Gerard Freriks wrote: > Thomas, > > What gets modeled where? I would argue that since OGTT is a very common test, it is an archetype which it is - see the glucose result archetype. For the sports ones, I would expect to see some archetypes as well, but I know very little about this part of medicine.
I was not making any suggestions about which is used however, only saying why the underlying reference mode is the way it is. > > Reading your examples I have the following questions: > - GTT, "want to see one graph". Are you talking about an Archetype or > a Template? > I think you are talking Templates where clinicians want to see something. > They use in the template archetypes. > > Archetypes are used to store information in the database. > When I store 10 Glucose measurements and patient states obtained at 10 > different dates and times in the data base each as single point > measurement using an Archetype each time to store that measurement, > and when I store 10 Glucose measurements and patient states obtained > at 10 different dates and times as a times series in the data base > using one Archetype, > can I retrieve *all* glucose data and patient states using one version > of the Archetype or do have have to use one Archetype version with all > measurements modeled separately and one Archetype with all > measurements modeled as a time series? this will be an application issue, since in reality more than one different time-series measurement has been made (even if some of the subsequent ones only have one sample). If one group of samples is indeed made using the same protocol, as a single test, then it should be represented as a single Observation with a single History containing N (usually 3) Events. - thomas
