Large scale, de-identified general practice (GP) patient data has been linked into existing Victorian health data sets for the first time, allowing better tracking of health trends over time and the possibility of creating predictive models to flag people at risk of hospitalisation.
The Victorian Agency for Health Information’s (VAHI) Cardiac Outcomes linked data project team worked with the Centre for Victorian Data Linkage (CVDL) to successfully link MedicineInsight GP data into datasets held by CVDL through their integrated data resource.
Victoria’s residents interact with general practice more than any other part of the health system, so having GP data as part of the datasets enables a much better understanding of the movement of people through the system.
Following the successful linkage, the project team has been undertaking a thorough process of exploratory analysis to understand how patient groups interact with their GP and hospital. Having access to more than 10 years’ worth of data means the team has been able to generate broad insights and understand trends over time, initially focusing on the pre-pandemic period. A project working group, which includes GPs and health consumers, has been guiding the exploratory analysis and will shape more targeted analyses and product development over the second half of 2023.
While the project has been funded to focus on cardiovascular conditions, it also aims to demonstrate the broader value of the linked primary care–hospital data asset to be used to explore other chronic conditions, such as diabetes, in future. The linkage was completed using privacy preserving record linkage techniques, meaning patient identifying details were fully encrypted before leaving the GP software system with no identifiers supplied to CVDL. Only a de-identified data set is created for analysis by the project team.