Aloha OHDSI,
My name is Breck Yunits and I’m writing from the UH Cancer Center. We are working on a new type of medical record system here called Pau. (GitHub - treenotation/pau: Medical records you can copy and paste)
Even better than us writing Pau though would be to help you all use our one innovation, Tree Notation, to make OMOP CDM better.
Tree Notation let’s you encode data with all the bells and whistles of current technology in a much simpler way–no weird syntax characters like "’(){};
etc. In Tree Notation data and/or code are like spreadsheets with rows (nodes) and columns (words). You can indent a row to make one row a child of a parent row (similar to Python). You can then create things called “Tree Languages” which give meaning to these words. These are like your schemas/grammars.
What I propose for OMOP is this: we take your semantics and turn them into Tree Grammars, and then your code will be self-documenting, you’ll get autocomplete, syntax highlighting, type checking, program synthesis, readmes, ETL, and more all for free. That’s kind of a lot to take in, I know, so perhaps next step would be to have a discussion on a Zoom or something to explain more about what the heck I’m talking about.
Mahalo!
-Breck
yunits@hawaii.edu