Stream: genomics
Topic: code for a Group that means family
Andrew Patterson (May 13 2018 at 23:56):
Hi, I have a requirement to store an identifier for a 'family' - as part of a familial genomic order (trio). So whilst all the participants are linked via FamilyHistory into a proper pedigree - I have no resource in my model to store a single 'identifier' for this family unit. I'm toying with introducing a Group that will reference all these patients and have the family identifier - and now need a code to say what this group is. Current best suggestion is 'Family' in the social context hierarchy of SNOMED 35359004|Family (social concept). Any thoughts out there as to a better code or code system? Is the basic premise of using a FHIR Group for this correct? Thanks
Lloyd McKenzie (May 14 2018 at 05:15):
How do you define the bounds of a family?
Kevin Power (May 14 2018 at 10:38):
I would propose adding the family identifier as an extension on FamilyMemberHistory, and possibly even requesting to add it as a first class attribute. It is a common request for pedigrees.
Lloyd McKenzie (May 14 2018 at 11:50):
I'm fine with adding an extension, we just need a robust definition of what the boundaries are for a "family"
Andrew Patterson (May 15 2018 at 04:56):
I think 'family' is one of those ones that would be impossible to define in a general context - but would be possible to define and bound in the context of a genomic test. 'An identifier for all blood related members involved in a genomic test'? FamilyMemberHistory itself has no definition of what it means by family I note..
Lloyd McKenzie (May 15 2018 at 06:38):
Yes, but it also doesn't try to define an identifier for the collection. If you want to capture cousins 5 times removed, you can just establish the path to that person. However, if we're going to "identify" a family, we need to have consisteincy around what it is that we're identifying
Donna Truran (Aug 13 2018 at 06:02):
Hi, I have a requirement to store an identifier for a 'family' - as part of a familial genomic order (trio). So whilst all the participants are linked via FamilyHistory into a proper pedigree - I have no resource in my model to store a single 'identifier' for this family unit. I'm toying with introducing a Group that will reference all these patients and have the family identifier - and now need a code to say what this group is. Current best suggestion is 'Family' in the social context hierarchy of SNOMED 35359004|Family (social concept). Any thoughts out there as to a better code or code system? Is the basic premise of using a FHIR Group for this correct? Thanks
I'd be tempted to request new concepts from SNOMED International (or ADHA for AU release)
Joel Schneider (Aug 13 2018 at 22:00):
Using Group.identifier to hold the identifier for a trio (family unit) seems reasonable to me, assuming there's not another resource or profile that fits the requirement more closely.
However, using SNOMED 35359004|Family (social concept) as the Group.code for a genetic trio doesn't seem quite correct.
Ammar Husami (Aug 15 2018 at 14:03):
Check PED file format http://valdarlab.unc.edu/pedformat.html it works well for genomics applications that I have been using
Last updated: Apr 12 2022 at 19:14 UTC
