Data linkage
Data linkage is a process of linking together information from people held in different data collections. This linking can create a study dataset that contains information about a person, a group or population that is relevant to specific health conditions, treatments or outcomes.
For example, a data linkage may help to identify modifiable risk factors for birth outcomes or enable the comparison of outcomes for diseases such as cardiovascular disease between different population groups.
Data linkage can be used for epidemiological research, population health surveillance, health system performance monitoring, program evaluation and social research. It helps government provide a way for researchers to study health and social factors in populations.
The Epidemiology Section is a founding partner of the Centre for Health Record Linkage (CHeReL) at the NSW Ministry of Health and facilitates the provision of data to researchers and government agencies. CHeReL provides the technical infrastructure to enable data linkage services.
The following health datasets have been integrated into CHeReL and are available for linkage studies:
- ACT Admitted Patient Collection
- ACT Cancer Registry
- ACT Emergency Department Information System
- ACT Maternal Perinatal Data Collection
- ACT Notifiable Diseases Management System
- ACT Births, Deaths and Marriages birth registrations
- ACT Births, Deaths and Marriages death registrations
- Australian Early Development Census
- ACT Kindergarten Health Check
- ACT Australia and New Zealand Dialysis and Transplant Registry
- ACT Ambulance Service data.
How data linkage works
Data custodians separate the content of their datasets into demographic and clinical files and retain the clinical file on-premises.
Data Managers in the Epidemiology Section provide the demographic datasets to CHeReL with an encrypted source record number and the demographic details for each record in their dataset. Under the current ACT model, no clinical information is provided to CHeReL. Custodians retain all clinical data on-premises.
Once CHeReL receives these demographic files, a probabilistic linkage algorithm calculates a score for each field that indicates for any pair of records, how likely it is that they both refer to the same person.
Pairs of records above a certain score are considered to be a link. Pairs below a certain score are considered to be a non-link. Scores in between (in other words, possible matches) often require clerical review to be resolved.
CHeReL assigns a Person ID to records that are estimated to belong to the same person. The Person ID and associated source record numbers create the basic foundation of the Master Linkage Key. The Master Linkage Key thereafter provides a ‘pointer’ to records for a given person in different datasets. The Person ID never leaves CHeReL.
When a researcher is approved to perform a data linkage study, CHeReL assigns a project-specific person number (PPN) for each person in the linked dataset.
CHeReL sends the custodian a list of PPNs for each person in the study, along with the encrypted record numbers.
Now that the PPNs and the encrypted record IDs are with the custodian, the data custodian decrypts the source record number and merges the PPN with the clinical variables approved for use in the project.
The source record ID is removed and the researcher is provided with the PPN and the clinical information.
Using linked data
Data linkage is conducted on a project-by-project basis, with each project requiring the approval of relevant Human Research Ethics Committees and data custodians.
The following list provides an overview of the process of accessing linked data from the Data Linkage Team at ACT Health.
- Plan your study
- Contact the CHeReL to discuss your project
- Complete the CHeReL Combined Protocol and Application for Data form
- Complete the required ethics approval forms
- Provide document to the CHeReL for a technical feasibility review
- Obtain ethics approvals from ACT Health Human Research Ethics Committee
- Request data custodian approval via the online data linkage request form
- The ACT Health Data Linkage Team will coordinate approval from data custodians and notify the researcher and the CHeReL of approval
- CHeReL will process the data linkage
- The ACT Health Data Linkage Team will extract and release the linked data to the researcher ensuring that signed ACT Health Conditions of Release are current and all appropriate approvals are in place
- Data analysis is conducted with findings provided to the ACT Health Data Linkage Team before release or publication of results.
For more information about data linkage in the ACT, go to:
Contact
To contact the ACT Health Data Linkage Team, email HealthInfo@act.gov.au.