Comparing biomedical collections over time

The IHMHS is the second time the ABS has collected information on biomedical indicators. Biomedical indicators were first collected as part of the National Health Measures Survey (NHMS) 2011–12 and the National Aboriginal and Torres Strait Islander Health Measures Survey (NATSIHMS) 2012–13. In both collections, information was collected for the general population and for Aboriginal and Torres Strait Islander peoples. The ABS anticipates that users will have interest in comparing these collections to understand changes in biomedical information over time.

However, there are a range of factors that may impact the comparability of survey results across time. In some cases, it may be difficult to distinguish between changes due to methods, changes due to equipment updates, and/or changes due to biological trends. Users should be aware of these factors when comparing results between ABS biomedical collections. This page presents information on known changes in pathology test methodology and equipment between 2011–13 and 2022–24 for users looking to compare data across time. 

The approximately 10-year time period between the 2011–13 and 2022–24 biomedical collections significantly exceeds the life of most laboratory equipment. While the same pathology provider conducted the testing in 2011–13 and 2022–24 using a single laboratory, in some cases equipment for specific tests changed and/or the method of analysis changed in that time.

Verification and validation processes were conducted by the pathology provider each time a method and/or equipment was upgraded. Whilst these processes were robust, it was not always possible to directly verify current instrumentation and methods of analysis used in the 2022–24 against that used for in 2011–13 due to more than one change having been made. In addition to equipment changes, reagent lots may have changed, and calibrators and traceability may have improved. Internal procedures were in place for new batches of reagents and assays to check and adjust for results drift.

The precision of the methods (± x%) for each biomedical test needs to be considered when comparing results from old and new methods, in this case when interpreting observed changes in patient means between the two survey periods (i.e. whether observed changes are within the accepted precision range or not). New instruments tend to be more sensitive than older ones, with positive results at the lower end of the distribution that would not have been reported using older instruments.

To understand the impacts of changes in pathology methodology and equipment, the ABS commissioned expert advice on the equivalence of the individual tests in the 2022–24 with those undertaken in 2011–13. This work looked at the viability of using a range of various analytical tools to determine comparability of biomedical results over time. This included Quality Control material, Quality Assurance Program material and patient means data from the pathology provider.

Patient means data are routinely used by pathology companies to inform comparability of their results over time (e.g. mean or median fasting plasma glucose levels for patients attending pathology clinics). Patient means data is based on individual test results and cannot distinguish individual people. Patient means data were extracted by the pathology provider for each test from their database for comparable time periods (2011–13, 2022–24) and included patient numbers, means, medians, and standard deviations of the tested pathology by age and sex. Patients who attended the pathology clinic for the purposes of the ABS collections were excluded from patient means analysis.

The numbers of data points available for tests included in standard pathology tests for individuals (such as cholesterol and haemoglobin) were much higher than those for tests which are usually only reported on special request by a health professional for an individual or recorded through studies of population health (such as sodium, potassium, and iodine). Results indicated that patient means data were not normally distributed. As patient means data cannot distinguish individual people, it cannot identify whether a person has been tested multiple times for medical reasons in the period assessed or only once.

In determining the comparability of results over time for each test, the patient means data should not be given the same weight by the user as the instrument verification and validation results, because the populations from which they were drawn were different in the two time periods assessed.

Over time, there have been changes in clinical practice and government health policy that may influence the type of clients attending pathology clinics. For example, the HbA1c test for diabetes is now used by health professionals to diagnose diabetes as well as to assess longer term management of diabetes (WHO 2019; D’Emden et al. 2012; Australian Diabetes Society 2015).

The age at which diabetes screening occurs has been lowered since 2011–13. For the general population, screening should now be every 3 years starting at age 40, and for people with obesity or other risk factors, screening should be every 3 years, not limited by age (RACGP 2016, 2020; Bell et al. 2020). For Aboriginal and Torres Strait Islander peoples, the screening age was initially lowered from that for the general population based on evidence from previous ABS biomedical surveys, but a more recent recommendation from the Royal Australian College of General Practitioners was for annual diabetes screening from 18 years of age and from 10 years of age for children with one or more risk factors (Burrow and Ride 2016; RACGP 2016, 2024; NACCHO 2024).

Recommendations for requesting a vitamin D test have also changed since 2011–13. The vitamin D test is currently not routinely recommended for adults, children or healthy infants, and there are criteria for eligibility for coverage of the cost of the test by the Pharmaceutical Benefits Scheme for ‘at risk’ groups (RACGP 2024).

Changes to clinical practice can impact the patient population used to derive quality metrics like patient means, but the impact on these data sets is unknown. In some cases, for example HbA1c and serum ferritin, changes in patient means data between the two time periods assessed were different in magnitude and direction for different age and sex groups. This indicates a potential change in client population rather than changes in results due to instrument or method upgrades.

The tables below summarise changes in methods of analysis from 2011–13 to 2022–24, with comments on the equivalence of methods for consideration by users.

When comparing the 2011–13 and 2022–24 studies it is important to also consider impacts that changes to survey methodology may have on time series analysis. For more information, see the National Health Measures Survey methodology, 2022–24. Information on NATSIHMS 2022–24 will become available when first results are released.

Australian Diabetes Society (2015, updated 2023), Guidance concerning the use of glycated haemoglobin for the diagnosis of diabetes mellitus, Australian Diabetes Society, accessed 20/02/2025.

Bell K, Shaw JE, Maple-Brown L, Ferris W, Gray S, Murfet G, Flavel R, Maynard B, Ryrie H, Pritchard B, Freeman R, Gordan BA (2020), A position statement on screening and management of prediabetes in primary care in Australia, Diabetes Research and Clinical Practice, 164:108188, accessed 20/02/2025.

Burrow S, Ride K (2016), Review of diabetes among Aboriginal and Torres Strait Islander people, Australian Indigenous HealthInfoNet, accessed 20/02/2025.

d’Emden MC, Shaw JE, Colman PG, Colagiuri S, Twigg SM, Jones GRD, Goodall I, Schneider HG and Cheung NW (2012), The role of HbA1c in the diagnosis of diabetes mellitus in Australia, Medical Journal of Australia, 197(4):220-221, accessed 20/02/2025.

National Aboriginal Community Controlled Health Organisation (NACCHO) and The Royal Australian College of General Practitioners (RACGP) (2024), National guide to preventive healthcare for Aboriginal and Torres Strait Islander people: Recommendations. 4th edition., RACGP, accessed 20/02/2025.

Royal Australian College of General Practitioners (RACGP) and Diabetes Australia (2016), General practice management of type 2 diabetes: 2016–18 [PDF 3267 KB], RACGP, accessed 20/02/2025.

Royal Australian College of General Practitioners (RACGP) (2024), Management of type 2 diabetes: A handbook for general practice, RACGP, accessed 20/02/2025.

Royal Australian College of General Practitioners (RACGP) (2024), ‘Vitamin D testing’, First Do No Harm: a guide to choosing wisely in general practice, RACGP website, accessed 20/02/2025.

World Health Organization (WHO) (2019), Classification of diabetes mellitus, WHO, accessed 20/02/2025.