This work is copyright. Apart from any use as permitted under the Copyright Act 1968, no part may be reproduced without prior written permission from the Australian Institute of Health and Welfare. Requests and enquiries concerning reproduction and rights should be directed to the Head, Media and Communications Unit, Australian Institute of Health and Welfare, GPO Box 570, Canberra ACT 2601.
A complete list of the Institute’s publications is available from the Institute’s website .
ISBN 978 1 74024 902 7
Australian Institute of Health and Welfare 2009. International Group for Indigenous Health Measurement, Canberra 2006. Cat. no. IHW 26.Canberra: AIHW.
Any enquiries about or comments on this publication should be directed to:
Social and Indigenous Group
Australian Institute of Health and Welfare
GPO Box 570
Canberra ACT 2601
Phone: (02) 6244 1146
Published by the Australian Institute of Health and Welfare
Executive summary vi
Part A: Meeting proceedings 8
A1 International collaboration on indigenous health measurement—Canberra 2006 meeting 8
A2 Agenda 13
A3Presentation abstracts 16
A4Meeting participants 24
5Biographies of presenters 29
A6Media Release by the AIHW 36
Part B: Data on indigenous health 37
B1 Census 37
B2Vital statistics 41
B3Administrative data collections 47
B4Health surveillance 49
B6Comparisons of indigenous health data 53
Appendix 1:List of AIHW publications provided to meeting participants 56
List of tables 61
A meeting of the International Group for Indigenous Health Measurement was held in Canberra on 28–30 November 2006. The meeting brought together a range of people who are involved in this arena including people from indigenous organisations, national statistical agencies, departments of health, and research institution from Australia, Canada, New Zealand, and the United States. The Australian Institute of Health and Welfare (AIHW) was honoured to host this meeting, which was co-sponsored by the Office for Aboriginal and Torres Strait Islander Health.
The purpose of the meeting was to further international collaboration to tackle health measurement issues for indigenous populations, building on the work that began at the inaugural meeting in Vancouver, Canada, in 2005.
The meeting highlighted the continued disparities between the health and wellbeing of the indigenous and non-indigenous populations in the four countries. The collection of high-quality data is essential to policy makers as it helps in monitoring and evaluating programs aimed at reducing these disparities. The value of the International Group for Indigenous Health Measurement was confirmed, and the commitment of the agencies represented to improving health information about indigenous people, for the benefit of indigenous people, was highlighted.
I would like to acknowledge the valuable contributions made by the participants during this meeting. Participants presented information on the work they are doing in their own countries to overcome current data deficiencies and to ensure that indigenous health is measured in appropriate ways and that all indigenous people are included in data collections. The importance of national level frameworks for data collection, health improvement and accountability underpinned many of the discussions.
I hope this report of the meeting is useful to participants and others with interest in indigenous health measurement issues.
The main author of this report is Fadwa Al Yaman. Other people who have contributed to various Parts of this report include Deanna Pagnini, Ilona Papajcsik, Michelle Gourley and Ann Darcy.
The author would like to also acknowledge the contribution of the various country coordinators on the International Group for Indigenous Health Measurement for collating their in-country comments on the report. These include Sam Notzon, National Center for Health Statistics, Paula Searle and Natalie Paki Paki, Ministry of Health, New Zealand and Rene Dion and Don Fiddler, Health Canada. Thanks are also due to Lisa Jackson Pulver (University of New South Wales), Ian Ring (University of Wollongong), Anneke Outred, Australian Bureau of Statistics and Kirrily Harrison, Office for Aboriginal and Torres Strait Islander Health, Department of Health and Ageing for their valuable comments.
The meeting was jointly funded by the Australian Institute of Health and Welfare and the Office for Aboriginal and Torres Strait Torres Strait Islander Health, Department of Health and Ageing.
The AIHW received financial support from Office for Aboriginal and Torres Strait Islander Health, Department of Health and Ageing for the production of this report.
This publication provides an overview of the second meeting of the International Group for Indigenous Health Measurement held in Canberra, Australia on 28–30 November 2006. The International Group for Indigenous Health Measurement includes a range of people who are involved in this arena, including people from indigenous organisations, national statistical agencies, departments of health, and research institutions from Australia, Canada, New Zealand, and the United States who are committed to working together to ensure the highest quality collection, dissemination, and use of data related to indigenous health.
The Introduction to this report describes the development of the International Group for Indigenous Health Measurement, then presents an overview of indigenous population statistics for each of the four countries, highlighting difficulties in data definitions and data collection processes, indigenous/non-indigenous disparities in infant mortality and life expectancy, and the differences in data governance and input into the collection and use of indigenous health data.
The proceedings from the Canberra meeting are included in Part A of the report. A summary of the meeting is presented, and the themes from the 34 presentations are organised into nine categories: Who counts whom?, capacity building, information governance issues, importance of collaborative efforts, the importance of the community seeing benefits from the data, the importance of cultural factors/holistic approach to well-being, data issues, international collaboration, and the way forward. The meeting agenda is included, along with abstracts from all the presentations, a list of meeting participants, and biographies of the presenters.
Part B goes beyond the meeting summary to present detailed information on data collection processes and data quality issues related to indigenous health measures in each of the four countries. The report includes information regarding the identification of indigenous people/households in the Census, vital statistics data (births and deaths), administrative data, health surveillance measures, and health related surveys, then ends with a comparative overview of indigenous data which includes developments in improving indigenous coverage and data quality.
The report continually highlights the necessity of collecting high-quality data of relevance to both policy organisations and indigenous people for reducing the health disparities between indigenous and non-indigenous people in all four countries.
Indigenous and non-indigenous statisticians, researchers and organisations in numerous countries have been working to improve definitions, collection methods and uses of data on indigenous health, recognising the benefits of both high-quality data and the benefits of collaborating with colleagues in countries with similar issues. High-quality data are necessary to measure the health of indigenous people and to prioritise health needs. These data are fundamental to understanding the factors which underlie health disparities between indigenous and non-indigenous people. High-quality data can also guide public debate and ensure that policies and programs are evidence-based. Without high-quality data it is difficult to evaluate the effectiveness of health programs.
In 2004, discussions were held between health professionals in Australia and the United States on the establishment of a formal structure for working on indigenous health measurement issues. Canada and New Zealand subsequently joined the discussions, which led to the formation of the International Group for Indigenous Health Measurement (International Measurement Group).
At its inaugural meeting in Vancouver, Canada, in 2005, the International Measurement Group agreed that it would consist of participants from Australia, Canada, New Zealand and the United States, and that it would have two main purposes (AIHW 2006:11):
1. To improve the collection, analysis, interpretation and dissemination of information useful for improving the health of indigenous populations.
2. To develop an international network on indigenous health measurement that enables meaningful comparisons, exchange, mutual learning and collaborative projects that inform national policy making oriented to health gain.
The International Measurement Group considered the principles that should govern its development and operation, and agreed that indigenous people had a right to be counted, that indigenous people should be both leaders and participants (and thus needed increased capacity to assemble, analyse and use data), and that all data collection, analysis and dissemination should be linked to improvements in indigenous health.
The second meeting of the International Measurement Group was held in Canberra, Australia, in 2006. This report provides an account of the 2006 meeting in Part A, including the meeting agenda, abstracts, list of participants, and biographies of the presenters. Part B provides additional information on data related to indigenous health in Australia, New Zealand, Canada, and the United States, including census data, vital statistics, administrative data collections, health surveillance and health surveys. This introduction provides an overview of indigenous population statistics, including key indicators and issues related to data governance.
Overview of indigenous population statistics
One of the most important factors in collecting high-quality data on indigenous health is the ability to define and identify the indigenous population in a consistent and meaningful way. Definitions are often determined by why the data are being collected. For example, if legal entitlements are linked to indigenous status, the question of ancestry or formal registration status with the government is likely to be used. For broader issues of health and wellbeing, the more socially meaningful measure of self-identification may be used.
Table 1 provides a summary of the size and composition of the indigenous populations in Australia, Canada, United States and New Zealand (the four countries constituting the International Measurement Group), along with relevant treaties or legal status definitions which affect the provision of health care or the collection of health-related statistics.
1,172,790 people identified themselves as aboriginal(b)
First Nations (60%)
Multiple and other responses (3%)
Treaty status is given to those of First Nations heritage who have registered with the federal government. Health Canada provides free health care to the First Nations and Inuit populations who live on reserves and/or Inuit communities. The Métis are not eligible for treaty status or free health care.
New Zealand (NZ)
565,329 (Maori ethnic group)(c)
643,977 (Maori descent)(c)
The Treaty of Waitangi (1840) established British control while setting out Maori rights. The Treaty is integrated into health policy, with Maori people given the rights to partnership, participation and protection in health-related policies.
United States (US)
American Indian or Alaska Native (AI/AN) only (60%)
AI/AN plus another race (40%)
562 federally recognised tribes exist as sovereign entities. The US government is obligated to provide free health care to federally recognised American Indians and Alaska Natives.
(a) Estimated resident population—30 June 2006 preliminary, based on the 2006 Census and post-enumeration survey (ABS 2007).
(b) 2006 Census count (Statistics Canada 2008a).
(c) 2006 New Zealand Census.
(d) US Census Bureau 2008.
Among the four countries, the United States has the largest indigenous population, with 4.5 million people, but is the smallest proportionally at 1.5% of the total US population. Canada has the second largest indigenous population, with approximately 1.2 million indigenous people who constitute nearly 4% of the population. Over half a million ethnic Maori make up almost 15% of the total New Zealand population and in Australia just over half a million Aboriginal and Torres Strait Islanders make up 2.5% of the total Australian population.
The size of the Australian Indigenous population shown in Table 1 is the estimated resident population on 30 June 2006. To derive this number, the 2006 Census count was adjusted for Australian Indigenous people missed or counted more than once by the Census, those temporarily overseas and those who did not respond to the question. A post-Census enumeration survey was undertaken to estimate the sizes of these adjustments.
The size of the Canadian aboriginal population shown in Table 1 is the adjusted count from the 2006 Census. Respondents self-identified as North American Indian (First Nations people), Métis or Inuit, or with more than one group. This count was adjusted by including Registered Indians or members of an Indian Band or First Nation who did not identify as aboriginal.
Two counts from the 2006 Census have been presented in Table 1 as representing the size of the Maori population: the self-identified Maori ethnic group and those who have Maori ancestry. The latter number is used for the purposes of Treaty of Waitangi claims and other rights and entitlements.
The size of the American Indian/Alaska Native population in the United States shown in Table 1 is the modified race count as at 1 July 2006, based on the 2000 Census. In the United States the census is conducted every 10 years, the most recent in 2000. The original race data from the 2000 Census has been modified to reallocate those in the ‘some other race’ category, and adjusted for births, deaths, net international migration and net movement of US Armed Forces overseas. Indigenous Hawaiians are included under the classification Native Hawaiian and other Pacific Islander.
Significant health disparities exist between indigenous and non-indigenous populations in Australia, Canada, New Zealand and the United States. In all four countries indigenous people have lower life expectancies, higher rates of chronic and preventable illnesses, poorer self-reported health and higher likelihood of hospitalisation (Bramley et al. 2004; Freemantle et al. 2007; ABS & AIHW 2008). Relevant and accurate data are necessary to shape programs aimed at reducing health inequalities. Current data on indigenous health and wellbeing in the four countries have problems of coverage and quality and this undermines the usefulness of the data for good policy decision making.
Infant mortality rates and life expectancy at birth are two of the key measures of health status for which data are available on indigenous persons in Australia, Canada, the United States and New Zealand. A summary of this information is presented below.
The infant mortality rate is defined as the number of infant deaths in the first year of life expressed as a proportion of 1,000 live births. Table 2 presents the indigenous and non-indigenous infant mortality rates in the four countries. Indigenous infant mortality rates in all four countries are higher than infant mortality rates for the non-indigenous populations.
The coverage and quality of indigenous birth and death data determine the accuracy of infant mortality rates. The lack of a mandatory aboriginal identifier on Canadian birth and death registrations means that the aboriginal infant mortality rate cannot be accurately estimated. In Australia the coverage of Australian Indigenous birth data is higher than the coverage of Australian Indigenous death data, so the Australian Indigenous infant mortality rate is likely to be underestimated. In both New Zealand and the United States, birth and death registrations are matched, so indigenous infant mortality rates can be estimated with greater accuracy.
Table 2: Infant mortality
Indigenous and non-indigenous infant mortality
Coverage and data quality issues
The Australian Indigenous infant mortality rate for the period 2002 to 2004 is estimated to be 11.5 per 1,000 live births, using mortality data from the four states and territories that have adequate levels of Indigenous identification in their mortality data collections. This compares with a non-Indigenous rate of 4.1 per 1,000 live births for the same states/territories over the same period (AIHW 2007a:311).
Australian Indigenous infant deaths and births are both underestimated; however, Indigenous identification in birth data is higher than in death data, so the Indigenous infant mortality rate is likely to be an underestimate (AIHW 2007a:322). Australian estimates could be improved by linking birth and death data, as occurs in the United States and New Zealand, as well as by linking birth registrations with perinatal data.
The infant mortality rate for First Nations peoples in 2000 was estimated to be 6.4 per 1,000 live births, using vital registrations data from the four western provinces that collect ethnicity data, combined with data collected from nursing stations for the rest of the country. The comparable rate for all births in Canada is 5.2 per 1,000 live births (Treasury Board of Canada Secretariat 2005)
Analyses of mortality data for regions with high percentages of aboriginal inhabitants have estimated higher infant mortality rates. For example, First Nations infant mortality in 1999 is estimated to have been 8 per 1,000 live births, as compared with the Canadian average of 5.5 per 1,000 live births (Kermode-Scott 2005).
The lack of a mandatory question on birth and death registration forms relating to aboriginal status means that infant mortality rates cannot be accurately estimated for the First Nations, Inuit or Métis populations.
For the period 2000 to 2004, the Maori infant mortality rate was 8.1 deaths per 1,000 live births, as compared with the non-Maori rate of 5.0 deaths per 1,000 live births (Robson and Harris 2007:45)
Because infant mortality rates are calculated from matched birth and death registrations, the Maori rate is considered to be relatively accurate.
Based on analysis of the linked birth and infant death data sets for the period 1995 to 2003, the AI/AN infant mortality rate was 8.7 per 1,000 live births, as compared with 6.8 for the whole of the United States (National Center for Health Statistics 2007:160; National Center for Health Statistics 2006:5).
The National Center for Health Statistics matches birth and death records and uses the race recorded on the birth certificate to calculate race-specific mortality rates (National Center for Health Statistics 2006:5). Once the 2003 amendments to the birth certificate are fully implemented in all states, the classification of race in birth data should be relatively accurate, which will result in relatively accurate estimates of AI/AN infant mortality, both alone and in combination with other races.
Life expectancy at birth
Life expectancy estimates are an important aspect of assessing indigenous disadvantage. Indigenous life expectancy at birth in all four countries is reported as being lower than the life expectancy at birth for the non-indigenous population. The coverage and quality of death data and census data determine the accuracy of the calculation of life expectancy at birth and all four countries have data-quality issues which affect the accuracy of their estimates.
In Australia, a number of indirect methods have been used to estimate Australian Indigenous life expectancy at birth. All of these methods rely on assumptions and expert opinions. There have been significant uncertainties attached to the level of life expectancy estimates for Aboriginal and Torres Strait Islander peoples in the past, and although there has been significant progress in terms of using better methods to estimate life expectancy, there continues to be a level of uncertainty around the estimate of life expectancy for Indigenous Australians.
The latest available experimental estimates of Australian Indigenous life expectancy at birth are for the period 1996–2001 and are 59 years for males and 65 years for females (as compared with 77 years for all males and 82 years for all females for the period 1998–2000). The difference is approximately 17 years for both sexes.
Because of coverage and quality issues associated with Australian Indigenous births, deaths and population data that are summarised in Tables B2.1 and B2.2, a number of indirect methods have been used to estimate the life expectancy of Indigenous Australians. According to the ABS (2008), these indirect methods are no longer appropriate and direct methods are recommended for future calculations.
In 2001: First Nations men 70.4 years, all Canadian men 77.1 years; FN women 75.5 years, all Canadian women 82.2 years (Treasury Board of Canada Secretariat 2005).
Inuit-inhabited areas (using death data for the period 1999–2002): Men 64.4 years, women 69.8 years (Statistics Canada 2008b).
Because of lack of aboriginal identifiers on death registrations, standard data sources and methods cannot be used to estimate basic health indicators. Instead a geographic-based approach is often used to estimate life expectancy for the entire population of areas having a high percentage of aboriginal inhabitants.
For the period 2000–2002, females of Maori ethnicity 73.2 years, non-Maori females 81.9 years; males of Maori ethnicity 69.0 years, non-Maori males 77.2 years (Statistics New Zealand 2008b)
Every 5 years Statistics New Zealand produces complete period life tables using average mortality rates for three successive years centred on a census year. Before 2000, methods such as ‘ever Maori’ were used to resolve the undercount in Maori deaths. However, since 2000, the quality of Maori death data has improved, and no undercount is now apparent (Robson and Harris 2007).
Life expectancy tables by race published by the US Government through the National Center for Health Statistics relate only to the white and black populations. No tables are available for the AI/AN population. The Indian Health Service has published estimates of life expectancy for the period 1999–2001, which state that life expectancy for the AI/AN population is 2.4 years less than for the all-race population of the United States (74.5 years, as compared with 76.9 years) (Indian Health Service 2006).
The undercounting of AI/AN deaths referred to in Table B2.2 means that AI/AN death rates are underestimated, hence life expectancy is overestimated for the AI/AN population. Despite the development of adjustment factors to correct the under-identification of AI/AN deaths, life expectancy estimates, which are based on age-specific death rates, should be interpreted with caution.
The Australian Bureau of Statistics (ABS) has recently published a discussion paper entitled Assessment of methods for developing life tables for Aboriginal and Torres Strait Islander Australians, 2006 (ABS 2008),which presents an assessment of various methods for adjusting incomplete Aboriginal and Torres Strait Islander death registration data for use with the latest available estimates of the Australian Indigenous population so as to compile Australian Indigenous life expectancy estimates. The key findings of this paper were that the indirect methods previously used by the ABS to estimate Indigenous Australian life expectancy are no longer considered appropriate in the Australian context; and that currently the best option for adjusting under-coverage of Australian Indigenous deaths is to use more direct methods such as that used in the ABS Indigenous Mortality Quality Study (part of the Census Data Enhancement project) which linked 2006 death records to Census records. Findings from this study suggest that the coverage of Australian Indigenous deaths in death registration data is higher than previously estimated.
In New Zealand the quality of Maori death data has improved since 2000, with no undercount now apparent. In the United States, the number of indigenous deaths is underestimated, which means that indigenous life expectancy at birth is overestimated. The lack of a mandatory aboriginal identifier on Canadian death registrations and the lack of a question on aboriginality on the majority of census forms means that aboriginal life expectancy at birth cannot be accurately estimated.
An overview of the agencies and committees which have a strategic role in relation to indigenous data in the four countries is presented in Table 4.
Table 4: National agencies and committees which provide strategic advice in relation to indigenous data and information
The National Advisory Group on Aboriginal and Torres Strait Islander Heath Information and Data (NAGATSIHID)
NAGATSIHID was established to:
• provide broad strategic advice to the Australian Health Ministers’ Advisory Council and its National E-Health Information Principal Committee on ways of improving the quality and availability of data and information on Aboriginal and Torres Strait Islander health and health service delivery
• draw together the range of Australian Indigenous health information activities into a coordinated and strategic process.
Membership comprises the Australian Bureau of Statistics, Australian Institute of Health and Welfare, Department of Health and Ageing, Department of Families, Housing, Community Services and Indigenous Affairs, Australian Institute of Aboriginal and Torres Strait Islander Studies, Torres Strait Regional Authority, National Health Information Statistics and Standards Committee, National Aboriginal and Torres Strait Islander Health Officials Network, National Aboriginal Community Controlled Health Organisation, Indigenous advisors on Aboriginal and Torres Strait Islander health and welfare and an epidemiologist with expertise in Indigenous health issues. A key priority of NAGATSIHID is to maintain an Australian Indigenous majority at meetings.
NAGATSIHID provides a forum for all key stakeholders involved in Indigenous Australians health policy and data to work together towards improving the data and reporting.
First Nations Statistical Institute (FNSI)
The Institute was set up by the First Nations Fiscal and Statistical Management Act 2002 to provide statistical information and analysis of key issues related to members of First Nations and other aboriginal groups, ensure data quality and compatibility, work in cooperation with Statistics Canada, and build statistical capacity within First Nations governments. The Institute is still in the developmental phase and is not yet fully operational.
First Nations Information Governance Committee (FNIGC)
FNIGC operates under the auspices of the Assembly of First Nations and is the responsible authority for the Regional Longitudinal Health Survey.
Maori Statistics Forum
The Forum provides Statistics New Zealand with expert external input into critical statistical issues in relation to Maori. A Maori statistics framework is currently being developed.
American Indian and Alaska Native Advisory Committee
The Committee advises the US Census Bureau on ways to increase the participation of AI/AN people in the census, facilitates the Bureau’s outreach to the AI/AN population during the planning and implementation of the census and the American Community Survey, and advises on ways census data can be disseminated for maximum usefulness to the AI/AN populations and other users. The nine-member Committee comprises representatives from the public and private sectors, community-based organisations, academic institutions and other individuals as appropriate from the public at large.
Tribal Consultation Advisory Committee (TCAC)
In 2004 the Department of Health and Human Services (HHS) implemented a Tribal Consultation Policy that applies to all HHS divisions. In compliance with the Policy, the Centers for Disease Control and Prevention (CDC), which is a division of HHS, established the Tribal Consultation Advisory Committee. The Committee advises the CDC Director and the Administrator of the Agency for Toxic Substances and Disease Registry on policy issues and broad strategies that may affect AI/AN peoples.