The Centre for Data Linkage
The Centre for Data linkage (CDL) is a nationally focused unit capable of securely linking and managing health information across Australia to support research. The CDL plays a key role in:
- developing and evaluating leading-edge data linkage technologies to support the network
- providing access to LinXmart a proprietary secure, fast and scalable data linkage software, with sophisticated data management features that handle growing datasets and dynamic linkages over time
- providing access to a Secure eResearch Platform (SeRP@Curtin) is the complete customisable solution for data sharing, linkage and analysis in a safe, secure and controlled environment
- providing a secure file transfer service (SUFEX) for the safe transmission of data files for linkage purposes and linkage studies
- providing expert advice on linkage methods and project design to national and international partners.
The CDL is an ISO27001 accredited facility providing a secure and efficient data linkage environment for accurate, large-scale linkage of datasets. The CDL currently supports several major projects including:
- WA Government Social Investment Data Resource a linked, administrative database containing de‑identified information from key government agencies
- The international platform, SeRP@Curtin which incorporates the data linkage capabilities provided by specialist software, LinXmart, developed by the CDL
- The Health Studies Australian National Data Asset (HeSANDA) program creating a national sharable platform using the data outputs of health research projects
The CDL adheres to all applicable legislation, standards and codes of practice relating to data collection, processing, linkage, sharing, publication and retention as well as actively engaging and implement common standards and promote good practice.
Our mission is to establish, support and promote the secure linkage of data through efficient, effective and innovative systems. To provide Australian researchers with state-of-the-art health data linkage facilities and services by:
- operating reliable and efficient data linkage and analysis services for researchers
- facilitating innovative research and promote infrastructure development
- expediting linked analysis to improve health services and outcomes.
We strive to provide a world-class data linkage and data analytics service built around an internationally recognized and inspired team providing cost-effective and reliable services, exploring opportunities for research and development, providing high-quality linked research information for government and researchers, and supporting high-quality research.
We will work collaboratively with other groups that have data linkage expertise to facilitate linkage between datasets. Our aim is to:
- make efficient , effective and innovative use of our resources
- be receptive and responsive to our stakeholder’s needs
- engage our stakeholders in meaningful consultation and communication
- identify and share good practice.
CDL is a Commonwealth Authorised Data Linkage Unit with IS027001 certification. It provides state-of-the-art linkage products and services that are secure, fast and scalable, with data management functionality that can help clients and collaborators to manage growing datasets and complexity dynamic linkages over time.
A team of data analysts and linkage experts undertake data linkage at the CDL. The Centre uses LinXmart software developed at Curtin University to perform data linkage. The software uses deterministic and probabilistic methods to determine if records in different data collections belong to the same person. The software takes into account missing information and errors in data (e.g. typographical errors) and produces highly accurate match results. Privacy-preserving record linkage is also available.
The CDL operates SUFEX – a secure online file transfer service that allows users to securely send and receive large data files from anywhere, at any time. Through a standard software browser, registered users can send and receive data files, track recipients, delete or withdraw files, and view file reports.
For more on SUFEX, contact email@example.com.
Research at the CDL and across the Data Analytics Hub is enabled through an advanced analytic platform. The CDL hosts a range of research facilities including systems for the secure curation and management of research data, authorised access mechanisms, and secure computing. Additional computational resources are available through Curtin’s Institute for Computation.
Our staff engage in high quality research across the following research domains:
- innovative linkage methods, including the implementation of privacy-preserving data linkage techniques
- linkage quality: measurement of linkage quality, developing methods for maximising linkage quality, investigations of the factors affecting linkage quality and research outcomes based on linked-data
- epidemiological studies of health and well-being, using linked-data
- crime and social justice research.
In recent years, the CDL has pioneered the implementation and use of privacy-preserving record linkage (PPRL) in operational settings.
Legal, administrative and technical issues have impacted on the ability of linkage units (nationally and internationally) to undertake linkage of particular datasets. New record linkage techniques, collectively referred to a privacy-preserving record linkage, significantly reduce privacy risks as they operate on de-identified information and do not require the release of personal identifiers.
Curtin University has developed software that implements Privacy Preserving Record Linkage (PPRL) for use in operational record linkage settings. PPRL involves carrying out record linkage activities using encrypted personally identifying information. Data is encrypted in a way which allows linkage to be carried out, without personal identifying information being disclosed. Data is encrypted by the data custodian (using specially provided software) before being sent to the linkage team, who carries out the linkage.
The Bloom filter method for PPRL provides strong protection as the encryption process is irreversible and the output is so ‘distorted’ that recognition of an individual is not possible (i.e. the identity of an individual cannot be reasonably ascertained from the information disclosed by a data custodian).
Ensuring the accuracy of linkage results is a priority for all linkage units (highlighted at the 2017 IPDLN Conference at Swansea University). The development of tools, measures and methods to maximise and maintain linkage quality ensure high standards and consistency of results in the data linkage field.
The CDL has developed automated checking algorithms and related tools to reduce the manual clerical burden on large-scale linkage systems. In one tool a reviewer expert’s knowledge is acquired based on current identifiers and is added incrementally i.e. the reviewer creates a rule for classifying cases corresponding to a particular context.
The CDL supports and uses the linkage of population datasets to address significant population health issues by evaluating current practices and guiding innovations that improve health services in Australia.
- Burns research. Together with researchers from the Fiona Wood Foundation (FWF), our team has investigated how burn injury impacts on the long term health of Australians. The WA Population-based Burn Injury Project (WAPBIP) has used linked data to produce a range of research papers.
- International collaboration. Working with researchers from Swansea University in Wales, the NHS National Services in Edinburgh, Scotland, the Institute for Clinical Evaluative Services in Toronto, Canada, the Ontario Institute for Cancer Research in Toronto, Canada, and the University of Western Australia, out team has recently determined there is no evidence to support a link between vasectomy and prostate cancer. Read full paper.
Researchers at the CDL also engage in social science and criminal justice-related research and contribute to national inquiries:
- ARC Linkage Project – Aboriginal and Non-Aboriginal sex offenders in Australia: Assessing risk for practice and policy
- ARC Linkage Project – Transitioning from out-of-home care: a longitudinal population-based study
- Prolific Offenders in Western Australia
- Enhancing the coordination of housing supports for individuals leaving institutional settings (funded by the Australian Housing and Urban Research Institute PRO/53210
- National inquiries:
A strong focus on cross-institutional alliances and national and international collaborations has helped us produce high-quality, relevant research outcomes. Our strategic partners include:
- Telethon Kids Institute
- Fiona Wood Foundation
- PHRN Partners
- Queensland Treasury
- Victorian government (Department of Premier and Cabinet)
- Swansea University (Wales)
- University of Duisburg-Essen (Germany)
- University of British Columbia (Canada).