Ms Rory Chen1, Dr Vibeke Catts1, Dr Darren Lipnicki1, Ms Naja Fragante1, Prof Perminder Sachdev1
1Centre for Healthy Brain Ageing (CHeBA), UNSW Sydney, Australia
Biography:
Rory joined the Centre for Healthy Brain Ageing (CHeBA) within the School of Psychiatry in January 2021. She is the manager of Dementias Platform Australia (DPAU).
Rory is a data scientist with an MSc in Statistics from UNSW and a BSc in Statistics from Beijing Normal University.
She previously worked on various clinical trials and observational studies at the National Drug and Alcohol Research Centre (NDARC) and the Chinese Center for Disease Control and Prevention (CDC).
Abstract:
The Dementias Platform Australia (DPAU), established by the Centre for Healthy Brain Ageing (CHeBA), aims to accelerate dementia research by enhancing the findability, accessibility, interoperability, and reusability (FAIR) of data. As part of this, we developed a semi-automated measurement-level data sharing workflow using Python, REDCap, and R.
Data from CHeBA’s longitudinal cohort studies: Sydney Memory and Ageing Study, Sydney Centenarian Study, and Older Australian Twins Study, have been shared on >500 occasions since 2010. The data cover 1000s of variables from >100 measurements in domains such as sociodemographic, cognition, medical history and lifestyle surveys. With the onboarding onto DPAU, opportunities for automation and efficiencies were explored.
Using Python, data files and variables are renamed following C-Surv Ontology, and metadata is extracted to create data dictionaries. REDCap is configured to capture study metadata and the project application process, including measurement-level data requests. Upon approval, Python extracts relevant data files for each project. R Shiny applications complement this through multiple dashboards tracking administrative workflows and provide tools for applicants to explore study metadata before submitting requests. R and Python are used to create data dictionaries and import data for REDCap.
Detailed diagrams and real-time demonstration of the integrated workflow illustrate the reduced repetitive manual work, improved operational efficiency, streamlined reporting, and enhanced transparency of the workflow to the e-research community. The current system is highly customised and adaptation by other users may require significant effort, including a need for technical expertise. Additional work is required to ensure compliance and data security.