CLIF as a relational SQL database

[08wparker]

we should make CLIF "SQL ready" because SQL databases have an explicit relational database structure that will help us achieve the goal of having each hospitalization represented coherently in CLIF.

patient: patient_id is the primary key hospitalization: hospitalization_id is the primary key. patient_id is a foreign key to link hospitalizations to patients vitals: hospitalization_id is a foreign key to link each set of vitals to a specific hospitalization. labs: hospitalization_id also serves as a foreign key to associate lab results with the corresponding hospital stay.

I also like this:

Foreign Key Constraints: By using SQL, you can enforce foreign key constraints to maintain referential integrity. This ensures that every hospitalization_id in the vitals, labs, or medications tables must exist in the hospitalization table. It helps prevent orphaned records (e.g., vitals records that reference a non-existent hospitalization).

SQL queries will be a very efficient way to do things like identify sepsis, etc.

Let's add a running to do list to this issue. I think we are pretty close!

[08wparker]

1. add a primary key for each table, e.g. vital_id
2. use SQL datatypes