Received Date: August 14, 2018 Accepted Date: September 05, 2018 Published Date: September 12, 2018
Background: This case study assessed the acceptability and experience of being asked a single question about Indigenous self-identity when patients registered at a core neighborhood hospital in a western Canadian province. Registration clerks were trained and requested to ask all patients about their identity during the admission process and record the information in the registration system. The information was used to link patients to cultural support and navigation services.
Methods: Two hundred and fifty-nine participants were interviewed to examine perceptions and experiences with being asked about their identity. Descriptive statistics including frequencies and percentages were used to summarize participant demographic characteristics and responses. Chi-squared or Fisher’s Exact tests were used to explore the relationship between demographic characteristics (age, gender, Indigenous identity) and domains of inquiry. A thematic analysis of participants’ responses was also done.
Findings: Sixty (23.1%) participants self-identified as Indigenous and just over half (n=134) of all participants were female. Participants supported universally offering the opportunity to disclose their Indigenous identity. However, perceived importance of asking was associated with Indigenous identity (χ2 =16.52, p<0.001) and age (χ2=11.08, p=0.023). Elderly participants and those who did not identify as Indigenous assigned lower importance to asking about Indigenous identity. Among Indigenous participants who had prior negative experiences with the health care system, there were concerns about misuse of information and adverse effects on care. Preferences for how identity information should be collected were influenced by expectations about roles of health care providers and practical considerations.
Conclusion: While acceptable, perceived value of collecting self-reported Indigenous identity at registration varies across patient groups. Priority should be given to targeted communication about the rationale for collection of identity information and how it will be used to improve the delivery of care.
What is known about this topic:
• There is limited research related to the collection of information about Indigenous identity in Canadian health care settings.
• Perceptions that public and patient support for ethnicity data collection is lacking.
• Debates about the value and implications of collecting ethnicity data in health care settings.
What this paper adds:
• Better understanding of patient preferences and concerns about collection and use of Indigenous identity data.
• Context and use of Indigenous identity information affects acceptability and experience with data collection in health settings.
• Insights about potential strategies to support implementation of data collection for equity purposes.
Indigenous identity; First Nations; Metis; Social determinants of health; Canada; Cultural safety
Indigenous peoples are culturally and linguistically heterogenous but share a distinct political context, history and ancestral connection to first inhabitants of a given region . In Canada, Indigenous people account for 4.9% of the total population and comprise a higher proportion (15% - 85%) of the population in West and Northern provinces and territories . Recent available data also indicate that the Indigenous population is growing faster and remains younger by almost a decade than the general population . Despite the youthful profile, Indigenous peoples bear a disproportionate share of the morbidity, mortality burden from chronic diseases and experience a shorter life expectancy than non-Indigenous Canadians [3-7]. Several studies also suggest higher rates of communicable diseases including tuberculosis, respiratory tract infections and some sexually transmitted infections [8-12]. The growing evidence of health disparities presents a compelling argument for equity-oriented approaches to delivery of care and opportunities to strengthen data systems for collection and use of identity information [13,14].
Indigenous identity (refers to an umbrella term for persons who self-identify as First Nations, Métis or Inuit) is an important determinant of health in Canada [15,16]. The legacy of colonization and culturally destructive processes of residential schools have traumatized several generations of Indigenous peoples . Over time, these complex forces have continued to shape behavioral risks and lived experience of many people who identify as First Nations, Métis or Inuit. While being careful not to reinforce stereotypes, the collection of information about Indigenous identity within health care settings could assist care providers in the delivery of patient-centred care as well as inform planning for more responsive services.
There is limited information available about self-identification as Indigenous in health care settings . Although Registered Indian Status (RIS) is recorded on health cards and is used to bill for Non-Insured Health Benefits, it is not integrated with clinical information systems. The Canadian Institute for Health Information (CIHI) has recognized the limited availability of sociodemographic data to support measurement of health inequity as an indicator of health system performance . In 2016, CIHI convened national stakeholders to identify core ‘stratifiers’, including Indigenous identity, for use in measuring health inequalities. Our local efforts support the national work and contribute to understanding the acceptability of collecting of information about Indigenous identity from the patient’s perspective. We report our experience with implementation to help others gauge feasibility and challenges in similar contexts.
The study was conducted in a medium sized Canadian city where taxation-based publicly funded services (Medicare) are delivered by a provincial health care system in parallel with fee for service physician models . The research was conducted at an inner city hospital where cultural support, interpretation and patient navigation services are available and offered to patients who identify as First Nations or Métis . Prior to the study, analysis of surnames and home community as listed on the daily hospital census was used to identify patients who would be offered information about these services. The limitations of this approach and opportunities for misclassification of individuals were recognized. The study addressed an important service gap and assessed the acceptability of asking about Indigenous identity at registration. It was the first step to understanding patients’ preferences and concerns related to collection of Indigenous identity data.
A site implementation team worked closely with the research team to determine priorities and information needs for the study. Relevant stakeholders, including Indigenous community health leaders and patient/family advisory councils were engaged prior to and during project implementation. The study protocol was reviewed and approved by the relevant community stakeholders. There was respect for established ceremonial protocols. A pipe ceremony was performed by elders to symbolize partnership, seek spiritual guidance for the project and favor to reach its objectives. Ethics approval was obtained from the relevant Institutional Behavioral Research Ethics Board. Operational approval was also obtained from the respective health region authorities.
A single question about Indigenous identity was added to the demographic tab of the electronic registration program. Registration clerks were trained and requested to ask all patients, “Would you like to self-identify as First Nations or Métis?” Response options were limited to “yes” or “no” due to limitations of a legacy registration system. The field was left blank if the question was not asked. If the patient declined to answer, this was recorded as “no” so that the question was not asked on a subsequent occasion.
The evaluation of implementation employed a mixed method case study research strategy to assess feasibility and acceptability of collection of self-reported Indigenous identity. Acceptability was defined as “the perception among implementation stakeholders that a given practice, service or innovation is agreeable, palatable or satisfactory” . In this study, acceptability was assessed based on perceptions and experiences of participants with the process of implementation. Multiple perspectives were examined including patients, registration staff and key informants. However, this article focuses on the patients’ experience in order to fully explore a key factor in the success of efforts to collect Indigenous identity data in health care settings.
We conducted intercept interviews with registered patients to understand their perceptions and experiences with collection of information about Indigenous identity. The intercept interview or ‘person on the street’ interview is commonly used in marketing research . In this method, an interviewer intercepts a sample of patients who pass by to ask if they would like to participate in a study. Those who agree are either interviewed on the spot or taken to a separate area for the interview. This approach to data collection was practical and appropriate given the study’s context.
Eligible participants were at least 18 years old, had been asked about their identity at main registration of the hospital, and were able to speak fluently in english. There was a different registration area for persons presenting with acute concerns to the Emergency Department. Recruitment of participants did not occur in this area. Persons who presented to main registration accessed a wide variety of hospital services including but not limited to day surgery, specialty outpatient clinics (e.g. cardiology, endocrinology, orthopedics, urology) and diagnostic imaging. Those persons with requests for laboratory services were not required to register prior to proceeding to the lab. It was important to ask individuals who had exposure to the experience of interest – being asked about Indigenous identity as this provided an actual point of reference for answering all interview questions.
Interviews were conducted over a six week period during the hours of 8 am to 4 pm. The schedule for data collection included alternate week days. An imaginary line was used to count patients as they were leaving the main registration area. Every fifth patient on a recruiting day was approached by one of two researchers and invited to participate. If an intercepted individual agreed and was eligible, the researcher explained the study objectives, procedures, and provided assurances about the voluntary nature of participation and anonymity of responses. Oral consent was documented prior to capturing the required information. When possible, reasons for refusal to participate were noted.
The two researchers (one a doctoral candidate) received training in research methods including advanced qualitative analyses. Both individuals also had experience working with Indigenous peoples. The research was also supported by other members of the project team with relevant methodological and content expertise.
The available local data suggested that differences exist in the pattern of health care utilization by Indigenous and non- Indigenous peoples. Inpatient participants who self-identified as Indigenous were purposively recruited with the assistance of First Nations and Métis patient navigators. These participants ensured that the perceptions of these primary stakeholders were adequately reflected in the evaluation. This subset of participants also captured experiences of patients who were more acutely ill and had been registered through the Emergency Department.
An intercept tracking form facilitated quick and easy recording in the field. The date of the interview was noted as well as participants’ self-reported age, gender and racial/ethnic identity. All questions were open-ended and asked in a standardized way. Questions focused on patients’ level of comfort with being asked about their identity, perceived importance of collecting Indigenous identity information, the appropriate target for data collection and preferences for mode of question administration. Responses were categorized using predetermined labels and longer explanations and were noted in an adjacent space.
Data were subsequently entered in an Excel spreadsheet. Descriptive statistics including frequencies and percentages were used to summarize participant demographic characteristics and responses. Chi-squared or Fisher’s Exact tests (if small cell sizes) were used to explore the relationship between demographic characteristics (age, gender, Indigenous identity) and domains of inquiry. A multivariable logistic regression model was fitted with perceived importance of asking about identity as the dependent variable and age group, sex and identity as independent variables to adjust for confounding effects of significant demographic variables on the outcome of interest. Odds ratios and 95% confidence intervals were calculated. A thematic analysis of participant responses was also done . An inductive approach to coding was applied with assignment of labels to segments of text until the entire data corpus was completed. The various codes were compared and then sorted into potential themes. These potential themes were refined by looking for coherence among codes as well as themes. Final themes were named and explained with a short narrative. Excerpts of the data were selected to capture the essence of a particular theme.
Data collection and analysis proceeded iteratively. The data was summarized and discussed weekly with members of the project team. This allowed for questioning of any assumptions and grounding of insights in the data that was collected. Preliminary results were also shared with stakeholders including managers and advisory councils who provided feedback and important context to help with interpretation of the results.
Among 459 patients who were approached, 259 (56.4%) adults agreed to be interviewed. The main reasons for declining included disinterest and time constraints. Several persons (n=130, 28.3%) who were approached were ineligible to be interviewed as they had not been asked about their identity at registration. Descriptive characteristics of the sample are summarized in (Table 1). Just over half (51.7%) of participants were female. Sixty (23.1%) participants self-identified as Indigenous. Among self-identified non-Indigenous participants, 193 (96.9%) described their racial/ethnic identity as Caucasian. Compared to non-Indigenous participants, a lower proportion (13.3% vs 31.7%) of those who self-identified as Indigenous were elderly (≥65 years). As previously explained, we intentionally recruited self-identified Indigenous inpatients to capture diverse perspectives.
|Self-identified Indigenous (N=60, %)||34
|Non-Indigenous (N=199, %)||100
Table 1: Descriptive characteristics of interview participants.
Most participants had positive experiences with data collection and the majority (95.4%) reported that they were comfortable being asked although they were surprised that the information was requested. A higher proportion of participants who were uncomfortable self-identified as Indigenous (10%) compared to their non-Indigenous counterparts (3%). Non-Indigenous persons who expressed discomfort also perceived that it was either inappropriate to ask or unnecessary for the delivery of care (Table 2). This was consistent with limited appreciation for the importance of asking about Indigenous identity. In contrast, Indigenous participants who expressed discomfort were concerned that identity information would be misused and adversely affect care provided.
|Level of comfort with being asked about identity||χ2 =0.015||p=0.90||χ2 =5.09||p=0.024|
|Uncomfortable/Do not know||12(4.6)||0||10(6.6)||1(1.4)||6(4.8)||6(4.5)||6(10)||6(3.0)|
|Perceived importance of asking||χ2 11.075||p=0.026||χ2 =3.70||p=0.156||χ2 =16.518||p=<0.001|
|Do not know||49(18.9)||8(24.2)||27(17.8)||13(18.3)||25(20)||24(17.9)||9(15)||40(20.1)|
|Who should be asked about their identity?||χ2 =1.8||p=0.406||χ2 =1.103||p=0.29||χ2=0.393||p=0.530|
|Other (Some people/No one)||55(21.2)||5(15.2)||29(19.1)||18(25.4)||30 (24)||25 (18.7)||11(18.3)||44 (22.1)|
|Most appropriate person to collect identity information||χ2=10.34||p=0.109||χ2=0.613||p=0.89||χ2 =14.85||p=0.001|
|No preference||42 (16.2)||9(27.2)||22(14.5)||10(14.1)||23(18.4)||25(18.6)||16(26.7)||32(16.1)|
|Care provider||33 (12.7)||5(15.2)||17(11.2)||10(14.1)||15(12)||17(12.7)||14(23.3)||18(9.0)|
*Three missing observations χ2 refers to chi-squared statistic
Table 2: Distribution of participant responses by demographic characteristics.
Participants were also asked about their perceptions of the most appropriate target population for data collection. There was some support from both Indigenous and non-Indigenous participants for a universal approach to avoid subjectivity and profiling some patients. One participant shared this view, “I think you should ask everyone or no one in order to be fair” (P15). There was also an appreciation that it could be difficult to distinguish identities based only on physical appearance. A participant who self-identified as Métis explained, “[You] can’t tell by appearance alone. Some are Aboriginal but don’t look it, so it is best to ask” (P1).
Although most participants appreciated the limitations of physical attributes as a marker of identity, a few participants argued that it was unnecessary to ask patients who appeared to be non-Indigenous. A non-Indigenous participant argued, “Do I look First Nations? I think that common sense should apply in some cases where you can see that the person is not native [Indigenous]” (P105).
Perceptions of importance of asking about Indigenous identity were associated with Indigenous self-identity (p<0.001) and age (p=0.026) in bivariate analyses. In adjusted multivariable regression analyses, age group (p=0.093) was no longer independently associated with perceptions of importance (Table 2). However, persons who self-identified as Indigenous had 3.1 times greater odds (p<0.001, 95% CI OR 1.64 -5.78) of supporting asking about identity that non-Indigenous persons. Participants who felt that it was not important to collect identity information cited equality arguments or that it was not personally relevant to them. Interviews pointed to several factors that influenced participant willingness to disclose their identity. These themes will be further explored in the following sections (Table 3).
|Model||OR||95% Confidence Intervals OR||P value|
|Self-identified Indigenous identity
18 -34 years
35 – 64 years
OR= Odds Ratio
Table 3: Results of multivariable logistic regression model for perceived importance of asking about identity, sex and age group showing odd ratios, 95% confidence intervals and p values.
Understanding the reason for data collection
Participants who perceived that it was important to ask about their identity understood the reasons for data collection. Some participants had seen the project poster, requested information from the clerk or were aware of cultural support services. In contrast, other participants were bewildered and felt unsettled because they didn’t know why they were being asked about their identity. This was exemplified by a non-Indigenous participant who stated, “Explaining why the information is being requested, can go a long way to encouraging persons to answer” (P76). It was further reinforced by an Indigenous participant who agreed, “It can be uncomfortable if you don’t know the reason that they are asking” (P99).
Personal relevance of the question
Although non-Indigenous participants were comfortable with being asked, they generally perceived that the identity question had less personal relevance. A non-Indigenous participant explained, “I don’t think that it is important for me because I am not First Nations” (P32). Some Indigenous participants assumed that the question was asked to bill for Non–Insured Health Benefits. An Indigenous participant commented, “I was ok with it. I thought they asked because of the benefits for our people” (PI_7).
Perceived benefits in the delivery of care
Participants who perceived that collection of Indigenous identity information was beneficial for the delivery of care were more likely to support disclosure of Indigenous identity. An Indigenous participant who was aware of the disproportionate burden from chronic diseases among Indigenous peoples felt that it was important to consider screening based on Indigenous identity. He explained that, “People who are First Nations have a higher rate of diabetes and some other conditions, so it is important to know so that you can check for those diseases and treat them” (PI_3). In contrast, among participants who felt that it was not important to ask, the most commonly cited reasons were that it was inappropriate to ask or unrelated to the delivery of care.
Among some participants who identified as First Nations or Métis, their identity was an embodiment of self and source of pride. One participant aptly captured these sentiments, “I was very comfortable because I am secure in my identity. If it bothered me I would have told them” (PI_5). Another participant shared her views. “If they want to know, I will tell them who I am because I am proud of who I am” (P181).
Fear of misuse of information
Fear that identity information would be misused and adversely affect care was an important concern among Indigenous participants who expressed discomfort and assigned low priority to data collection. This was often coupled with prior negative experiences with the health care system. An Indigenous participant expressed this view, “It is not important to me especially if it will affect how you are treated in a bad way” (PI_15).
A few participants were concerned about privacy. This was due to the physical layout of the department where patient-staff interactions were visible and audible to others who were waiting to be served. One non-Indigenous participant commented, “The registration clerks should try to ask more discreetly” (P157).
Another participant offered this strategy for mitigating the concern, “Privacy may be an issue. There should be a silent process” (P151).
Perceptions that data collection was contrary to principles of equality
A few participants thought that requesting this information was contrary to the principles of equality in the context of universal access to health care in Canada. These non-Indigenous participants felt that data collection perpetuated or magnified historical tensions between Indigenous and non-Indigenous peoples and that everyone should be treated the same, thus negating need for additional information about personal characteristics. A non-Indigenous participant offered this comment, “Do we want to be one or not? We can’t have it both ways. Time for healing and coming together. When we have to single people out, we reinforce stereotypes and stay stuck in the past” (P138). Another non-Indigenous participant said, “I am opposed to special treatment for people who are First Nations, so I don’t think that it is important” (P25).
Other non-Indigenous participants provided a broader argument that the identity question should be more inclusive and enquire about other groups. One non-Indigenous participant suggested, “We shouldn’t single out one group for support, but we should be concerned about the needs of all ethnic groups” (P19). One non-Indigenous participant asserted that it was discrimination against the majority to offer cultural supports for only one group, “Isn’t it discrimination to ask the question and treat people differently?” (P175).
Preferences for mode of data collection varied and there was an association with self-reported identity (χ2 =16.07, p=0.001). A higher proportion of Indigenous participants indicated no preference (26.7%) or care provider (23.3%) compared to non- Indigenous participants (16.1% and 9% respectively). This seemed to be influenced by role expectations and practical considerations. Some preferred to be asked at registration for convenience, as other demographic data was already being collected in this setting. Others without a clear preference were more concerned that persons were asked in a respectful way wherever the information was collected. A non-Indigenous participant commented, “It doesn’t matter as long as people are asked respectfully” (P20). Those who preferred providers to ask an identity question reasoned that doctors and nurses were uniquely placed to assess whether the information is needed for care. One Indigenous participant explained, “I prefer the doctor because they are the ones taking care of you. They know if that is important to your care” (PI_33). Another Indigenous participant who was an inpatient shared a similar view, “The nurse is best because they get to know you and care about you as a person” (PI_34). A few participants felt that doctors and nurses’ time could best be used in provision of clinical care. A non-Indigenous participant explained, “Doctors and nurses have other things to do” (P206). Self-administration was seldom mentioned and only by non-Indigenous participants as a preferred method for collection of identity data (Figure 1).
There is limited research related to the collection of information about Indigenous identity in Canadian health care settings. Varcoe, Browne, Wong and Smye (2009) suggest that careful consideration should be given to potential harms for racialized and vulnerable populations despite good intentions for collection and use of ethnicity data . Therefore, it was important to assess the process of implementation and understand patient concerns and preferences to inform decisions about optimal processes for standardized local data collection. The pattern of responses suggested differences between the perspectives of Indigenous and non-Indigenous participants. Non-Indigenous participants perceived little benefit and consequently assigned lower priority to data collection. In contrast, Indigenous participants shared multiple views depending on their experiences with the health care system. Some supported data collection when they perceived benefits to delivery of care, while others expressed concerns about discrimination or negative stereotyping in health care environments. The views expressed in this study were similar to Varcoe et al. (2009) in that participants located their responses in their own social identities and experiences. A dual approach is likely to be required to address the different information needs and concerns of Indigenous and non- Indigenous participants .
Health disparities between Indigenous and non-Indigenous peoples in Canada have been well documented [26-28]. Community concerns about the role of systemic racism and discrimination and the need for cultural safety have increasingly become part of health system discourse . Intergenerational trauma and distrust of institutions because of historical human rights violations plague relations between the state and Indigenous people [28,30]. In this context, it is to be anticipated that questions about Indigenous identity may be met with resistance and cause discomfort, particularly if reasons for asking are not clear. The findings were consistent with these views in that persons who knew why the information was being requested and how it could be used to improve care were more receptive to data collection. This suggests that greater priority should be accorded to explaining why identity information is being collected and how it will be used. This may increase patient comfort with disclosure about Indigenous identity, especially if there are concerns about the benefits of data collection.
A few non-Indigenous participants felt that information about identity should not be collected. They argued that care should be independent of other considerations such as social context. This showed little acknowledgement of white privilege or understanding of the concept of equity . Similarly, perceptions that equal treatment of all persons for equal need is contravened by provisions of services that are targeted to the specific needs of Indigenous peoples are also cause for concern in the local context. These prejudicial views expose an insidious modern racism that serves to perpetuate inequity and signal potential for non-acceptance among non-Indigenous participants . This should be addressed in future information, education and communication messages.
Staff concerns about offending patients can undermine efforts to collect sociodemographic information [33,34]. Other Canadian projects have urged that quality staff education is critical for successful data collection [35,36]. Staff comfort was a challenge for local data collection efforts. Health care organizations need to build capacity of staff at all levels in order to achieve cultural safety in all patient-provider interactions. This likely requires more intense coaching over an extended period than offered during the study.
Preferences for mode of administration of the identity question varied among study participants. While interesting, it needs to be further explored and may reflect nuances associated with the context in which care was provided. Although integration into the routine registration process offers the most seamless option, multiple options are desirable. This would require capacity building among other groups of staff in order to successfully collect and apply the information.
The implementation of Indigenous identity data collection responded to a critical need of the organization. It augmented the hospital’s efforts to reach persons who identified as Indigenous and who could benefit from cultural support and navigation services. The current procedures based on surname analysis are unreliable particularly for identification of individuals who are Métis. In sites without an organizational imperative or compatible vision for equity, implementation of identity data collection may experience more resistance. There was extensive organizational advocacy and several critical conversations with senior leaders to canvass support in this site prior to implementation of data collection.
The support of leadership at all levels and a meaningful engagement process of Indigenous stakeholders greatly facilitated the implementation of Indigenous identity data collection. This process ensured that stakeholder information needs and preferences were understood; the project remained relevant to organizational priorities and respectful of cultural norms and values.
The study provided valuable information about implementation however; it also had several limitations. Although registration is a high traffic area that allowed for efficient recruitment, the nature of the encounter meant that some patients were reluctant to participate. Further, participants may have been preoccupied by health concerns or deterred by limited privacy as interviews were conducted on the spot. The interview also had to be kept brief to minimize inconvenience for participants. This limited the breadth of topics that could be covered.
Only limited information was available about patients who were not asked about their identity. Although they seemed to be similar to those who did not identify as Indigenous, it is important to gain a better understanding of why they were not asked. This would illuminate staff training needs and increase coverage and accuracy of the data.
The limitations of Indigenous identity as the single marker of vulnerability must be acknowledged . Additional information about social context is needed for full risk assessment. Further it is important to confirm that an individual identifies as Indigenous before offering cultural services as it is impossible to know whether the recorded status was ascribed based on physical appearance or Registered Indian Status as indicated on one’s health card
It was acceptable to ask patients at registration about Indigenous identity in an urban acute care setting. Perceived importance of collecting of information about Indigenous identity varied by Indigenous status and age. Preferences for how the information should be collected were influenced by convenience and role expectations of providers. Targeted education and communication is required to address different information needs across participant groups. Organizational, staff and patient factors were important for successful collection of information.
HWR was supported by an NSERC Create ITraP Fellowship during the study.