Here are 7 mistakes I have seen made again and again in the first few months of FHIR projects.
Any one of them can put your project’s success in doubt.
Any one of them can lead to delays, expensive rewrites and unmet expectations.
1. Choosing the wrong FHIR server provider
- One FHIR server is NOT the same as another.
- Be very careful and thorough when you’re making this decision.
2. No plan for managing terminology
- The longer this is delayed the bigger the problem becomes.
- Start small, but do start.
3. No early meetings with legal, privacy and compliance
- What data can you store?
- What can you do with it?
- Do not leave these meetings until late in the project.
4. Pushing the multi-tenant architectural decisions out
- You need to be thinking about this from the beginning.
- Number 1 above will directly impact this.
5. Not filling critical expertise gaps early
- Clinical workflow expertise.
- Business domain expertise.
- FHIR & terminology expertise.
- The companies with the experience to know better are the very companies who make this mistake.
6. Making assumptions about incoming data
- A sample of 1 is worse than no sample data at all.
- Talk to stakeholders and get access to real data.
7. Not understanding FHIR’s limitations
- FHIR is an API, not a database.
- It’s not always possible to deliver what the business wants.
These early mistakes pop up more on FHIR projects than on other healthcare data projects.
The cost of fixing them increases as the project matures.
Fix them early and fix them fast.
---