APPENDIX 1 PRINCIPLES ON THE USE OF DIRECT AGE-STANDARDISATION
PRINCIPLES ON THE USE OF DIRECT AGE-STANDARDISATION IN ADMINISTRATIVE DATA COLLECTIONS
The Australian Bureau of Statistics (ABS) hosted a workshop of key stakeholders on age-standardisation on 19 April 2011. The aim of the workshop was to discuss the best method of age-standardisation (direct or indirect) and to produce a clear set of guidelines specifically for the analysis and reporting of COAG "Closing the Gap" indicators. The workshop was attended by representatives from the ABS, Australian Institute of Health and Welfare (AIHW), Office of Aboriginal and Torres Strait Islander Health at the Department of Health and Ageing, COAG Reform Council, Productivity Commission and the Department of Families, Housing, Community Services and Indigenous Affairs. Workshop participants agreed that the direct method is the most preferred method of age-standardisation as it allows for valid comparisons of mortality rates between different study populations and across time.
However, there were a number of concerns about the direct method, in particular small cell sizes, that required consideration before adopting the method. A small team led by the AIHW, working closely with the ABS, developed a paper outlining a set of principles for using the direct age-standardisation method. For more information, refer to Australian Institute of Health and Welfare 2011. Principles on the use of direct age-standardisation in administrative data collections: for measuring the gap between Indigenous and non-Indigenous Australians. Cat. no. CSI 12. Canberra: AIHW. The principles included in this appendix are taken from the above paper published by the AIHW.
ABS has applied these principles to data cube Table 9: Deaths, Country of Birth, Australia, 2010, available for download from the ABS website. The principles will be applied to sub-state data, where appropriate, in future releases of this publication.
The guiding principles to be followed when undertaking the direct method of age-standardisation are summarised below.
Overarching principle
Before undertaking age-standardisation, analysts must investigate the data being used to understand the age-specific distributions of the populations being compared, and any limitations that may impact on the results.
Principle 1
The standard population used should be the Australian Estimated Resident Population as at 30 June 2001 from the 2001 Census until population estimates from the 2011 Census become available.
The population used as the denominator for the calculation of Aboriginal and Torres Strait Islander age-standardised rates should be series B of Experimental Estimates and Projections, Aboriginal and Torres Strait Islander Australians, 1991 to 2021 (cat. no. 3238.0) based on the 2006 Census until population estimates from the 2011 Census become available.
Principle 2
If the denominator is less than 30 in any one age group, then do not attempt to produce age-standardised rates.
Age groups may be collapsed to obtain a denominator of 30 or more (provided that this is in accordance with Principles 3 and 4).
Principle 3
If the total number of events in one population, e.g. deaths in the Aboriginal and Torres Strait Islander population, is less than 20, then do not attempt to produce age-standardised rates.
Combining several years of data, or aggregating jurisdictions should be considered to obtain a total of 20 or more events.
If this does not meet the purpose (i.e. data are required for time series or jurisdictional comparisons), or does not result in greater than 20 events, then other measures and contextual information should be reported instead of age-standardised rates which could include total number of events, crude rates, age-specific rates, age-specific rate ratios and median age at the time of the event.
Principle 4
Age-standardised rates should be calculated using the five year age groupings of 0-4 to 75+ (provided Principles 2 and 3 for denominator and numerator are met).
10 year age groups may be used to overcome small numbers (20 year age groups are too wide and should not be used).
Principle 5
Contextual information (most importantly age-specific rates and ratios) should be provided in addition to age-standardised rates when:
- the age-standardised rates and rate ratios lie largely outside the range of the age-specific rates and rate ratios
- the pattern of age-specific rates of the Aboriginal and Torres Strait Islander and non-Indigenous populations differ substantially (e.g. deaths from a certain cause concentrate on younger ages for the Aboriginal and Torres Strait Islander population while for non-Indigenous population they may occur at older ages)
- age-specific death rates depart from the assumption of a uniform increase in death with age (e.g. injury which peaks in the young adult to middle ages and certain cancers amenable to treatment for some age groups).
Principle 6
For conditions restricted to a specific age group (e.g. conditions originating in the perinatal period, Sudden Infant Death Syndrome), it is recommended to report the age-specific rate for the age group of interest instead of the age-standardised rate.
Notes regarding application of these principles
Analysts should apply these principles keeping in mind that a consistent time series will need to be maintained and valid comparisons ensured. The guidelines should not result in a widely different methodology being used for data being compared across time or across jurisdictions. Analysts should make decisions regarding the application of these principles based on an assessment of the data from the outset of their analyses.
These principles may need to be reviewed at some point in the future in order to take into account any changes, for example, in Aboriginal and Torres Strait Islander mortality and population estimates. A sensible time for such a review would be to coincide with a change of standard population (i.e. every 10 years).
The impact of a change in standard population on the resulting age-standardised rates will need to be assessed by the ABS and AIHW as soon as practical following the release of the 2011 Census population estimates.