Converting FHIR to OMOP is becoming more and more common.
A big reason for doing so is to make de-identified FHIR data available for analytical purposes – something that FHIR is not good at.
The OMOP data often ends up in data platforms such as Snowflake, Databricks or Fabric where it can be aggressively queried in multiple ways.
If you’ve been landed with investigating or undertaking this task and are unsure where to start, here are 6 resources I’ve compiled to help you on your way.
- “A Guide to the OMOP Common Data Model” (4 minutes)
A quick and dirty introduction to OMOP to get you started.
https://www.iqvia.com/blogs/2025/06/a-guide-to-the-omop-common-data-model - “Overview of the OMOP CDM” (11 minutes)
A more in-depth video tutorial from Stanford that provides a visual overview of the OMOP CDM schema.
https://www.youtube.com/watch?v=oa-bAVHBog8 - “An Open Approach for Translating FHIR to OMOP” (25 minutes)
In this video Carl Anderson walks through a small project using OS tools and FHIR ndjson files to translate FHIR resources into OMOP. This should get you thinking about how to build your conversion process.
https://www.youtube.com/watch?v=WouS6lrRoV8 - FHIR to OMOP Implementation Guide (20 minutes)
The Vulcan Accelerator’s comprehensive IG. When you get into the detail this should be your “go to” resource. Start with the main page for an introduction then move on to number 5 below.
https://build.fhir.org/ig/HL7/fhir-omop-ig/ - “Common Challenges When Transforming FHIR to OMOP” (20 minutes)
Insights into how to handle de-identification, privacy, resource statuses, data integrity and data completeness.
https://build.fhir.org/ig/HL7/fhir-omop-ig/F2OGeneralIssues.html - The Book of OHDSI (? days)
Everything about OMOP in one place. Pay particular attention to “Chapter 5: Standardized Vocabularies”. If your FHIR to OMOP conversion is implemented well, this is where you’ll spend a lot of time.
https://ohdsi.github.io/TheBookOfOhdsi/
---