Stream: social
Topic: FHIR resources
Jain Philip (Mar 17 2020 at 16:13):
I am looking for guidance in transforming resources into a realational structure from the repository. Can any one guide me on this ?
Grahame Grieve (Mar 17 2020 at 18:42):
there's no standard way to do this because there's so many different ways to do it, but you could catch up on past discussions on #analytics on FHIR
Lloyd McKenzie (Mar 17 2020 at 19:22):
Also #storage for FHIR
Jain Philip (Mar 18 2020 at 12:35):
We are having an AWS Oracle RDS instance for FHIR repository. Can you point me to the table/column that will have resource body
Jain Philip (Mar 18 2020 at 12:37):
(deleted)
Jain Philip (Mar 18 2020 at 12:41):
this table /column seem close to the resource information; HFJ_RES_VER/res_text. The content of this a stream of uppercase characters and numbers. If this is an encoded text how can I convert it back to a plain text format.
James Agnew (Mar 18 2020 at 13:07):
Asuming this is a HAPI FHIR JPA server, the res_text column is Gzipped. You need to gunzip it to see the raw resource text.
Ps, you may wish to join the #hapi stream for hapi issues.
Jain Philip (Mar 19 2020 at 00:34):
@James Agnew HI James,
Thanks for clarifying my question. do you know if there is a oracle function I could use as part of the sql to gunzip the column content ?
James Agnew (Mar 19 2020 at 01:10):
Unfortunately that I am not sure. We do this at the application tier, and not at the DB tier.
I know I was surprised at one point to learn that Postgres has no such function, but that doesn't mean that Oracle doesn't..
Josh Collins (Mar 19 2020 at 13:12):
This could potentially help: https://oracle-base.com/articles/10g/utl_compress-compress-and-uncompress-data-from-plsql
Jose Guerrero (Apr 29 2020 at 04:14):
Hello everyone, I am new to this group! I am a rising medical student at UCLA David Geffen School of Medicine and have a few questions about the future of Electronic Health Records systems and interoperability. I am not sure if this is the place to ask but I was sent here by HL7 HQ. I have a very basic understanding of FHIR.
Will healthcare facilities be able to pull data from another facility simply with the resource id? Can a medical record number be used instead?
To be fully interoperable, does an API have to allow access to delete/update a fhir resource? To me, this sounds dangerous. Unless fhir resources are going to be stored an a separate database from the main database.
Will healthcare facilities have to register with one another for access, say through an API Key, to each facility's API?
Please excuse my lack of knowledge on this topic. I hope that I can better understand FHIR/interoperability since I spent my gap year before medical school working in two clinics where I faced current limitations in information exchange.
Lloyd McKenzie (Apr 29 2020 at 04:25):
Welcome to the community :)
FHIR's search capability should allow retrieval by MRN or various other search parameters. Retrieval by id only happens if you've already got a reference to the resource from some other resource.
Permission to update resources tends to be restricted to trusted parties and ability to delete tends to be even more restricted (if it's supported at all).
In most cases, healthcare facilities will need to register to talk with each other - either by site-to-site agreements or within an over-arching umbrella organization. That's not just to establish trust relationships, but also to establish mechanisms for validating user identity, rules around handling sensitive data, etc. The specific rules will vary by environment - how healthcare systems are managed and regulated varies significantly from country to country and those sorts of policies aren't ones that HL7 can enforce. You can find some of our recommendations here: http://www.hl7.org/fhir/safety.html
Mikael Rinnetmäki (Apr 29 2020 at 07:08):
In most cases, healthcare facilities will need to register to talk with each other
The other alternative, of course, is to make the process of sharing data between organizations mediated by patient, for instance using the International Patient Access specification. In this case the organizations do not need bilateral agreements or even an umbrella organization.
Kin Dia (Apr 29 2020 at 12:27):
HI I am new to this group I am leaning to use https://r4.ontoserver.csiro.au Terminology server . I am using $lookup operation. I want look up multiple code sets eg
https://r4.ontoserver.csiro.au/fhir/CodeSystem/$lookup?system=http://snomed.info/sct&code=11687002,45313011000036107
It is throwing a error.. Does terminology server support retrieving multiple codes in single query or is any other way I can use retrieve multiple codes in single query
Thanks
Kin
Jose Costa Teixeira (Apr 29 2020 at 12:33):
the cardinality of code is 0..1, so I guess you can't do a lookup with 2 codes
Lloyd McKenzie (Apr 29 2020 at 13:48):
@Kin Dia and @Jose Guerrero as a side note, technical questions are best asked on #implementers
Jose Guerrero (Apr 30 2020 at 02:07):
@Lloyd McKenzie Thank you for pointing me to the appropriate stream to ask questions!
Dave deBronkart (May 04 2020 at 21:16):
Mikael Rinnetmäki said:
The other alternative, of course, is to make the process of sharing data between organizations mediated by patient, for instance using the International Patient Access specification.
?!? How could I have never heard about this?? Links please! (I can go googling but that's no guarantee I'll find the best stuff.)
A top question for me is how this specification dovetails with the CURES Act rules, if at all.
Grahame Grieve (May 04 2020 at 21:27):
International Patient Access
is something I'm working on, which is really just an internationalisation of US Core + Argonaut
Dave deBronkart (May 04 2020 at 21:28):
And for those (including me) who aren't certain, does that mean they're US federal policy, i.e. we can count on them to be widely implemented someday?
Grahame Grieve (May 04 2020 at 21:40):
I think that adopting US core is federal policy (not IPA since that is the internationalization of US Core). But that doesn't make it certain to be widely implemented some day. Nothing is certain (except death and taxes)
Brandi Burgess (Jun 12 2020 at 18:38):
I know there are a lot of conversations on the FHIR resources. I have a very high level question that I am not sure has been answered about the resources on the HL7 website. Why does the HL7 reources give a differential view and a full view? What does it mean for the differential view? Are those the required fields? Is there an easier navigation to getting to what the list of resources that are needed for the Formulary and Clinical data build ?
Lloyd McKenzie (Jun 12 2020 at 18:44):
The differential view and full view is something we provide when documenting profiles. The differential view shows "what's changed?". The full view shows everything that's available when those changes are taken into account. Reviewers often only care about "what's changed?" but implementers need to take into account "what's available?". Thus we provide both views.
What do you mean by "the Formulary and Clinical data build"?
Lloyd McKenzie (Jun 12 2020 at 18:44):
(Side note - questions about the spec are generally raised in #implementers rather than #social )
Brandi Burgess (Jun 12 2020 at 19:09):
For our project, we are doing the CMS/ONC Interoperability (Rule/Mandate) and we are to use the FHIR resources to accomplish. Drug Formulary and Clinical Data are included.
Lloyd McKenzie (Jun 12 2020 at 19:13):
Then you should be looking at the Da Vinci PDex and PDex drug formulary IGs. The current draft of the IGs are here:
Discussion and questions related to those IGs are here:
- https://chat.fhir.org/#narrow/stream/235286-Da-Vinci.20PDex
- https://chat.fhir.org/#narrow/stream/197730-Da-Vinci.20PDex.20Drug.20Formulary
Brandi Burgess (Jun 12 2020 at 19:51):
That's the site that confused me for the Formulary resources. Is it only the CoveragePlan and Formulary resources (2) to be sent?
Lloyd McKenzie (Jun 12 2020 at 19:52):
Please ask IG-specific questions in the stream that's specific to the IG that I provided above.
Brandi Burgess (Jun 12 2020 at 20:12):
The question I have is on the https://build.fhir.org/ig/HL7/davinci-pdex-formulary IG guide and if there are only 2 resources included for that IG guide. Which are as follows:
CoveragePlan:
FormularyDrug:
Michele Mottini (Jun 12 2020 at 20:32):
There are only two resources - but they are List and MedicationKnowledge. CoveragePlan is the name of the List profile, FormularyDrug is the name of the MedicationKnowledge profile. You should ask for details in https://chat.fhir.org/#narrow/stream/197730-Da-Vinci.20PDex.20Drug.20Formulary as Lloyd said
Last updated: Apr 12 2022 at 19:14 UTC
