Stream: implementers
Topic: Birth outcomes
Jose Costa Teixeira (Jul 12 2021 at 10:54):
How should one list the birth outcomes in a simple way? What I want is to say
- Pregancy 1 had 2 births at 37 weeks,
- first weighing 3,4 kg, etc,
- second was 2,4 kg
- Pregnancy 2 resulted in 1 still birth
Jose Costa Teixeira (Jul 12 2021 at 11:02):
I can use a Patient for the mother, plus one patient for each child (stillborn also?) and use the Link + RelatedPerson. But in this case since my focus is really the mother only, would this be solved in another way?
Craig Newman (Jul 12 2021 at 12:58):
The US Birth and Fetal Death IG is doing some of this https://build.fhir.org/ig/HL7/fhir-bfdr/index.html and there's also the US Vital Records common library http://build.fhir.org/ig/HL7/fhir-vr-common-ig/branches/master/index.html. I'm not sure there is a standard way of doing this but @Sarah Gaunt has done a lot of work on pregnancy and delivery related Observations (and other profiles). There is also the CHOICE group (https://confluence.hl7.org/pages/viewpage.action?pageId=86973863) which conveniently meets this Wednesday that is working in the space and might have ideas.
Craig Newman (Jul 12 2021 at 13:01):
My general thought is that that any observation about the pregnancy has the Mother's patient record as the subject but if it's unique to the individual (a newborn's birth weight or the pregnancy outcome (live birth, still birth, etc) ) then there needs to be a Patient for the child. Grahame did suggest using .focus for some of the Observations where the parent may not have a Patient record (eg. you need an observation about the father's education level or the information is about the Mother but she's not actually a patient (eg. the birth was at another location but the current provider is collecting the data for their uses). In this case, .focus could point to the person's RelatedPerson resource
Sarah Gaunt (Jul 12 2021 at 22:35):
I think @Craig Newman has covered what we've done in BFDR pretty well.
We use Patient.multipleBirth = number in the delivery (if multiple) - there is a Patient for a child and a Patient for a deceased fetus. We use an observation-gestational-age-at-delivery (subject = baby or fetus) for gestation estimate (37 weeks) and another Vital Sign Observation (Observation-birth-weight) (subject = baby or fetus) for birth/delivery weight.
We don't actually model the Pregnancy in BFDR, but we do in eCR and there will be a Public Health common profile for that (subject = mother): https://build.fhir.org/ig/HL7/fhir-us-ph-common-library-ig/StructureDefinition-lib-copy-us-ph-pregnancy-status-observation.html. There is also a Pregnancy Outcome profile in that common library. (BTW - the PH common library isn't a project yet, so it's not possible to simply reference any of the profiles - for now we are making copies of them to use in various IGs - hopefully will go to ballot in Jan 22).
In BFDR we follow the guidance here to relate everything: https://www.hl7.org/fhir/patient.html#maternity
Elliot Silver (Jul 12 2021 at 22:50):
@Jose Costa Teixeira IPS also pregnancy outcomes, you should look at that.
Last updated: Apr 12 2022 at 19:14 UTC
