Diabetes is known to cause substantial illness and death in Australia. Knowing how many Australians are affected by diabetes informs decision makers who plan and allocate resources for health services. The continuing availability of time series data on the prevalence of diabetes in Australia is particularly important for planning and for disease monitoring.

There are several sources of diabetes prevalence data, each providing a different estimate of the number of cases of diabetes in Australia. Given this, the National Centre for Monitoring Diabetes has assessed data sources that provide national estimates of diagnosed diabetes prevalence. These are based on cases where a diagnosis of diabetes has been certified by a doctor, nurse or credentialed diabetes educator, or where an individual has reported having been told by a doctor or nurse that they have diabetes.


The national centre identified five national data sources that could be used to assess diagnosed diabetes prevalence: the Australian Diabetes and Obesity Lifestyle Study (AusDiab), the National Diabetes Services Scheme (NDSS) database, the National Health Survey (NHS), the Medicare Benefits Schedule (MBS) database and the Pharmaceutical Benefits Scheme (PBS) database.

Three criteria were used to describe and compare the estimates of diagnosed diabetes prevalence derived from these data sources: how much of the diagnosed diabetes population is captured (coverage), how current the data are (currency) and how often new data are made available for the purpose of time series analysis and disease monitoring (frequency).


  • The NHS and NDSS data are the best available sources for monitoring diagnosed diabetes prevalence in Australia, yielding prevalence rates of 3.6% in 2004-05.
  • Users of these estimates must be aware of coverage limitations in each of the preferred sources, the limitations of self-reported data in the NHS and the fact that estimates of diabetes subtypes have not been assessed.
  • Linking the NDSS database, the MBS database and the PBS database would assist in finding further cases of diabetes not counted in the NDSS data.