Very nice points.
We should have such a thing for the community and crowdsource. Do you mind sharing?
This is a good one, it comes up often. The question is: Should an OMOP instance be built on top of an existing data warehouse, or should it be the warehouse. I have seen both. It’s easier for the ETLer if there is already a warehouse, but the overall effort of course is a little duplicated.
There is more: almost all of the data munging that typically takes 70% of a study goes away.
The other ones that keep popping up:
- How much effort is it to do an ETL? They typically don’t ask about the refreshing.
- Who else has done it who is like me?
- Can OMOP support therapeutic area XYZ (typically pediatrics and cancer)?
- Isn’t there data loss during the ETL?
- Is it HIPAA/GDPR/other-patient-protection-schema compliant?
- What am I getting if I do that? What’s in it for me?
And the most important question, which rarely gets asked openly and honestly:
- Will I still keep control over my data, or is any Tom, Dick or Harry going to come in and do crazy queries, which may hurt my reputation?