FHIR Chat · ATNA AuditMessage to FHIR AuditEvent · implementers

Stream: implementers

Topic: ATNA AuditMessage to FHIR AuditEvent


view this post on Zulip Veliyan Georgiev (Oct 28 2020 at 17:52):

Hello experts. I was wondering if any of you have had the need to translate ATNA (DICOM) AuditMessage to FHIR AuditEvent. The mapping is here https://www.hl7.org/fhir/auditevent-mappings.html#dicom but I'm curious if someone has actually done it and potential feedback associated with it.

view this post on Zulip Vassil Peytchev (Oct 28 2020 at 17:57):

Have you looked at the IHE supplements using AuditEvent?

view this post on Zulip John Moehrke (Oct 28 2020 at 18:13):

If you have recommendations, I would be glad to field them. I manage the specifications in IHE, HL7, and DICOM on this topic.

view this post on Zulip Veliyan Georgiev (Oct 28 2020 at 19:14):

so ... here is my challenge..and it is probably more technical than anything else. For legacy apps (non-FHIR) we use ATNA (DICOM) AuditMessage.. for FHIR ones -> FHIR AuditEvent. The challenge with the ATNA one is parsing it out later. Here is an example -> in a hosted AWS app those events are generated from a K8S cluster/pod through stdout (which BTW there is a 16K limit on a container stdout entry that is a separate topic) and then picked up and stored in CloudWatch. The tools that are available are notoriously bad when dealing with XML (i.e. Grok, Regex etc). The options are -> 1. do ETL in the cloud (it will cost money) or convert from ATNA to FHIR JSON (either via transform or directly from code.
So nothing wrong with the spec/standard really - just curious if someone has tried to convert ATNA to FHIR and how well it went. I;'m trying to avoid having the non FHIR apps change internal logic to generate FHIR events when they already have code to generate ATNA

view this post on Zulip John Moehrke (Oct 28 2020 at 22:24):

I think @Grahame Grieve may have done that

view this post on Zulip John Moehrke (Oct 28 2020 at 22:25):

you might also find some help on the IHE discussion forums - like https://groups.google.com/g/ihe-xds-implementors

view this post on Zulip Veliyan Georgiev (Nov 05 2020 at 13:26):

I couldn't find what i needed so I decided to write a library yesterday. It is here currently but I'll put it on NPM soon https://github.com/upmc-enterprises/ATNA2FHIR
I started validating the generated resource and I'm having a predicament with ".recorded" The instructions here https://www.hl7.org/fhir/auditevent-mappings.html#dicom don't have anything on it but it is a required field. I see the case for being a timestamp when the event is generated but why not being the same as ".period" - equal to the DICOM's EventDateTime since it is mapped?

view this post on Zulip John Moehrke (Nov 05 2020 at 13:46):

the concept of .recorded is not found in the DICOM schema. you could replicate the eventDateTime, but it is not the same concept. I would expect your tooling would just set .recorded to the time that you are creating the FHIR AuditEvent instance... as that is more aligned with the meaning of .recorded.

view this post on Zulip John Moehrke (Nov 05 2020 at 13:47):

I would very much like to promote your open source project on this. thanks for doing this

view this post on Zulip Veliyan Georgiev (Nov 05 2020 at 13:58):

That makes sense. Thanks John. Much appreciated. I could use a sanity check when it is somewhat done in the next few days.

view this post on Zulip Veliyan Georgiev (Nov 10 2020 at 20:58):

@John Moehrke npm install @upmc-enterprises/atna2fhir promote away :-)

view this post on Zulip John Moehrke (Nov 10 2020 at 21:08):

could you describe in the readme what problem this is trying to solve? I would like to promote something that is self promoting.

view this post on Zulip Grahame Grieve (Nov 12 2020 at 10:17):

i did it once but the code is long gone

view this post on Zulip Veliyan Georgiev (Nov 12 2020 at 14:10):

@John Moehrke readme updated and added bunch of more mappings. Thanks for the feedback. Keep it coming


Last updated: Apr 12 2022 at 19:14 UTC