SA-NT DataLink
Supporting health, social and economic research, education and policy in South Australia and the Northern Territory.

Frequently Asked Questions

What is data linkage?

Data linkage is the bringing together of two or more sets of population information so connections between the information can be examined and relationships determined. Data linkage is also known as “record linkage”.

Data linkage enables researchers to look at large volumes of data from various sources and make links and study variables between those data. 
An example of this may be the relationship between birth weight and Year 12 scores.  A researcher may want to examine whether the birth weight of a child has a long term impact on educational outcomes.  Using data linkage the researcher is able to look at this relationship over many years using many thousands of sets of de-identified data.

Data linkage enables extremely powerful research and Australia is in an excellent position to undertake strong data linkage projects due to its long term, comprehensive and stable population data sets.
 
Information used in data linkage projects is always “de-identified” meaning that no information that could identify any individual is ever supplied to researchers. 

 
Is Data Linkage new?

Data linkage has been used for many years by researchers primarily looking at the relationship between disease and lifestyle factors such as smoking and coronary disease.   It is thought that the discovery of the relationship between Thalidomide and congenital defects is an early (and renowned) example of data linkage.

Technological advances and the ability to link large, seemingly unconnected, databases now provides the means for health and social researchers to examine relationships that may not be clearly evident but nonetheless just as important.  For example a researcher may wish to investigate whether there is a relationship between early literacy scores and criminality?  By linking education and justice records, the researcher will be able to investigate this relationship.

A recent study undertaken in France linking data between educational levels and cancer mortality over a fifteen year period demonstrated a strong relationship between lower educational levels and certain types of cancer in men but not in women. The researchers commented:
“Recently, it was made possible to carry out a mortality follow-up of a representative sample of the French population, through record linkage between population census records and the national French cause-of-death file. This linkage possibility allowed for the first time in France analyses of social inequalities by cause of death that did not suffer from the previous methodological problems."

Menvill, G., et al (2005) Social inequalities and cancer mortality in France 1975-1990. Cancer causes and control.  June 2005 #13.

Technological advances are now also enabling researchers to work with data bases whose sizes are measured in terabytes, or trillions of bytes, thereby allowing for the comparison of very large sets of data.

 
Who uses data linkage?

Researchers and policy makers who have received approval from the appropriate Data Custodians and properly constituted and registered human research ethics committees can undertake statistical linkage projects for the purposes of research, policy development and evaluation, quality assurance and monitoring.

All output is population level and not individual based. The linked data are not used for direct service or clinical provision for individuals.

 
What data is included?

Administrative, clinical and service datasets are linked using a minimum number of variables. De-identified, linked variables are then made available to researchers by Data Custodians (the agency from which the data originates) for ethically approved statistical linkage projects.

 
How does SA NT DataLink protect your privacy?

The linkage methodology adopted by SA NT DataLink is based on the best practice Western Australia model. Data remains the property of the relevant custodians, and linkage “keys” are developed by SA NT DataLink based on probabilistic matching of demographic information provided from the relevant datasets.

SA NT DataLink does not hold identified data related to services provided or clinical information. Strict security procedures and ethical approval processes have been developed and agreed to by the Privacy Committee of South Australia. The SA NT DataLink linkage unit has high level security for every project using linkage keys. Staff of the linkage unit are all SA Government employees. All projects must firstly be approved by at least one NHMRC approved Human Research Ethics Committee.