Stream: v2 to FHIR
Topic: ratio+comparator
Christoph Müller (Oct 17 2019 at 15:20):
From our lab System we receive messages containing a ratio plus a comparator in OBX.5 (sth. like "< 1:80"). Assuming, that this style is valid (standard procedure for such cases in v2), then how to I transform this into a FHIR-resource correctly?
Intuitively i would do it via modeling the ratio first as a valueRatio and then taking the "<" into the comparator-property of the numerator. Would that be the way to do it?
Grahame Grieve (Oct 17 2019 at 21:32):
I think so
Christoph Müller (Oct 18 2019 at 06:45):
ok, great. Thank you!
Christoph Müller (Oct 23 2019 at 12:28):
I am sorry, but unfortunately I have to reopen this topic...
As for valueRatio this works perfectly fine. For referenceRange though, one may only provide SimpleQuantitites. In the case of titer for example, where the referenceRange would be sth. like "<1:20", the way to write it as discussed in my previous post would not work out. (I mean sure i could put this into referenceRange.text, but this seems not very FHIR and makes not much sence after alle the "effort", coding the value itself correct)
Does anybody have any experience on that, or maybe even some extension at hand?
Grahame Grieve (Oct 23 2019 at 17:32):
this is a reference range with high = 1:20, and no low, right?
Christoph Müller (Oct 24 2019 at 08:48):
@Grahame Grieve yes, the Ratio "1:20" is the high.
Christoph Müller (Nov 07 2019 at 13:18):
We now use an Extension to store that Ratio as a high. So tahnk you for that! :)
But, we have got another Problem at hand: within our transformation, cloverleaf replaces ">" and "<" by ">" and "<" respectively. This behavior seems legit, since the prior are not valid for XML, on the other hand the fhir spec name them as valid codes. The question now is: will the receiving systems all understand the ">" and "<" as for what the are, or is there another, better solution?
Thank you already!
Christoph
Lloyd McKenzie (Nov 07 2019 at 15:45):
The tool should be escaping and unescaping instances transparently...
Christoph Müller (Nov 08 2019 at 10:59):
ahahaha only now i saw, that the " & l t ;" above had already been converted. All problems solved so far then. Thank you for the help! :)
Last updated: Apr 12 2022 at 19:14 UTC
