The Sheffield Tobacco and Alcohol Policy Modelling Platform

The aim of the STAPM research programme is to identify and evaluate approaches to reducing the harm from tobacco and alcohol, with the aim of improving commissioning in a public health policy context, i.e. providing knowledge to support benefits achieved by policymakers.

Beginning in 2014, as part of the UK Centre for Tobacco and Alcohol Studies (UKCTAS), the modelling approach used in the Sheffield Alcohol Policy Model (SAPM) was extended to model the effects of tobacco policies and interventions on the dynamics of smoking, and to investigate the effects of interactions between policy effects on tobacco and alcohol consumption.

The two objectives of the STAPM research programme are:

The STAPM platform is not linked to the tobacco or alcohol industry and is conducted without industry funding or influence.


Background

As part of the UK Centre for Tobacco and Alcohol Studies (UKCTAS), the Sheffield Alcohol Policy Model (SAPM) was extended to model the effects of interventions on tobacco smoking as well as alcohol consumption. The goal was to be able to assess the potential scale of effect of tobacco and alcohol policy on health and economic outcomes within a common modelling framework.

To guide development of the joint tobacco and alcohol modelling, a workshop was organised to which a range of UK public health policy and academic experts in tobacco and alcohol were invited. The outcome of this workshop was a description of the main elements and mechanisms within the joint tobacco and alcohol policy system, which covered five policy/intervention themes: price, place, person, promotion, prescriptive and a cross-cutting theme of industry regulation.

Development of the STAPM modelling platform is continuing as part of the SPECTRUM UK Prevention Research Partnership consortium. This work extends the modelling already developed for England to Scotland and Wales, develops further health and economic outcomes, and continues to evaluate and appraise policies and interventions.


Approach to modelling

The aim of this website is to make the processes within the Sheffield Tobacco and Alcohol Policy Modelling as open and well explained as possible. The main purpose of the website is to help navigate model documentation, data and code across multiple projects that build on the STAPM platform. Many parts of this website are therefore a working draft that are added to, reviewed and edited as the STAPM platform evolves and new projects are added and completed.

Data - The STAPM platform utilises several different data sources, resulting in a wide array of code to prepare the data inputs to the modelling, to run the models, and to present the outcomes. The datasets used to inform the modelling are managed so that they can be used and re-used in a consistent way across projects - this involves managing both the information governance and the code used to conduct basic data cleaning operations.

Code - The modelling was developed in R and R studio, with particular attention to utilising new developments in the software environment to organise, document, and version control code. A set of internal STAPM R packages have been developed that contain code used for particular purposes, e.g. to estimate smoking state transition probabilities from cross-sectional survey data. Packaging up code into modular functions and packages makes the processes applied in modelling easier to document, adapt and reuse across projects. The long term goal is to make the STAPM code open source. Some STAPM code is starting to become publicly available on Github, but the majority is currently available only to the project team on Gitlab.








 

Last updated: