MIMIC-III Concepts for Sqlite3 or DuckDB (probably now DuckDB)

Prerequisites

• Put an X between the brackets on this line if you have done all of the following:
  • Checked the online documentation: https://mimic.mit.edu/
  • Checked that your issue isn't already addressed: https://github.com/MIT-LCP/mimic-code/issues?utf8=%E2%9C%93&q=

Description

Filing this because I am considering adding it: Create a set of scripts that will add the MIMIC-III concepts views to the Sqlite3 version of the database.

Has anyone tried this before, and if so any encountered difficulties? I'm having some initial success just with regexp-ing the Postgres scripts, though that may or may not be a practical solution for the whole task. Sqlite does not have functions/procedures, so some things might have to be done with python functions--which would preclude embedding them into materialized views (pretty sure), however, simply building the concepts entities as static tables may be an acceptable solution for the kinds of use cases you'd want to use this data and Sqlite for. Or maybe regexping the Postgres/BigQuery SQL will get me to views in Sqlite3 dialect.

Start by reviewing the existing PostgreSQL and BigQuery concept SQL and the MIMIC-III documentation. Determine which concepts can be translated to SQLite or DuckDB, and whether Python functions or static tables are needed for unsupported database features. Done means the concepts are reproducibly created and usable in the target database.