Should I use FHIR in production?

A question was asked today on the Implementer’s chat that has been asked many times before: “Should I be using FHIR in production today?”

There are a number of answers to this – ranging from the fact that quite a lot of people actually are, through to some comments in Grahame’s keynote at the Amsterdam devdays concerning “Working Interoperability” and talking about the resource maturity model.

But I thought the most succinct one was this from James (HAPI creator)

the point that i’ve always made when trying to make the “net benefit” argument is that even if you’re using FHIR in a completely closed environment, the alternative is to do your own data modeling and API design. no matter how good a job you do of that, you still have to do the work, and you’ll still never come up with something as robust as FHIR since FHIR has many peoples’ implementation experience baked into it. add to that the fact that FHIR is well documented (so you don’t need to do that part either) and you have a very compelling case.

Posted here so I can find it when the question is next asked!