I have dealt with this scenario before. The type of data is called "slowly changing dimension" or SCD. SCDs might never change, or change daily. It all depends

I have dealt with this scenario before. The type of data is called "slowly changing dimension" or SCD. SCDs might never change, or change daily. It all depends on the business logic. To implement this tracking, I have seen row-level historical snapshots for auditing and dimension-level tables to track the less-important dimension changes. For health data, row-level snapshots with a key generated from the data might be most appropriate. For data that has less legal implications, a table that has "id", "dimension value", and "timestamp" may be enough. From that, using from advanced SQL, you can even produce dimension time intervals. An example where that would be useful is tracking the average time a support ticket spends in "Waiting on Customer" or a call spends waiting in queue to be answered.