Indigenous identity identification in administrative health care data globally: A scoping review

Objective Both Indigenous and non-Indigenous governments and organizations have increasingly called for improved Indigenous health data in order to improve health equity among Indigenous peoples. This scoping review identifies best practices, potential consequences and barriers for advancing Indigenous health data and Indigenous data sovereignty globally. Methods A scoping review was conducted to capture the breadth and nature of the academic and grey literature. We searched academic databases for academic records published between 2000 and 2021. We used Google to conduct a review of the grey literature. We applied Harfield’s Aboriginal and Torres Strait Islander Quality Appraisal Tool (QAT) to all original research articles included in the review to assess the quality of health information from an Indigenous perspective. Results In total, 77 academic articles and 49 grey literature records were included. Much of the academic literature was published in the last 12 years, demonstrating a more recent interest in Indigenous health data. Overall, we identified two ways for Indigenous health data to be retrieved. The first approach is health care organizations asking clients to voluntarily self-identify as Indigenous. The other approach is through data linkage. Both approaches to improving Indigenous health data require awareness of the intergenerational consequences of settler colonialism along with a general mistrust in health care systems among Indigenous peoples. This context also presents special considerations for health care systems that wish to engage with Indigenous communities around the intention, purpose, and uses of the identification of Indigenous status in administrative databases and in health care settings. Partnerships with local Indigenous nations should be developed prior to the systematic collection of Indigenous identifiers in health administrative data. The QAT revealed that many research articles do not include adequate information to describe how Indigenous communities and stakeholders have been involved in this research. Conclusion There is consensus within the academic literature that improving Indigenous health should be of high priority for health care systems globally. To address data disparities, governments and health organizations are encouraged to work in collaboration with local Indigenous nations and stakeholders at every step from conceptualization, data collection, analysis, to ownership. This finding highlights the need for future research to provide transparent explanation of how meaningful Indigenous collaboration is achieved in their research.

The International Group for Indigenous Health Measurement (IGIHM) has pursued several activities to improve health measurement of Indigenous people globally.There is an ongoing challenge to measure Indigenous health due to the misidentification of Indigenous people on administrative and health records.The major issue underlying this is a problem of identificationwho is considered Indigenous and how is this documented in health data?The IGIHM is attempting to address these challenges through sharing of best practices, lobbying national statistical organizations, and reaching out to countries beyond the four founding members of the group.This work is intended to increase the awareness within national statistical organizations of the need to improve data and measurement methods, to better describe and help improve the health status of their Indigenous populations.
IGIHM includes Indigenous stakeholders and researchers.The key element of the group development was to connect people and ideas to focus on the unifying principle of the group: "The right of Indigenous peoples to count and be counted."Terms of reference to ensure that Indigenous populations have rights to all their data, can expect accurate data and contribute strengthening data collection, analysis, and interpretation at all levels.
Clark et al (2020) Commentary Canada An Inuit specific approach to discussing key issues relating to data, inequities, monitoring, and future research opportunities because of the COVID-19 pandemic.Historically, there have been no mechanism to identify Inuit in health systems, at the point of care and in administrative health data.Identification is essential to ensuring that universal coverage to response health services exists as well as the ability to monitor and evaluate health systems to identify gaps.The inclusion of an Inuit identifier to COVID-19 is an exceptional tool for public health officials and the ability for the Manitoba Inuit Association to access Inuit specific real time data.

IGIHM Recommendations:
1.There should be Indigenous identifiers on national census, birth and death certificates using the same definitions and nomenclature.The methods used to identify Indigenous people have been imposed by colonizers.Historically, Indian Agents were the ones to determine who was First Nations in relation to the Indian Act which continues to shape data collection today.There is no gold standard for Indigenous identification but that does not mean it should not be included in national censuses and in birth and death registrations.
There is an urgent need for national efforts to develops a standard process for Indigenous identification.
Historically Indigenous societies have been characterized and labelled by settler or colonizing groups.There is an urgent need for Indigenous groups to apply their own identification principles.

Minore, Kat & Hill (2009) Qualitative Canada
The capacity to document Indigenous health and service utilization varied significantly and depended upon existing provincial/territorial health data sets.
The inability to identify Indigenous Ontarians in the provincial health information system poses a significant challenge for those charged with planning and delivering healthcare.
The authors Identified seven key themes: (1) Data gap (2) Absence of ethnic identifier within administrative databases (3) barriers to database linkage (4) crossjurisdictional care (5) privacy and confidentiality concerns (6) access and governance issues and ( 7) technical or capacity requirements.

Recognition of OCAP ® and that
Indigenous people have a collective entitlement to information about their communities, which must be recognized.Indigenous stakeholders must be engaged throughout the process and procedures must be in place to ensure that the information is of high quality and kept in strict confidence.
Negotiation shall occur on a strict project-by-project basis to ensure that the information meets the needs of Indigenous communities, healthcare planners and providers.Indigenous people must be partners in data collection and governance to aid with the design, collection, and use of information.

Smylie & Firestone (2015)
Commentary Canada There are current Indigenous data quality challenges including misclassification errors and non-response bias systematically contributing to the underestimation of inequities in health determinants, health status and healthcare access among Indigenous and non-Indigenous people in Canada.This problem requires the revision of core health data sources to include relevant, consistent, and inclusive Indigenous self-identification.
Changes to healthcare data collection must be done in collaboration with Indigenous peoples and governing organizations.Indigenous right to govern and manage their own knowledge and information systems is inherent to self-determination.

Sarfati & Robson (2015) Commentary Australia
To monitor cancer incidence among Indigenous populations, Indigenous status needs to be recorded accurately in cancer registries and relevant population denominator data must be used to estimate incidence using a similar process that linked census-cancer data are used when ethnic origin is collected.Most countries rely on an unlinked method, but Indigenous status is measured differently in cancer and census data resulting in a numerator-denominator bias that most likely underestimates incidence of cancer among Indigenous people.There are issues of inconsistent data collection of ethnicity data because of assumptions of clerical staff.
Engagement with Indigenous people.In the original dataset, Indigenous status was unknown for most of the notifications but could have been ascribed with data linkage.The data linkage led to substantial increase in notification rates for Indigenous people.There is a range of methods available for linkage: "ever identified" methodology is the most used and most simple.The authors identify the SEEDS principles for Indigenous population health data linkages: 1) prioritize Indigenous Peoples' right to self-determination 2) make space for Indigenous Peoples to exercise sovereignty 3) adhere to ethical protocols 4) acknowledges and respects data stewardship and governance and 5) works to support reconciliation between Indigenous nations and settler states.

Australia
Self-identification in primary care Argue that physicians should "just ask" about Indigenous identity to ensure that the identifier question and explanation are conveyed consistently and appropriately.There are several themes that would influence decision to identify: previous racism; patient-doctor relationship; perception discussing identity would lengthen consultation times; assumed motives for asking; recognition of the culture and diversity of Indigenous Australians.The principal message was that the process for asking needs to be brief and simple.

census question that is more attentive to the complexities of Métis history is to link Métis self-identification to Métis organization attachment (i.e., Are you a member of the Métis Nation?)
Andersen (2016)Commentary Canada Health data for Métis people is poorly collected.The data that does exist relies on colonial logic and a misunderstanding of Métis identity.The lack of data makes it difficult to create meaningful health policy.Argues that a

typically held by the state. Ownership models such as formal data use agreement can be used to give communities control over analysis and dissemination.
6. Best practice linkage practices should be established to enhance vital statistics.7. Back casting for estimation of health trends needs to be based on sound underpinning assumptions of mortality trends to avoid circularity.Journal of Health Services Research & Policy Indigenous

The authors suggest solutions that can be implemented to strengthen the visibility of indigenous peoples, including, acknowledgment of the central importance of full engagement of indigenous peoples with all data- related processes, encouraging the use of indigenous identifiers in national and regional data sets and mitigation and/or careful assessment of biases inherent in cancer surveillance methods for indigenous peoples.
Engagement of Indigenous people.
(2)the need for meaningful single registry of a subpopulation of Indigenous people.The existing data deficiencies mask health inequality.Data can also demonstrate progress (i.e., improvement in smoking rates, vaccine coverage, cardiovascular deaths).

Indigenous health surveys, census, birth and mortality data collection Note:
Data governance column excluded for the remaining tables because most papers did not address it.

Due to the complexity and diversity of Indigenous identity, it is not possible to provide a definitive direction on the best way to define Indigenous populations to measure inequalities
. The proportion of Indigenous women changes significantly according to the Indigenous identification criterion used.Women who self-identify as Indigenous higher than those who identified that they speak an Indigenous language.From a

to migration and marriage, there is greater need to move from single group- race-based measure towards culturally based complex ethnicity measures
. In 2001, 7.9% gave more than one response to the question.For Indigenous people, those who recorded Maori, only 56% solely reported Maori identity.For

It is crucial to be able to easily identify Indigenous cancer patients in health services in the primary health care setting to monitor progress, improve and evaluate care.
It is also necessary for primary health care center staff to receive training and support to use the PCIS.The authors identified a knowledge gap between staff members and between primary care centers in their ability to use the PCIS.The accurate recording of Indigenous status varied among the participating emergency departments ranging from 45.5% to 87.2% with an average of 76% (n=8).In two of the ED the quality improvement program was associated with an increase in the accurate recording of Indigenous patients.

Indigenous patients will not identify as Indigenous if they perceive the environment is culturally unsafe and there is missing data on Aboriginal experiences in the ED due to this. There are several technical barriers for primary care providers to collect Indigenous identity data. It is argued that software providers make necessary changes to better capture the data.
This would be beneficial at numerous levels.For general practitioners this would mean better clinical care and billing.For patients this would result in improved access to preventative care, support services and more affordable medicine.At the population level, better data would improve service delivery.O'Loughlin et al(2020)

research found that it was more common for under-identification of Indigenous status rather than overidentification
. Sensitivity (self-identifying Indigenous people recorded by EDIS as being Indigenous) was 85.7%, specificity was 98.6% and overall accuracy was 97%.Comparison of (1) mail vs. telephone vs. clinic visit questionnaire and (2) compare front desk method vs. exam room.Collecting data during clinic visits had the highest response rate.The

front desk method yielded higher response rate than in the exam room. This labour would require one FTE for data entry.
Pulver et al., (2003)dating in providing the information with a high response rate.Pulver et al., (2003)Indigenous people giving birth at an urban obstetric hospital are under-represented in official statistics.Authors recommend developing sensitive but specific questions to allow patients to disclose their status.This study found that the participants did not mind being asked about Indigenous identity.

reasons are identified including: the lack of staff recognition that a problem exists, practice environments that do not promote identification, entrenched attitudes among staff and software limitations.
Scotney et al., (2010)or a policy change mandating Indigenous identification and recording in general practice.Scotney et al., (2010)