Vaccine Preventable Diseases and Vaccination Coverage in Aboriginal and Torres Strait Islander People, Australia, 2003 to 2006

Data quality and notes on interpreting data

Disclaimer: Produced by the National Centre for Immunisation Research and Surveillance of Vaccine Preventable Diseases and the Australian Institute of Health and Welfare on behalf of the Australian Government Department of Health and Ageing. Published as a supplement to the Communicable Diseases Intelligence journal Volume 32, June 2008.

Page last updated: 30 June 2008

Comparisons between the notification, hospitalisation and death databases should be made with caution as they differ in their purposes, reporting mechanisms and accuracy. To provide the most recent information available, and to account for the varied reporting formats, different time periods have been reviewed for each dataset. Due to variations in data quality, data from different jurisdictions have been included for notifications compared with hospitalisations and deaths, while data from all jurisdictions were used for vaccination coverage. As there were no unique identifying codes to link records for the same individual across databases, and because of differences in the accuracy of each database, it was not possible to analyse deaths and hospitalisations as a subset of notifications.

Comparing data from different collections is therefore problematic and should take account of the various factors outlined below.

Indigenous identification

The quality of Aboriginal and Torres Strait Islander health statistics depends on the accuracy of Indigenous population estimates and the level of completeness and accuracy of reporting achieved in the collection of Indigenous status for the condition of interest. Considerable work has been done in recent years on assessing and improving the quality of Aboriginal and Torres Strait Islander statistics in national, state and territory administrative data collections.2,8,17 More work is needed to improve the quality of the data, as large variations in quality exist between data collections, and, within the same data collections, there are variations between jurisdictions and over time.


Indigenous identification

The proportion of notifications lacking identification of Indigenous status were analysed by jurisdiction, year and disease. Adequate levels of completeness of Indigenous status identification between 2003 and 2006 were defined as at least 60% for a substantial majority of the diseases analysed. This level of completeness was achieved for New South Wales, the Northern Territory, South Australia, Victoria and Western Australia. After establishing that notification incidence estimates were not dominated by any one of these five jurisdictions of interest (data not shown), estimates are presented for the five jurisdictions combined. While the first report covering the period 2000 to 2002 excluded Victorian notification data, data completeness for this state between 2003 and 2006 substantially improved to be above the 60% threshold for all diseases except pertussis. Indigenous status was reported for approximately 50% of pertussis notification data for New South Wales and Victoria. Additional caution is therefore needed when analysing pertussis notification rates. The accuracy of Indigenous identification within the notification system has not been validated and therefore the possibility that some misreporting might have occurred should be considered in interpreting the data.

As a high proportion of influenza notifications lacked identification of Indigenous status in all states and territories, other than the Northern Territory, influenza was excluded from further analysis.

Other issues

A major limitation of the notification data is that, for most diseases, they represent only a proportion of the total cases occurring in the community, i.e. only those cases for which health care was sought and a diagnosis made, followed by a notification to health authorities.11 This proportion may vary between diseases and over time, with infections diagnosed by a laboratory test more likely to be notified. States and territories may have varying reporting requirements by medical practitioners, laboratories and hospitals. Under-reporting of notifiable diseases by doctors and from hospitals has been documented in Australia21 and this may vary between jurisdictions.22

Top of page


Indigenous identification

Aggregated hospital separation data for New South Wales, the Northern Territory, Queensland, South Australia and Western Australia are used in this report. These aggregate data cover hospital use for a majority (60%) of the Aboriginal and Torres Strait Islander Australian population. It should be noted that the data from these five jurisdictions do not necessarily represent the national picture, as their hospital experiences will not necessarily be representative of Aboriginal and Torres Strait Islander people living in the other jurisdictions.2 Jurisdictional differences in data quality, including the degree of Indigenous under-identification, should also be considered when interpreting the results.2

The analysis of hospitalisation rates over time should also be interpreted with caution, as hospitalisation rates for Aboriginal and Torres Strait Islander patients may be affected to a varying degree by improved identification over the period being analysed.2

Other issues

Recorded hospitalisations generally represent only the most severe end of the morbidity spectrum, so the ability of ICD codes to capture infectious disease-related morbidity varies between diseases. It must also be noted that the hospitalisation database contains a record for each admission, which means that there are separate records for each readmission or inter-hospital transfer. This is unlikely to have a major impact on the numbers reported for most diseases reviewed, as they are acute illnesses. For hospitalisations where the code of interest was not the principal diagnosis, the code of interest will have been recorded as a co-morbidity (additional or secondary diagnosis), the relative importance of which cannot be gauged.

Top of page


Indigenous identification

The accuracy of reporting Indigenous status on deaths has been previously evaluated by comparing the reported number with an expected, or predicted, number of Aboriginal and Torres Strait Islander deaths.8 Reporting was found to be acceptable for deaths in 1997–1999 in the Northern Territory, Queensland, South Australia and Western Australia. Following previous practice,8 reported deaths from these four jurisdictions only have been presented in this report. These combined rates may still underestimate Aboriginal and Torres Strait Islander death rates due to under-reporting, and the validity of the Indigenous status data has not been assessed. Mortality rates of Aboriginal and Torres Strait Islander people reported here may not be representative of mortality in the other four jurisdictions.

Other issues

Mortality data were analysed by year of registration rather than by year of death, thereby avoiding incomplete data for the latest available year. In recent years, less than 5% of deaths in a particular calendar year are registered in the subsequent year,23 with the bulk comprising that calendar year’s December deaths.

Vaccination coverage data

Indigenous status identification in the Australian Childhood Immunisation Register

In contrast with the previous report,1 in which ACIR vaccination coverage data were reported only from selected jurisdictions due to data quality issues with Indigenous status identification, data from all jurisdictions have been used in this report. Rank and Menzies17 showed that the reporting of Indigenous status in the ACIR had improved from 42% of the estimated national cohort of Aboriginal and Torres Strait Islander children aged 12–14 months in 2002 to 95% in 2006. With the exception of South Australia (where completeness remained at 72%–77%), Indigenous status identification was greater than 90% of expected levels in all jurisdictions by 2005, with the greatest improvement between 2003 and 2004 for most jurisdictions. This probably resulted from several initiatives during this period that improved recording of Indigenous status in the ACIR. These included promotional efforts for reporting of Indigenous status, commencement of regular transfer of demographic data from Medicare to ACIR, and commencement of transfer of data from some jurisdictional immunisation registers to ACIR.

The use of ACIR coverage estimates for Aboriginal and Torres Strait Islander children relies on the assumption that, in addition to the completeness of recording, the recorded Indigenous status is valid. While the validity of the data has not been formally assessed, two lines of evidence support it. First, previous analysis has found that children reported as Indigenous on the ACIR were more likely to have been reported as receiving vaccines recommended only for Aboriginal and Torres Strait Islander children1, and secondly, that the coverage estimates were similar to those obtained through a face-to-face survey supported by written records.18

Records with no data on Indigenous status have been classified as non-Indigenous for the purpose of this analysis.

General under-reporting in the Australian Childhood Immunisation Register

General limitations of data available from the ACIR must be considered when estimating vaccination coverage. A study conducted in 2001 found that the ACIR underestimated overall Australian immunisation coverage by 2.7% at 12 months of age and by 5% at 24 months of age.24 Coverage is calculated only for children registered on Medicare; however, data have shown that, by the age of 12 months, practically all Australian children have been registered with Medicare (Kathi Williams, Health Insurance Commission, personal communication, April 2004).

Validity of reported vaccination status in the National Aboriginal and Torres Strait Islander Health Survey

Vaccination status data were collected by patient recall only. For adult vaccinations, some respondents were unfamiliar with the term ‘pneumococcus’, and some were confused between the influenza and pneumococcal vaccinations. However, it was found this mainly applied to persons who had not had either vaccination and that those who had been vaccinated could generally report with certainty.19 The validity of self-reported vaccination status in elderly individuals has been shown to be higher for receipt of influenza vaccine within 12 months than for receipt of pneumococcal vaccine within the previous five years. Population coverage estimates for influenza vaccine using self-report have generally been slightly higher than provider-validated estimates (1%–10%).25 For pneumococcal vaccine the differences have been more variable, but self-reported estimates have generally been lower than validated estimates by around 10%.25 Validity has not been assessed in Indigenous or remote populations.

Top of page
Risk factor prevalence information from the 2004–05 National Aboriginal and Torres Strait Islander Health Survey

While data from the 2004–05 NATSIHS enable, to a certain extent, estimation of the prevalence of risk factors for which the influenza and pneumococcal vaccine were recommended in Aboriginal and Torres Strait Islander people aged 15–49 years,20 the survey was not primarily designed for collecting this information. Due to the questionnaire design, the classification of responses and the necessary data aggregation procedures undertaken by the ABS, conditions included in the prevalence estimates in this report do not exactly match the vaccination recommendations. Some specific conditions on the recommendation list could not be individually selected and included (e.g. cerebrospinal fluid leak, organ transplant recipients, chronic suppurative lung diseases), and chronicity and severity of some conditions (e.g. asthma) could not be differentiated. Overall, the selected list of conditions would probably provide the lower-end estimate of the true prevalence.

Aboriginal and Torres Strait Islander population estimates

Estimation of the size and age composition of the Aboriginal and Torres Strait Islander population is difficult. Increases in census counts of Aboriginal and Torres Strait Islander people between 1966 and 1996 are far greater than can be explained by simple demographic factors (births minus deaths).26 Other factors thought to be important include changes in: the propensity to identify as Aboriginal and Torres Strait Islander in the Census; the proportion of children with only one Aboriginal and Torres Strait Islander parent identified as Indigenous; and Census enumeration procedures.6,8,26

In this report, Australian Bureau of Statistics experimental estimates and projections low series, Aboriginal and Torres Strait Islander Australians population figures, based on 2001 Census data, are used.