What is Linkage?

Data linkage is the bringing together or linkage of records of an individual, household, business unit or other entity from either within or across a number of datasets. That is, records are compared and if a number of variables from each record are deemed to be the same then those records are considered to belong to the same person or entity.

The process of bringing together information from different sources paves the way for a broader range of questions to be answered than was otherwise possible. Through linkage it becomes possible to examine underlying relationships between various aspects of society or service delivery, thus improving our knowledge and understanding about a particular subject.

SA NT DataLink will use probabilistic linkage, a linkage methodology that involves the use of inexact variables that are common to both files (for example names, addresses, date of birth and sex). Mathematical probabilities are used to assign match weights for each variable that agrees or disagrees. Probabilistic linking is complex and sophisticated data linkage software is required in order to achieve high quality results.

SA NT DataLink will be linking data only for statistical, research and policy purposes. All research output is at the population level and not individual based. The linked data are not used for direct service or clinical provision for individuals.

The process and model used are based on the Western Australian model. In this model there is a clear separation between the custodial, linkage and research areas to minimise the risk of loss of an individual’s privacy or confidential information. The separations are:
 

• the Data Custodian is the only entity to hold all of the data, that is the personal, event and service or clinical data;
• the Linkage Unit only holds the variables required for record linkage, personal and potentially limited event data; and
   
• the researchers only ever obtain, de-identified data, with a project specific linkage key and must destroy the data after a given time. The use of project specific linkage keys ensures that data for individual projects cannot be linked together.
   

All data are secured physically and staffed by SA Health employees.
 

Preservation of individual’s privacy and protection of data

The model used by SA NT DataLink is recognised as best practice due largely to the separation of roles and the minimal and controlled use of identified data. One of the main features is that data custodians retain control of their data and how they are used.

Well defined approval processes for the addition of datasets to the Master Linkage File and for statistical linkage projects have been agreed with the Privacy Committee of South Australia. Appropriate ethics approvals are required for any project which will take into account methodologies and minimal cell size for reporting.

The Data are protected following the highest levels of security (Defence level standards similar to the ones used in the UniSA Department of Veterans Affairs project).