Data & Evidence
A secure, multi-layered data ecosystem for public health modelling
The STAPM research programme utilizes a comprehensive data ecosystem. We integrate administrative health records, national surveys, and commercial market intelligence to produce independent evidence for public health policy across the UK.
While this page highlights our foundational datasets, our team continuously integrates supplementary sources tailored to specific policy appraisals.
🔒 Secure Data Infrastructure
To manage our data assets securely and scalably, STAPM processes key datasets within the University of Sheffield’s Data Connect Secure Data Environment. This audited, highly secure infrastructure ensures that sensitive information is strictly controlled, anonymised where appropriate, and used solely for public-benefit research.
📊 Core Data Themes
Our modelling relies on robust partnerships with key data providers and funders to ensure our simulations reflect real-world populations and behaviours.
🏥 Administrative Health
Grounding our models in health outcomes and demographic realities across Great Britain.
National Hospital Admissions Data We apply for and maintain secure agreements to access hospital admissions data across the nations. This includes Hospital Episode Statistics for England, as well as the equivalent administrative datasets for Scotland and Wales. Maintaining these parallel datasets is essential for translating consumption changes into tangible healthcare impacts across the different devolved health systems. (See our Privacy Notice for our English data governance details)
Demographics & Mortality We utilize demographic and mortality data from the Office for National Statistics for England and Wales, and the National Records of Scotland. Held under secure access agreements, this provides the vital population counts for all our public health modelling.
📋 Behavioural Surveys
Tracking population-level consumption of alcohol and tobacco.
University College London & Cancer Research UK Through an ongoing collaboration with the Tobacco and Alcohol Research Group at University College London, we utilise the Smoking and Alcohol Toolkit Studies. Integrating these datasets, which are funded by Cancer Research UK, allows us to capture high-frequency, real-world behavioural trends.
National Health Surveys We rely on national health surveys via the UK Data Service—specifically the Health Survey for England, the Scottish Health Survey, and the National Survey for Wales—to establish our baseline understanding of public consumption.
🛒 Market & Economic
Understanding the economic drivers of consumption, pricing, and tax pass-through.
Consumer Spending Surveys We utilize the Living Costs and Food Survey to track how economic pressures shift household expenditure patterns.
Commercial Market Intelligence To simulate real-world market dynamics—from localized pricing to massive-scale sales volumes—we incorporate granular commercial data. We track independent retail sales via The Retail Data Partnership (ShopMate), monitor supermarket trends through NIQ and Circana, and link purchases to household demographics using consumer panel data from Kantar.
Together, this data is essential for modelling how fiscal policies will impact both consumer wallets and industry revenues.
🗺️ Conceptual Modelling & Evidence Mapping
Before building a mathematical model, we systematically map the evidence base to understand the structure of the policy problem. We do not simply input available datasets; rather, we systematically review the evidence to identify the most robust data for specific public health challenges.
Our approach begins with conceptual modelling to map out the system and its mechanisms. In fact, the STAPM platform itself is built upon a foundational conceptual model that maps out how changes to tobacco and alcohol policy affect a single, interlinked system.
By establishing this theoretical framework first, we can rigorously map the available evidence against it. This process ensures our models are theoretically sound, identifies critical data gaps, and highlights exactly where primary data collection must be focused.