You mean the ‘results’ schema But that’s good that it is working for you.
Right. As of now, there’s no specification about how those cohorts in the OMOP’s cohort table can be identified. Is it a concept ID? Is it some globally unique ID? It’s hard to write tools for standardized analysis when this critical piece of information is not specified.
What sort of information did you enter into the OMOP CDM tables?
I’m not quite understanding the ‘either-or’ phrasing of your question: The Atlas tools don’t use don’t use the OMOP cohort tables (and I’m not aware of any OHSI tools that do). Is atlas the only way through wich we can create cohorts? No, you can write your own custom code to create cohorts, but Atlas only knows about cohorts that Atlas creates (makes sense, right?) for use in the other functions of the tool: Cohort Characterization, Treatment Pathways, Incidence Rate calculations, Estimation and Prediction.
To do more custom things, you’ll have to drop into ‘raw SQL’ or ‘raw R programming’ mode. For example, with Estimation and Prediction, you can create a project referencing atlas cohorts, but then when you download the R code, you are free to modify it to use your custom cohorts or custom code. Likewise, with cohort definitions, you can always create a cohort definition, export the SQL, and modify it to reference the OMOP cohort table, if that’s what you need, or even reference other external data that you can import into a separate table. Just remember: if you add custom elements to your queries/code, it may prevent you from sharing those designs/specifications with other members in the OHDSI network.