How to use Pauá

The idea behind sitting the model in GitHub is to be able to run the different studies with a server inside an institution that pulls the scripts to run from the repository and runs over the database.

A user friendly serverless tool runs on a browser as an html page (docs folder of the same Github repository) and edits a json version of the parameters that define the selected cohort which are converted to insert statements and stored on the repository, as a new version of the parameters table.

Shiny will be a nice addition to make the same edition of parameters for the R scripts. Both scripts (SQL and R) are intended to be run by a simple Continuous Integration (CI) Server (like Jenkins) behind the institution, triggered by a request of a researcher, that forked the original repository and stated on it his study parameters, sending the repository address and version number.

The CI Server peaks the scripts from the repository and runs over the EHR database returning the results to the email of the sender. This way we ensure a repeatable process and references to the original work with means it is easy to check modifications made to the original scripts.

Even more, as collaboration evolves, changes can be pushed to the original repository to enhance studies. We have limited resources (time and man power) to put everything on place right now, but we are committed to do so and really understand that this way we are paving our road to a better understanding of large solutions as OMOP model which is indeed the right way to go.

How to run:

You need to setup a server with an account and an schema with access to the EHR's database and enough permissions to create tables, views and materialized views. Server may be any PC (any OS) setup to run:

• RStudio
• JenkinsCI

This is not ready yet. So be patient and come back to check for updates.
By now, you can check the tools for edit parameters for the model, the CohortEx Editor.