a double-edge sword, as Krieger (2004) writes, ‘‘we must often use population data appearing in categories that are far from ideal . . . [by] those who seek to use these data to illuminate and oppose social inequalities in health’’ (632). Given the clear refutation of race as a biological entity, the usefulness of ethnicity and race data depends on these categories operating as proxies for social experiences. Recently, a growing body of epidemiologic research using ethnicity/race data has demonstrated that experiences of racism are harmful to health via diverse pathways involving structural, institutional, interpersonal, and internalized racism (Stuber, Meyer, and Link 2008; Williams and Mohammed 2009; Krieger et al. 2010). For example, epidemiological analyses have revealed how racism and discrimination interact with socioeconomic status, ethnocultural identity, gender, and neighborhood characteristics to negatively influence health and access to healthcare (Krieger et al. 2005; Bhopal 2007; Afshari and Bhopal 2010; Krieger 2011). Whereas European countries routinely include socioeconomic data as part of health statistics to illustrate interactions among variables, in the USA, most public health and population-level surveillance studies remain focused on race/ethnicity as the primary variable of concern (Krieger et al. 2005; Krieger 2011). The net effect has been to obscure understandings of the pervasive patterning of health inequities by socioeconomic position within and across social groupings stratified by ethnic/racial category, gender, age and ability (Thomas, Dorling, and Smith 2010). These concerns have prompted calls for population-based socioeconomic data and measures of racial discrimination to be linked with race/ethnicity data in order to understand, monitor and address social inequities in health (Beckfield and Krieger 2009; Thomas, Dorling, and Smith 2010). In the USA, New Zealand and the UK, arguments about the value of race/ ethnicity data at the population level have been taken-up within organizations, agencies and governments to justify the collection of ethnicity data from patients at the individual level at the point of care (Gerrish 2000; Institute of Medicine 2009; Statistics New Zealand 2009; Cormack and Robson 2010). The primary justification rests on the assumption that collecting individual-level data is needed to address health and access inequities within organizations. Arguments about the analytical power and value of population-based data infuse decisions related to the collection of individual-level data in healthcare organizations. Underpinning these trends toward collecting individual-level ethnicity data are assumptions about the ostensible value of having such data to: (a) identify health and healthcare inequities, (b) implement prevention and intervention programs and accountability standards to foster greater equity and (c) facilitate the provision of culturally and linguistically appropriate healthcare (Institute of Medicine 2009; Statistics New Zealand 2009; Cormack and Robson 2010; Hasnain-Wynia et al. 2010; Hasnain-Wynia, TaylorClark, and Anise 2011; Thorlby et al. 2011). These justifications for individual-level data continue to be propagated despite evidence spanning from the late 1990s to the present, that such data, where collected, have not led to decreases in health inequities or significant improvements in healthcare (Bhopal 1998; Salway et al. 2009; Thorlby et al. 2011).
Downloaded by [Flinders University of South Australia] at 01:06 10 February 2015
Ethnicity & Health
245
In part, rising interest in collecting individual-level ethnicity data stems from the assumption that such data are required to reveal issues that contribute to health inequities most notably, inequitable treatment, discrimination, or services that are poorly aligned with patients’ presumed ethnocultural preferences (Institute of Medicine 2009; Salway et al. 2009; Statistics New Zealand 2009; Thorlby et al. 2011). Ethnicity data is assumed to be adequate as a proxy measure for experiences of racialization or discrimination, and culturally-based health behaviors are presumed to be linked to ethnicity and languages spoken. Research, however, shows that responses regarding ethnic origin are poor indicators of culturally-based practices, preferences or ethnocultural identity (Rummens 2003; Varcoe et al. 2009). Another prevalent justification for collecting individual-level ethnicity data is that knowledge of patients’ stated ethnicity is needed to determine risks for genetically-mediated diseases (Quan et al. 2006). Ethnicity in this context is presumed to be a proxy for ‘‘race’’, with the assumption that diseases are distributed according to racial classifications. These assumptions persist despite extensive research showing that genetic differences account for only a very small proportion of the variation in disease patterns across so-called ‘‘racial’’ groups3 (Fee 2006; Poudrier 2007). In part, this conflation tends to persist because of the very real negative effects that social experiences such as racism and other forms of discrimination have on health effects that may manifest at a biological level (Stuber, Meyer, and Link 2008; Krieger 2011). Issues particular to the Canadian context In Canada, collection of ethnicity data in healthcare contexts has not been mandated by provincial or federal governments or by local health authorities as it has in other countries, reflecting Canada’s particular sociopolitical context (Rummens 2003). Therefore, until recently, ethnicity data has not been collected at the point of care or as part of organizational-level administrative data. Exceptions include collection of individual patients’ ethnicity as part of (a) specific health screening programs (e.g., breast cancer screening and HIV testing), (b) health programming aimed at health issues known to be prevalent among specific ethnocultural groups (e.g., Aboriginal health programs), (c) a particular agency’s mandate to track Aboriginal status4 as demographic information, and (d) agencies or organizations that specifically target particular population groups such as immigrant, refugee or Aboriginal people. Most Canadian research on ethnicity and health, therefore, has relied on analyses of data from large national-level population surveys, such as the Canadian Community Health Survey, the National Population Health Survey, and the Canadian census. This is starting to shift, however, and health authorities in several regions in Canada are implementing mechanisms for collecting individual-level ethnicity data from patients either at the point of care (e.g., through the insertion of an ethnicity variable into patient demographic information collected at intake), or at an administrative level (e.g., with other health insurance data collection). This contrasts with other jurisdictions in Canada where provincial ministries of health and health authorities have taken an explicit stance against the collection of individual-level ethnicity data by healthcare organizations, arguing that such data would infringe on rights and freedoms related to protection of privacy in Canada. These diverse stances on the value of collecting ethnicity data within organizations signal a need to explore the implications of collecting such data in the Canadian context. Our study provided an
246 A.J. Browne et al. opportunity to consider the potential utility of such data at the healthcare organizational level to address health and healthcare inequities.
Key challenges to utility
Downloaded by [Flinders University of South Australia] at 01:06 10 February 2015
Our analyses identified three central challenges to the quality and potential utility of ethnicity data in addressing inequities in healthcare access when such data are collected from individuals for use at the level of local healthcare organizations.
Quality of the data The first challenge to the potential usefulness of ethnicity data collected from individuals accessing local healthcare organizations relates to the validity and reliability of such data challenges that mirror issues encountered when ethnicity data are collected at the population level. Although ‘‘ethnicity’’ has now overtaken ‘‘race’’ as the most commonly used epidemiological variable in health literature (Afshari and Bhopal 2010), the construct ‘‘race/ethnicity’’ continues to be pervasive in the literature reinforcing the notion that these concepts (and their measurement) are interchangeable. This contributes to the persistence of conceptual confusion about what information is actually being collected when patients are asked to identify their ethnicity. In Canada, concerns regarding the quality of ethnicity data collected at the population level (Rummens 2003; Potvin 2005; Canadian Institute for Health Information 2008) are relevant to data collected at the organizational level. The methodological challenge regarding how to reliably and accurately categorize people according to ethnicity has particular ramifications, given the increasing proportion of people who report ‘‘I am Canadian’’ as the sole response to the census question on ethnic origin (Rummens 2003; Thomas 2005). ‘‘Canadian’’ remains the most reported ethnic origin, and over half of the 10 million persons in the 2006 census who reported ‘‘Canadian’’ described their ancestry as exclusively Canadian without identifying any other ethnic affiliations. Findings from our study similarly showed that ‘‘Canadian’’ was a frequent response (Varcoe et al. 2009). Presuming the goal is to better understand the links between ethnicity and health inequities, such responses will be of questionable usefulness, because they do not yield information about people’s visible minority status, experiences of racialization, cultural preferences, languages spoken, or experiences of discrimination as issues of central concern in relation to health and healthcare inequities. Nor can it be assumed that responses to questions about ethnicity will reflect genetic predisposition to disease, lived cultural practices, or of central concern in relation to health inequities, perceived constraints on access to healthcare or quality of care. As our and others’ research continues to demonstrate, even when protocols regarding ethnicity data collection call for selfidentification, admitting clerks often ‘‘assign’’ ethnicity on the basis of appearances, perpetuating the ongoing (and often taken for granted) processes of categorizing people by ‘race’ (Gomez et al. 2005; Institute of Medicine 2009; Hasnain-Wynia et al. 2010). In our research, for example, a staff member who was expected to administer an ethnic identity question as part of her agency’s intake procedures discussed these challenges:
Ethnicity & Health
247
Downloaded by [Flinders University of South Australia] at 01:06 10 February 2015
. . . Because face to face, I can see them, makes it a little bit more obvious. . .I never ask. I only, you know, based on appearance, make a judgment as to what I thought that they were, and base my stats at the end of the day on that . . . [Staff 2]
The continued conflation of race with ethnicity also results in a narrow conceptualization of ethnicity as a categorical variable that is assumed to be readily elicited by patients and easily recorded by staff at the point of care. These attempts to operationalize ethnicity for the purposes of data collection stand in contrast to commonly cited definitions of ethnicity as a complex, fluid, and multifaceted concept that is linked to local and global sociopolitical contexts, and that reflects the shifting nature of identity politics and nationalism (Rummens 2003; Salway et al. 2009). Attempts to elicit quality data thus are compromised by conceptual muddling and methodological challenges about what information is actually desired, and the significance of that data in relation to health and healthcare inequities.
Data quality may be most problematic for those with the greatest vulnerability to the negative effects of health inequities Our findings and others’ show that ethnicity data collection is not a value-neutral process, particularly for individuals and members of groups who are most likely to experience the impact of health and healthcare inequities (e.g., Kandula et al. 2009; Varcoe et al. 2009; Hasnain-Wynia, Taylor-Clark, and Anise 2011). In our research, participants’ overwhelming response to the idea of being asked to identify their ethnicity in healthcare settings was deep concern about being judged negatively on the basis of assumptions and stereotypes associated with their ethnicity, and the possibility of receiving poorer care based on such judgments. Repeatedly, patients from visible minorities and Aboriginal groups drew on their personal experiences of discrimination in healthcare, and the challenges faced in accessing services to explain their rationale for not wanting to declare their ethnicity at the point of care. Many asserted that they would not reveal their ethnicity because of concerns that they might be treated differentially on the basis of the answer they provided, as described by this woman who identified as Aboriginal: It would make you uncomfortable saying I’m Sikh, I’m Indian, I’m Japanese and then, you know, do they [healthcare staff] get to stand behind the curtain and say I don’t want to touch someone who is Sikh, I don’t want to touch someone who is Aboriginal, you know? Everyone should be on an equal playing ground when it comes to healthcare . . . It suddenly could become a stamp on my file. [Patient 18]
The anticipation of negative treatment has been linked to chronic stress stemming from people’s perceived need to brace themselves, or be on the alert for ways to counter or respond to negative treatment (Stuber, Meyer, and Link 2008). In our research, many of the patient participants explained how they would ‘dress for success’ to either mask their identities or pre-empt being judged negatively or in relation to a stereotype, as this woman who identified as Aboriginal explained: Depending on how I dress and how I look at the time, they’ll [healthcare staff] ask me if I’m Aboriginal or not. If I’ve dressed really nicely and have all my make up on and my hair is done, you know, they usually won’t. [Patient 31]
248 A.J. Browne et al. In contrast, patients who did not anticipate racial stereotyping and discrimination pointed out that their comfort responding to ethnicity data collection questions depended on what they understood to be their relative privilege, as expressed by this Euro-Canadian respondent:
Downloaded by [Flinders University of South Australia] at 01:06 10 February 2015
It [an ethnicity question] wouldn’t bother me at all. . .because. . . I’m a white Canadian guy and I just can’t relate to what it would be like for someone who is ethnic to be asked those questions. I guess it would bother them but it doesn’t bother me. [Patient 35]
These excerpts point to the potential harms that may occur in terms of inadvertently invoking anxiety about negative judgments or discrimination, particularly among people who have experienced discrimination and healthcare inequities, and how these concerns may lead some people, particularly those most likely to anticipate racism and stereotyping, to offer less meaningful responses to ethnicity data collection questions. The current context of social inequities and their disproportionate impact on particular groups of people raises specific concerns about the potential quality of the data collected from those people the data is intended to serve most.
Individual-level ethnicity data does not necessarily reflect structural disadvantages and discrimination More consideration is needed to clarify what types of data are needed for healthcare organizations to address issues of health and healthcare inequities. Health and healthcare inequities persist because of intersecting structural factors and social processes (e.g. poverty, welfare policies, lack of social housing), social exclusion, individual and systemic discrimination, and marginalizing practices in healthcare not people’s self-identified or assigned ethnicity, culture, or genetic predisposition. For example, high rates of HIV/AIDS among Aboriginal women in Canada are not due to cultural traits or behaviors related to Aboriginality as an ethnocultural category; such rates are a manifestation of structural and interpersonal violence, economic marginalization, and unequal power relations (Varcoe and Dick 2008). A non-English speaking elderly woman living alone may lack access to healthcare, because of the lack of interpreter services, funds for transportation, or family or social support, without her ethnicity being the primary issue of importance. Focusing on the collection of ethnicity data in local healthcare settings in the absence of other variables related to inequities will not point to the kinds of organizational and system-wide changes needed to mitigate these root causes of health or healthcare inequities, or the negative health effects of such inequities. Exposure to various forms of individual or systemic discrimination and social exclusion are gaining attention as key pathways to health and healthcare inequities, yet little is known about how to track such data within specific organizations (Stuber, Meyer, and Link 2008; Williams and Mohammed 2009; Afshari and Bhopal 2010; Krieger et al. 2010). Issues of discrimination and inequitable access to healthcare also intersect with socioeconomic status, and evidence shows that these inequities are significantly greater for people in lower income brackets (Canadian Institute for Health Information 2010; Thomas, Dorling, and Smith 2010). Providers’ assumptions about class thus intersect with ethnocultural stereotypes (Tang and Browne 2008; Reutter et al. 2009; Browne et al. 2011; Smye et al. 2011). These have tangible
Ethnicity & Health
249
Downloaded by [Flinders University of South Australia] at 01:06 10 February 2015
effects on health inequities. Our findings showed that concerns about perceived discrimination, intersected with patients’ concerns about being judged negatively on the basis of their appearance as people living in poverty, as residents of the inner-city neighborhood, and often, as people affected by substance use and addictions. Patients made conscious efforts to mitigate the effects of ‘poverty stigma’ to avoid the potential dismissal of their health concerns if they appeared disheveled, as this man, who was a member of a visible minority, described: . . . If I was coming in here [an urgent care centre] looking pretty grubby they might push me aside and take the next person that looks, you know, may have more money or something like that. . . [Patient 46].
Patients also described their assumption that identifying as ‘Canadian’ would help them to avoid being categorized as a member of an ethnocultural group known to experience racism or discrimination, as the man above, who was a member of a visible minority reiterated, ‘I’m Canadian, I’m Canadian. . . that’s it’ [Patient 46]. In these cases, collecting ethnicity data runs the risk both of being irrelevant, particularly for people most likely to be at risk for discrimination and stereotyping, and of diverting attention away from underlying issues of perceived discrimination in healthcare. While we have suggested elsewhere that the quality of data may be improved and the harms reduced by collecting ethnicity data at administrative levels (Varcoe et al. 2009), the ethnic categories created are likely to be of limited usefulness unless combined with measures of perceived discrimination or other ways of assessing how issues of discrimination operate within healthcare systems. Further, race or ethnicity categories are likely to be poor proxies for the determinants of inequities, particularly when collected without other information such as socioeconomic data. Finally, having data does not ensure action. As Thomas, Dorling, and Smith (2010) describe of the evolving situation in the UK, ‘Despite government interventions, inequalities in health have not diminished; indeed in some cases the gap might have widened over the past 10 years’ (1). The possibility also exists that both collecting and having data will provide the appearance of action, appearance that paradoxically will reduce meaningful attention to strategies to reduce inequities based on ethnic discrimination.
Recommendations Discussions of how to elicit and use quality data about the root causes of health and healthcare inequities are challenging to address within specific healthcare organizations (as opposed to at the population level), because data about ethnicity intersect with perceived discrimination, racialization, poverty stigma, and other marginalizing processes that shape health and access to healthcare. Strategies to identify and address inequities will be needed at all levels in direct clinical practice, within specific healthcare organizations, within the healthcare system more broadly, and in society in general. The emphasis in healthcare organizations, particularly in the USA, the UK, New Zealand, and increasingly Canada, however, continues to be on the collection of ethnicity data as the variable of primary interest in relation to tracking, monitoring, and redressing health inequities. This is problematic, because ethnicity data will not be sufficient to understand or address the complex ways that
Downloaded by [Flinders University of South Australia] at 01:06 10 February 2015
250 A.J. Browne et al. inequities are embodied or sustained in healthcare systems. The particular situation in Canada requires special consideration; because of Canada’s unique sociopolitical and healthcare context, the collection of ethnicity data is only now being considered, and there are no federal, provincial or local mandates or directives related to how or where it is collected, or how such data can be linked to other variables. Given this environment, Canadian decision- and policy-makers, and those in other countries where ethnicity data are not collected, have a unique opportunity to develop new and innovative approaches to tracking, measuring and using various kinds of data to address and redress health and healthcare inequities. Our research in Canada, and research conducted in other international contexts, has shown that the quality of ethnicity data collected within healthcare organizations is often unreliable, particularly for people from racialized or visible minority groups who are most at risk of experiencing inequities. Additionally, other data needed to address inequities are often not collected at the organizational level. Therefore, the usefulness of ethnicity data is seriously constrained, and we suggest, may harmfully create the appearance of action. At the same time, reliable data pertaining to health inequities are urgently needed to: (a) show the links between perceived discrimination, marginalizing practices, and health or healthcare inequities, and (b) generate actions to mitigate the occurrence and impact of such inequities for particular groups. Quality measures for collecting data reflecting ethnocultural identity in specific healthcare organizations may well be warranted but only if mechanisms exist or are developed for linking ethnicity data with measures of perceived discrimination, stigmatization, income level, and other known contributors to inequities. To date, however, methods for linking these kinds of data remain underdeveloped or non-existent in most healthcare organizations, agencies or institutions. Regardless of whether or not such data are collected or available, healthcare organizations need to commit to actions to mitigate health and healthcare inequities (Browne et al. 2012). This raises questions about the desire, capacity and responsibility of healthcare organizations to be responsive to the broader policy and structural conditions giving rise to growing levels of health and healthcare inequities in diverse international contexts. In many cases, the growing body of evidence can provide sufficient grounds upon which to base such actions. We, therefore, put forward the following recommendations for healthcare organizations in tackling issues of equity:
Recommendation #1: Take action universally to decrease discrimination and stereotyping within organizations, and at the point of care. Action does not depend on collecting or having more data. For example, respectful practices (e.g. asking patients, ‘is there anything I should know that will help me provide you with better care?’) or inquiries about family or genetic histories do not require knowing a person’s declared ethnicity. However, where such data are collected or exist, they are only useful insofar as actions are taken to address the inequities they reveal. Recommendation #2: Make better use of existing data. Healthcare organizations can marshal existing evidence about what is already known about health and healthcare inequities, and how they are sustained. This could be achieved by analyzing disaggregated local-level data derived from currently available
Ethnicity & Health
Downloaded by [Flinders University of South Australia] at 01:06 10 February 2015
251
population-level data for direction on how to tailor services to mitigate inequities and better meet the needs of particular populations or groups. Recommendation #3: Develop ways of collecting organizational-level data on perceived discrimination in healthcare. Tracking and monitoring health and healthcare inequities require data that extends beyond ethnicity variables. What is urgently needed are ways of measuring perceived discrimination or experiences of discrimination (intersecting along the axes of ethnicity, visible minority status, class, sexual orientation, age, health status, etc.), and links to health and healthcare experiences. While qualitative data on these dimensions are available (and could be collected), few quantitative measures exist.
In closing, examining and addressing the root causes of health and healthcare inequities are integral to improving healthcare quality, and ultimately the health of the population. In the Canadian context as in other international jurisdictions, ethnicity data collected within organizations may be useful in monitoring and addressing inequities, particularly among populations who are subject to systemic racialization, discrimination and economic disadvantages. This potential utility, however, can only be realized when methods and mechanisms for linking such data with other variables of concern are further developed and refined. This will require further research into the complexities associated with examining the intersecting factors that shape health and healthcare inequities, and the role of healthcare organizations in addressing those complexities. Key messages (1) The collection of quality data reflecting ethnocultural identity in healthcare organizations may be warranted, but will be meaningful only if steps are taken to mitigate potential harms of collection, and mechanisms exist or are developed for linking ethnicity data with measures of perceived discrimination, stigmatization, income level, and other known contributors to inequities. To date, however, methods for linking these kinds of data remain underdeveloped or non-existent in most healthcare organizations, agencies or institutions. (2) Regardless of whether or not ethnicity data are collected or available, healthcare organizations need to commit to actions to mitigate health and healthcare inequities. In many cases, action does not depend on collecting or having more ethnicity data, but on making meaningful use of existing research-based evidence and knowledge. Acknowledgments This study was generously funded by the Michael Smith Foundation for Health Research. The authors thank the members of our research team for contributing to the analysis represented here including, Nadine Caplette, Elizabeth Stanger, Ron Peters, Betty Calam, Laurel Jebamani, and Tej Sandhu.
Notes 1. In Canada, the term ‘Aboriginal peoples’ generally refers to Indigenous groups comprising First Nations, Me´tis and Inuit peoples.
Downloaded by [Flinders University of South Australia] at 01:06 10 February 2015
252 A.J. Browne et al. 2. In Canada, ‘visible minority’ is defined by the Employment Equity Act, and includes ‘persons, other than Aboriginal people, who are non-Caucasian in race or non-white in colour’ (Statistics Canada 2011). This definition is revealing of the conflation of race and ethnicity in Canada, and contrasts with both the USA context, where race has legal status, and the UK context where ethnicity has legal status, whereas race does not. 3. Cystic fibrosis, for example, is more common in European-descended people than in African or Asian-descended people; sickle cell anemia is far more common among people from West African or Mediterranean heritage, yet neither disease reflects commonly defined ethnic or ‘racial’ categories commonly reported in Canada. 4. Aboriginal peoples’ identities are defined by the federal Indian Act, which sub-categorizes their socio-legal statuses. Those who claim ‘status’ as a ‘Registered Indian’ receive limited rights and entitlements (e.g., tax-exemptions on reserve lands; limited healthcare or educational benefits) (Fiske 2006). Rarely is demographic information collected for nonstatus groups, limiting analyses of much needed disaggregated data (Smylie et al. 2006).
References Adelson, N. 2005. ‘‘The Embodiment of Inequity: Health Disparities in Aboriginal Canada.’’ Canadian Journal of Public Health 96 (Suppl. 2): S45S61. http://journal.cpha.ca/index.php/ cjph/article/view/1490/1679. Afshari, R., and R. S. Bhopal. 2010. ‘‘Ethnicity has Overtaken Race in Medical Science: MEDLINE-based Comparison of Trends in the USA and the Rest of the World, 1965 2005.’’ International Journal of Epidemiology 39 (6): 16821683. doi:10.1093/ije/dyp382. Beckfield, J., and N. Krieger. 2009. ‘‘Epi Demos Cracy: Linking Political Systems and Priorities to the Magnitude of Health Inequities- Evidence, Gaps, and a Research Agenda.’’ Epidemiologic Reviews 31 (1): 152177. doi:10.1093/epirev/mxp002. Bhopal, R. 1998. ‘‘Spectre of Racism in Health and Health Care: Lessons from History and the United States.’’ British Medical Journal 316 (7149): 19701973. doi:10.1136/bmj.316. 7149.1970. Bhopal, R. S. 2007. ‘‘Racism in Health and Health Care in Europe: Reality or Mirage?’’ European Journal of Public Health 17 (3): 238241. doi:10.1093/eurpub/ckm039. Browne, A. J., V. L. Smye, P. Rodney, S. Y. Tang, B. Mussell, and J. O’Neil. 2011. ‘‘Access to Primary Care from the Perspective of Aboriginal Patients at an Urban Emergency Department.’’ Qualitative Health Research 21 (3): 333348. doi:10.1177/1049732310385824. Browne, A. J., C. Varcoe, S. Wong, V. Smye, J. Lavoie, D. Littlejohn, D. Tu, et al. 2012. ‘‘Closing the Health Equity Gap: Evidence-based Strategies for Primary Health Care Organizations.’’ International Journal for Equity in Health 11 (59): 115. doi:10.1186/14759276-11-59. Canadian Institute for Health Information. 2008. Reducing Gaps in Health: A Focus on Socioeconomic Status in Urban Canada. Ottawa, ON: Canadian Institute for Health Information. Canadian Institute for Health Information. 2010. Hospitalization Disparities by Socioeconomic Status for Males and Females. Ottawa, ON: Canadian Institute for Health Information. Canadian Institute for Health Information. 2012. Disparities in Primary Health Care Experiences among Canadians with Ambulatory Care Sensitive Conditions. Ottawa, ON: Canadian Institute for Health Information. Cormack, D., and C. Robson. 2010. Ethnicity, National Identity and ‘New Zealanders’: Considerations for Monitoring Maori Health and Ethnic Inequalities. Wellington: Te Ro¯pu¯ Rangahau Hauora a Eru Po¯mare. Fee, M. 2006. ‘‘Racializing Narratives: Obesity, Diabetes and the ‘Aboriginal’ Thrifty Genotype.’’ Social Science & Medicine 62 (12): 29882997. doi:10.1016/j.socscimed.2005. 11.062. Fiske, J.-A. 2006. ‘‘Boundary Crossings: Power and Marginalisation in the Formation of Canadian Aboriginal Women’s Identities.’’ Gender & Development 14 (2): 247258. doi:10.1080/13552070600747255.
Downloaded by [Flinders University of South Australia] at 01:06 10 February 2015
Ethnicity & Health
253
Gerrish, K. 2000. ‘‘Researching Ethnic Diversity in the British NHS: Methodological and Practical Concerns.’’ Journal of Advanced Nursing 31 (4): 918925. doi:10.1046/j.13652648.2000.01357.x. Gomez, S. L., J. L. Kelsey, S. L. Glaser, M. M. Lee, and S. Sidney. 2005. ‘‘Inconsistencies between Self-reported Ethnicity and Ethnicity Recorded in a Health Maintenance Organization.’’ Annals of Epidemiology 15 (1): 7179. doi:10.1016/j.annepidem.2004.03. 002. Harper, S., N. B. King, S. C. Meersman, M. E. Reichman, N. Breen, and J. Lynch. 2010. ‘‘Implicit Value Judgments in the Measurement of Health Inequalities.’’ The Milbank Quarterly 88 (1): 429. doi:10.1111/j.1468-0009.2010.00587.x. Hasnain-Wynia, R., K. Taylor-Clark, and A. Anise. 2011. ‘‘Collecting Race, Ethnicity, and Language Data to Identify and Reduce Health Disparities: Perceptions of Health Plan Enrollees.’’ Medical Care Research and Review 68 (3): 367381. doi:10.1177/107755871 0382549. Hasnain-Wynia, R., K. Van Dyke, M. Youdelman, C. Krautkramer, S. L. Ivey, R. Gilchick, E. Kaleba, and M. K. Wynia. 2010. ‘‘Barriers to Collecting Patient Race, Ethnicity, and Primary Language Data in Physician Practices: An Exploratory Study.’’ Journal of National Medical Association 102 (9): 769775. http://66.101.212.31/~nma2010/publications/ September%202010/OC769.pdf. Health Council of Canada. 2010. Stepping it Up: Moving the Focus from Health Care in Canada to a Healthier Canada. Toronto: Health Council of Canada. Institute of Medicine. 2009. Race, Ethnicity, and Language Data: Standardization for Health Care Quality Improvement. Washington, DC: The National Academies Press. Kandula, N. R., R. Hasnain-Wynia, J. A. Thompson, E. R. Brown, and D. W. Baker. 2009. ‘‘Association between Prior Experiences of Discrimination and Patients’ Attitudes towards Health Care Providers Collecting Information about Race and Ethnicity.’’ Journal of General Internal Medicine 24 (7): 789794. doi:10.1007/s11606-009-0991-z. Krieger, N. 2004. ‘‘Data, "Race," and Politics: A Commentary on the Epidemiological Significance of California’s Proposition 54.’’ Journal of Epidemiology and Community Health 58 (8): 632633. doi:10.1136/jech.2003.018549. Krieger, N. 2011. Epidemiology and the People’s Health: Theory and Context Oxford. England: Oxford Scholarship Online. Krieger, N., D. Carney, K. Lancaster, P. D. Waterman, A. Kosheleva, and M. Banaji. 2010. ‘‘Combining Explicit and Implicit Measures of Racial Discrimination in Health Research.’’ American Journal of Public Health 100 (8): 14851492. doi:10.2105/AJPH.2009.159517. Krieger, N., J. T. Chen, P. D. Waterman, D. H. Rehkopf, and S. V. Subramanian. 2005. ‘‘Painting a Truer Picture of US Socioeconomic and Racial/Ethnic Health Inequalities: The Public Health Disparities Geocoding Project.’’ American Journal of Public Health 95 (2): 312323. doi:10.2105/AJPH.2003.032482. Nazroo, J. Y. 2003. ‘‘The Structuring of Ethnic Inequalities in Health: Economic Position, Racial Discrimination, and Racism.’’ American Journal of Public Health 93 (2): 277284. doi:10.2105/AJPH.93.2.277. Potvin, M. 2005. ‘‘The Role of Statistics on Ethnic Origin and ‘Race’ in Canadian Antidiscrimination Policy.’’ International Social Science Journal 57 (183): 2742. doi:10.1111/ j.0020-8701.2005.00529.x. Poudrier, J. 2007. ‘‘The Geneticization of Aboriginal Diabetes and Obesity: Adding Another Scene to the Story of the Thrifty Gene.’’ Canadian Review of Sociology/Revue Canadienne De Sociologie 44 (2): 237261. doi:10.1111/j.1755-618X.2007.tb01136.x. Quan, H., A. Wong, D. Johnson, and W. Ghali. 2006. ‘‘The Public Endorses Collection of Ethnicity Information in Hospital: Implications for Routine Data Capture in Canadian Health Systems.’’ Healthcare Policy 1 (3): 5564. doi:10.12927/hcpol..18115. Reutter, L. I., M. J. Stewart, G. Veenstra, R. Love, D. Raphael, and E. Makwarimba. 2009. ‘‘Who Do they Think we are, Anyway?’’: Perceptions of and Responses to Poverty Stigma.’’ Qualitative Health Research 19 (3): 297311. doi:10.1177/1049732308330246. Rummens, J. A. 2003. ‘‘Ethnic Ancestry, Culture, Identity, and Health: Using Ethnic Origin Data from the 2001 Canadian Census.’’ Canadian Ethnic Studies 35 (1): 84112.
Downloaded by [Flinders University of South Australia] at 01:06 10 February 2015
254 A.J. Browne et al. Salway, S., P. Allmark, R. Barley, G. Higginbottom, K. Gerrish, and G. T. H. Ellison. 2009. ‘‘Researching Ethnic Inequalities.’’ Social Research Update 58: 14. http://sru.soc.surrey.ac. uk. Smye, V., A. J. Browne, C. Varcoe, and V. Josewski. 2011. ‘‘Harm Reduction, Methadone Maintenance Treatment and the Root Causes of Health and Social Inequities: An Intersectional Lens in the Canadian Context.’’ Harm Reduction Journal 8 (1): 17. doi:10.1186/1477-7517-8-17. Smylie, J., I. Anderson, M. Ratima, S. Crengle, and M. Anderson. 2006. ‘‘Indigenous Health Performance Measurement Systems in Canada, Australia, and New Zealand.’’ The Lancet 367 (9527): 20292031. doi:10.1016/S0140-6736(06)68893-4. Statistics Canada. 2011. Visible Minority of Person [online]. Ottawa, ON, Canada: Statistics Canada. http://www.statcan.gc.ca/concepts/definitions/minority-minorite1-eng.htm. Statistics New Zealand. 2009. Final Report of a Review of the Official Ethnicity Statistical Standard 2009. Wellington: Statistics New Zealand. Stuber, J., I. Meyer, and B. Link. 2008. ‘‘Stigma, Prejudice, Discrimination and Health.’’ Social Science and Medicine 67 (3): 351357. doi:10.1016/j.socscimed.2008.03.023. Tang, S. Y., and A. J. Browne. 2008. ‘‘‘Race’ Matters: Racialization and Egalitarian Discourses Involving Aboriginal People in the Canadian Health Care Context.’’ Ethnicity & Health 13 (2): 109127. doi:10.1080/13557850701830307. Thomas, D. 2005. ‘‘I am Canadian’’. Canadian Social Trends 76: 17. http://www5.statcan.gc. ca/bsolc/olc-cel/olc-cel?catno=11-008-X&chropg=1&lang=eng. Thomas, B., D. Dorling, and G. D. Smith. 2010. ‘‘Inequalities in Premature Mortality in Britain: Observational Study from 1921 to 2007.’’ British Medical Journal 341: 16. doi:10.1136/bmj.c3639. Thorlby, R., S. Jorgensen, B. Siegel, and J. Z. Ayanian. 2011. ‘‘How Health Care Organizations are Using Data on Patients’ Race and Ethnicity to Improve Quality of Care.’’ Milbank Quarterly 89 (2): 226255. doi:10.1111/j.1468-0009.2011.00627.x. Varcoe, C., A. J. Browne, S. Wong, and V. L. Smye. 2009. ‘‘Harms and Benefits: Collecting Ethnicity Data in a Clinical Context.’’ Social Science & Medicine 68 (9): 16591666. doi:10.1016/j.socscimed.2009.02.034. Varcoe, C., and S. Dick. 2008. ‘‘The Intersecting Risks of Violence and HIV for Rural Aboriginal Women in a Neo-colonial Canadian Context.’’ Journal of Aboriginal Health 4 (1): 4252. http://www.naho.ca/journal/2008/01/11/the-intersecting-risks-of-violence-andhiv-for-rural-aboriginal-women-in-a-neo-colonial-canadian-context/. Veenstra, G. 2011. ‘‘Race, Gender, Class, and Sexual Orientation: Intersecting Axes of Inequality and Self-rated Health in Canada.’’ International Journal of Equity in Health 10 (3): 111. doi:10.1186/1475-9276-10-3. Williams, D. R., and S. A. Mohammed. 2009. ‘‘Discrimination and Racial Disparities in Health: Evidence and Needed Research.’’ Journal of Behavioral Medicine 32 (1): 2047. doi:10.1007/s10865-008-9185-0.