Long Covid and PASC Studyathon - Study package

Package overview

This package contains all the required code to run for the study on Long Covid and PASC taking place in Oxford, April 2023.

Package extraction

You can clone this repository in your console like so, from cmd:

git clone git@github.com:oxford-pharmacoepi/Studyathon_LCPASC.git

or directly download it as a .zip file into your computer, from the menu above.

How to use it

You should only need to use the file CodetoRun.R in the main folder. First fill all the ... with the information from your database, and change the required variables (dates, names and database information availability) to suit your particular needs.

When you run that script, the whole code will start running. It will first instantiate the cohorts and then calculate everything for objectives 1, 2 and 3. If you re-run the code and the cohorts are already instantiated, you can set instantiateCohorts = FALSE so as to not repeat long calculations. Same with any other part of the study, like the creation of the study cohorts, getStudyCohorts = FALSE, the incidence calculations, doIncidencePrevalence = FALSE, etc.

Results

All the results will be zipped into a folder called [your database name]_Results. Some of them will have to be inspected and tuned afterwards, during the Studyathon, for instance the LCA clustering.

Trajectories

For the trajectories part, note that it might not work if your database engine is not SQLite or Postgresql. If so, be aware that the package has not been tested against those types of database engines, so you might have to set it as FALSE.

Plots

The package does offer some visualisations, which you can check in your final .zip file, but these are mainly orientative, to check it has worked well. Further plots will be done after receipt of raw outputs.