The Data, Digital Technology & AI (D2A!) Core provides data infrastructure, data governance and aims to develop machine learning (ML) and data analytics expertise, focusing on structured and non-imaging unstructured data, within NHCS and NHRIS. D2A! supports NHCS research groups with the initiation, execution, analysis and governance of their data-related/data-science projects and provides essential biostatistical services. It is a one-stop platform for external partners (e.g., industry collaborators) to interface with for data-related projects.
The core also provides clinicians/scientists with access to resources such as SingCLOUD, Singapore Cardiac Data Bank (SCDB), ODySSEy and the upcoming National Supercomputer Centre (NSCC) @ Outram. The goal of the core is to develop data science projects and promote new collaborations between NHCS, industry and research institutions in the digital space. D2A! is supported by the Singapore Cardiac Data bank (SCDB), a national registry collecting epidemiological and clinical information on cardiovascular diseases and procedures such as percutaneous coronary interventions (PCI), coronary bypass surgery, electrophysiology and pacing as well as heart failure admissions. All restructured hospitals in Singapore contribute to SCDB.
D2A! is headed by Clinical A/Prof Yeo Khung Keong, senior consultant cardiologist and Deputy Medical Director (Data Science and Innovation). He is also SingHealth’s Deputy Group Chief Medical Informatics Officer (Research), Deputy Chairperson of SingHealth Innovation Council and Chair of the Data and Emerging Technologies focus area of the SingHealth Duke-NUS Academic Medicine Innovation Institute.
Clinical A/Prof Yeo Khung Keong Assoc Prof Gao Fei Miss Wendy Tan
In the data-driven/data-informed landscape, new digital technologies and associated artificial intelligence (AI) will play a significant role in the future management and treatment of our patients.
These technologies have the potential to accelerate, streamline and optimize clinical procedures and reduce costs. The core has three work streams:
It is important to establish and develop a strong and highly trained team to manage and analyze the rapidly increasing and complex data. D2A! will build expertise and capabilities to organize data with more coherence to elucidate meaningful research answers and give NHCS an edge in this space. This will poise the institution/cluster to create more customized, informed, and personalized treatment strategies for future patients. D2A! is also positioned to be a conduit and integrated platform for discovery of possible collaboration within and beyond the institution (e.g., with IHL and industry players), fostering an eco-system which allows for future innovations and initiatives.
Our four main strategies include:
1. Singapore Cardiac Data bank (SCDB)
SCDB was established in 2000 as a national data bank of cardiovascular diseases and procedures. It is a rich source of data with 80% of hospital care delivered in public institutions. Current modules include cath/PCI, adult cardiac surgery, ICD, pacemaker, electrophysiology studies, ablation and heart failure admissions. The data is utilised for a number of audit/quality reporting purposes and research projects. Biennial data quality audits conducted by a 3rd-party auditor to assess data validity and reliability have shown data agreement of ~97% for both cardiac surgery and Cath/PCI data. A list of the focus areas and participating institutions can be seen in the figure below:
2. Singapore Cardiac Longitudinal OUtcome Database (SingCLOUD) Study
SingCLOUD is a research project initiated by NHCS and its participating institutions which includes MOH, all restructured hospitals and polyclinics.
List of participating institutions in the SingCLOUD programme
It aims to perform data extraction, integration and analysis across multiple clinical, administrative, financial and pharmacy databases in the participating institutions. The key data components will include cardiovascular specific data, procedural data, financial and administrative data (e.g. cost of medications and supplies, length of stay, diagnosis codes etc.), short-to-long term clinical outcomes (including hospital re-admissions, cardiovascular events, death), and quality-of-care indices (e.g. compliance to guideline-recommended therapy and medications). SingCLOUD leverages on BRAIN (Business Research Analytics Insights Network), formerly known as Health Data Grid, which is a virtual repository that enables access to anonymised and merged data from disparate databases at multiple institutions. The structures of BRAIN will preserve patients’ privacy and confidentiality, ensure IT security and allow access only by authorised persons. The building of BRAIN has been completed and SingCLOUD is currently at the stage of data analysis. Plans for SingCLOUD 2.0 are underway, and once it’s live, it will have data updated till 2021, which will allow more meaningful research to be conducted.
3. SingHEART
SingHEART is a 20-year study on a cohort of relatively healthy individuals, using comprehensive clinical information, advanced imaging, cardiovascular exercise physiology, metabolomics and genetic analysis. These findings will be correlated with adverse clinical outcomes including death, stroke and myocardial infarction. In selected cases, follow-up imaging and biomarker samples will also be obtained. This longitudinal study will enable us to address a critical gap in our knowledge as to how cardiovascular disease develops in previously healthy individuals.
Being in collaboration with NHCS Biobank, SingHEART recruit volunteers from the Biobank cohort. As of December 2021, 1007 healthy volunteers were recruited into SingHEART and has completed all baseline procedures.
Leveraging on SingHEART's extensive collection of phenotypic data, we were able to work with the SingHealth Duke-NUS Institute of PRecIsion Medicine (PRISM), which is the Academic Medical Center's flagship initiative in precision medicine. SingHEART has contributed to their signature program, SPECTRA, where its aim is to collect a multi-dimensional database of healthy Asian normality incorporating genomic, clinical, lifestyle, and imaging data. This therefore serves as an invaluable reference database for disease-oriented studies at SingHealth/Duke-NUS and across the nation.
Apart from SPECTRA, SingHEART has ongoing collaborations with a nationwide level "SG100K" project and AI.SG.
Data Available in SingHEART
4. External collaborations
There are several ongoing collaborations with national entities, academics institutions, pharmaceutical and technical companies. These collaborations include but not limited to the National Registry of Disease Office (NRDO), Singapore Consortium of Cohort Studies (SCCS), Singapore Eye Research Institute (SERI), Agency for Science, Technology and Research (A*Star), National University of Singapore (NUS) and AstraZeneca.
Subscribe to our mailing list to get the updates to your email inbox...
