/HEAT_DRF

Is it feasible to adapt the HEAT model to incorporate PA distributions and DRFs

Primary LanguageHTML

GitHub last commit GitHub top language GitHub repo size GitHub release License: MIT

A comparison of WHO-HEAT model results using a non-linear physical activity dose response function with results from the existing tool.

Authors: Robert Smith1, Chloe Thomas1, Hazel Squires1 & Elizabeth Goyder1

1 School of Health and Related Research, University of Sheffield, Regents Court, UK, S1 4DA

Abstract

The WHO-Europe’s Health Economic Assessment Tool is a tool used to estimate the costs and benefits of changes in walking and cycling. Due to data limitations the tool’s physical activity module assumes a linear dose response relationship be-tween physical activity and mortality.

This study estimates baseline population physical activity distributions for 44 coun-tries included in the HEAT. It then compares, for three different scenarios, the results generated by the current method, using a linear dose-response relationship, with results generated using a non-linear dose-response relationship.

The study finds that estimated deaths averted are relatively higher (lower) using the non-linear effect in countries with less (more) active populations. This difference is largest for interventions which affect the activity levels of the least active the most. Since more active populations, e.g. in Eastern Europe, also tend to have lower Value of a Statistical Life estimates the net monetary benefit estimated by the scenarios are much higher in western-Europe than eastern-Europe.

Using a non-linear dose response function results in materially different estimateswhere populations are particularly inactive or particularly active. Estimating baseline distributions is possible with limited additional data requirements, although the method has yet to be validated. Given the significant role of the physical activity module within the HEAT tool it is likely that in the evaluation of many interventions the monetary benefit estimates will be sensitive to the choice of the physical activity dose response function.

Repository structure & replication

  • the scripts folder contains the scripts necessary to replicate the analysis conducted in the paper.
    • The distributions_create.R file creates the country level physical activity distributions as described by Hafner et al.
    • The analysis.R undertakes the analysis discussed in the paper.
  • the R folder contains all functions required for the analysis.
  • the output folder contains the outputs generated by running the analysis.R file.
  • the data folder contains data which is read in during the cleaning phase.
  • This work was originally undertaken during a trip to Zurich in 2018, when the author was new to R. I apologise sincerely for the structure of the code, ideally I would refactor this code & set up the repository from scratch.

To replicate the analysis run:

source("scripts/analysis.R")

This will run the analysis.R script which:

  1. loads and cleans all necessary data, stored in the data folder, data and merges the data into a single dataframe.
  2. reads in the model function in R/model.R used to compare the three scenarios using the HEAT with a linear and a dose-response function.
  3. reads in the plotting functions in R/plotfunctions.R to visualise the results.
  4. uses R/relativerisksplot.R to plot the relative risk functions based upon Woodcock et al.
  5. runs the analysis described in the publication for three different scenarios and stores results as .
  6. uses R/all_plots.R to read in all the results from the csvs created above, and plots the results, saving them to the 'outputs' folder.

As I mentioned, it would be nice to have time to refactor this code, but unfortunately this project was time limited. If you are interested in building upon this work please contact me directly.

Acknowledgements

The authors would like to thank Martin Stepanek & Marco Hafner for detailed explanations explaining their method to estimate country physical activity distributions. We would also like to thank Sonja Kahlmeier, Thomas Gostski & Alberto Castro Fernandez for providing details of and access to the HEAT model and data.

Funding Sources

Robert Smith is joint funded by the Wellcome Trust Doctoral Training Centre in Public Health Economics and Decision Science [108903/Z/19/Z] and the University of Sheffield