Best practice for maintaining multiple versions?


We’re looking into implementing LOINC and are wondering how others go about supporting multiple versions of the schema at the same time. For example, an old record in our system might have some data that was recorded using a previous version, and then we want to have new data recorded using the new version, but when someone views the old version, the system should be able to use the old schema to display/manage that data.



A question just for clarification…does your system re-use local codes over time, possibly changing meaning or intent of the lab test?

Thanks in advance,


I would say that it’s possible that a given observation might need to be interpreted slightly differently depending on the version in use at the time of the observation.

But I think I’m mostly asking about whether anyone has come across anything related to versioning where they found doing it a certain way worked really well, or similarly where they later regretted how they implemented it and would therefore be a practice to avoid.