This report presents an analysis of per-capita health expenditure by income for Indigenous and non-Indigenous Australians based on the utilisation of a range of health services as set out by questions asked in the 1995 NHS. This survey was conducted on a multi-stage area sample of private dwellings and a list sample of non-private dwellings (hotels, motels etc.). Hospitals, nursing homes and convalescent homes were excluded from the survey, as were prisons, reformatories and single quarters of military establishments. A base sample size approximating one-third of one per cent of the population was initially chosen. Inclusion of the Indigenous status question on the survey form yielded a total of 1100 Indigenous persons. To enhance the reliability of data for the Indigenous population an additional 1100 Indigenous respondents were sought. In doing this, the ABS used a sampling methodology which ensured that Indigenous respondents were representative of the population from which they were drawn. Thus, the total NHS sample included 2168 people who identified as being of Aboriginal or Torres Strait Islander origin.

It is important to note that, due to concerns about the quality of some of the responses from Indigenous participants who do not speak English at home, NHS estimates exclude Indigenous and non-Indigenous people living in those areas identified by the ABS as being sparsely settled, as in these areas non-English speakers predominate.^[2] In total, 539 records from survey participants in such areas were excluded, of which 461 were Indigenous. The final Indigenous sample for this report is based on 1536 Indigenous respondents in non-sparsely settled areas for whom there were valid data on household income. The weighted estimates for 1995 show that this Indigenous sample was representative of 82 per cent of the Australia-wide Indigenous population (ABS 2000c: 34).

The utilisation data from the NHS include ‘out-of-hospital visits to general practitioners or medical specialists’, ‘other health professionals’, ‘admitted hospital patients’, ‘non-admitted hospital patients’, ‘prescription medications’, and ‘over-the-counter medications’. In establishing expenditures based on these data, consideration was given to the age and sex patterns of utilisation which were found to vary significantly. It should be noted that changes to NHS questions on the utilisation of health services mean that the estimates of hospital utilisation are very unreliable for the Indigenous population. In the 1990 NHS, hospital utilisation was measured over the 12 months prior to the survey, whereas in the 1995 NHS the reference period was only the prior two weeks. Given the small size of the Indigenous sample and the low frequency of hospital visits, there are insufficient cases to ensure statistical reliability. Note that when converting these utilisation data to estimates of national expenditure, the amounts spent on each health service reported in the NHS were obtained from AIHW using a composite of administrative and supplementary survey data aggregated across different sources for varying geographic levels.