# Use of RM:provider
**URL:** https://discourse.openehr.org/t/use-of-rm-provider/15475
**Category:** Technical (archive)
**Created:** [16 January 2017 13:14 UTC](https://discourse.openehr.org/t/use-of-rm-provider/15475 "2017-01-16T13:14:50Z")
**Posts on this page:** 9
**Page:** 1
### Author:  [@siljelb](https://discourse.openehr.org/u/siljelb)
#### Post date: [16 January 2017 13:14 UTC](https://discourse.openehr.org/t/use-of-rm-provider/15475/1 "2017-01-16T13:14:50Z")
Hi everyone,
I’ve got a problem about where to put non-identifying information about the source of information for an ENTRY. The value set we need to store is the code set identified by the OID 2.16.578.1.12.4.1.7498 (Source of information), as following:
1. Result of test/analysis
2. Observed by treating physician
3. Patient’s own information
4. Information from next of kin
5. Obtained from other records
6. Other
7. Reported by responsible clinician
Our first hypothesis was that the reference model element “provider” should be used for this, but then someone pointed out that the “provider” should be an identifiable person or entity, and not just a generalised coded text like this. So, where should this information go, if not in RM:provider?
Kind regards,
**Silje Ljosland Bakke**
Information Architect, RN
Coordinator, National Editorial Board for Archetypes
Nasjonal IKT HF
Tel. +47 40203298
Web: [http://arketyper.no](http://arketyper.no/) / Twitter: [@arketyper\_no](https://twitter.com/arketyper_no)
---
### Author:  [@system](https://discourse.openehr.org/u/system)
#### Post date: [17 January 2017 10:14 UTC](https://discourse.openehr.org/t/use-of-rm-provider/15475/2 "2017-01-17T10:14:36Z")
Hi Silje,
there is no immediate equivalent of these codes, which have indirect equivalents, i.e.
1. Result = OBSERVATION; provider = lab; limited to certain archetypes; also comes under SOAP ‘O’ Heading
2. Observed by physician = OBSERVATION, provider = PARTY\_IDENTIFIED (physician); limited to certain archetypes; also comes under SOAP ‘O’ Heading
3. Patient information = OBSERVATION, with provider = PARTY\_SELF, possibly limited to specific archetypes ; also comes under SOAP ‘S’ heading
4. Info from next of kin = OBSERVATION, with provider = PARTY\_IDENTIFIED (next of kin)
5. Info from other records = any ENTRY, with FEEDER\_AUDIT set appropriately
6. ??
7. Reported by responsible clinician could be 1, 2, most EVALUATIONs, most/all INSTRUCTIONs some or many ACTIONs; provider = PARTY\_IDENTIFIED (clinician)
I’d say it can be mostly set by using ENTRY.provider. 5 is a different thing - it’s about provenance of data obtained from elsewhere. Presumably 6 means ‘not any of 1-5 or 7’.
I’d also say it isn’t a very well designed code-set, and I don’t know what use it would be in real life…
- thomas
---
### Author:  [@system](https://discourse.openehr.org/u/system)
#### Post date: [17 January 2017 10:35 UTC](https://discourse.openehr.org/t/use-of-rm-provider/15475/3 "2017-01-17T10:35:43Z")
Hi Silje,
I would agree with your and Thomas’s assessment. This codeset does not really fit with provider, or indeed with any other RM attributes, although many but not all of these items could be calculated/ derived from existing attributes.
I guess this is part of a national requirement, and is a similar issue to the one we faced in Sweden, where the V-TIM standard was largely aligned with openEHR but had some extra specific metadata around Contsys-2 that needed to be captured.
This was exactly the purpose for the Extension slot that we are adding to new archetypes, so that would be my suggestion. Having said that, I do wonder about the purpose of this data -where is the value, over and above what is already captured by native openEHR RM. This feels like largely a derived set of data for reporting purposes
e,g,
> 1. Result of test/analysis
>
> We know that from the archetype name
>
> 1. Observed by treating physician
>
> Captured potentially by provider
>
> 1. Patient’s own information
> Party = Self
>
> 2. Information from next of kin
> Party = carer
>
> 3. Obtained from other records
>
> Can use Feed\_audit but practically very difficult to manage.
>
> 1. Other
>
> 2. Reported by responsible clinician
>
> Captured by composer
Can you tell us more about the background?
Ian
---
### Author:  [@siljelb](https://discourse.openehr.org/u/siljelb)
#### Post date: [17 January 2017 11:22 UTC](https://discourse.openehr.org/t/use-of-rm-provider/15475/4 "2017-01-17T11:22:50Z")
Thank you Thomas and Ian!
This is indeed a national requirement, and one where we do need to represent the chosen value in a coded text element. The background here is an entry in the critical information part of the national summary record, ie an adverse reaction, complication from anaesthesia, critical condition, ongoing treatment, implant, change of treatment routine, or infection. Each of these will be either an EVALUATION.adverse\_reaction\_risk, EVALUATION.problem\_diagnosis, or EVALUATION.precaution. The patient’s GP normally records the information, and this code set is supposed to be used to specify where the GP got the information about each of the entries from.
Regards,
**Silje**
---
### Author:  [@system](https://discourse.openehr.org/u/system)
#### Post date: [17 January 2017 12:34 UTC](https://discourse.openehr.org/t/use-of-rm-provider/15475/5 "2017-01-17T12:34:13Z")
Ideally there would be one or more classifiers at the ENTRY level, something that does not exist today. There are some others that we will include, e.g. relating to epistemic\_status.
I would follow Ian’s suggestion on the extension slot; it may be that the coding recorded there may need to be adjusted to a new place in relevant archetypes later on. Not ideal, but not a big problem either, assuming they are not used to create data before that is done.
Ian - do we have a related PR mooted for RM Release-1.0.4?
- thomas
---
### Author:  [@system](https://discourse.openehr.org/u/system)
#### Post date: [17 January 2017 12:42 UTC](https://discourse.openehr.org/t/use-of-rm-provider/15475/6 "2017-01-17T12:42:55Z")
Hi Thomas,
I’m not convinced as yet that this is a universally useful requirement, particularly as we carry much of this source/ provenance metadata already.
@Silje - are the GPs expected to add this extra information to every Entry in the summary? That seems like a significant burden, and actually in many cases unknowable.
Ian
---
### Author:  [@system](https://discourse.openehr.org/u/system)
#### Post date: [18 January 2017 10:24 UTC](https://discourse.openehr.org/t/use-of-rm-provider/15475/7 "2017-01-18T10:24:13Z")
Hi
The specified terminology ( OID 2.16.578.1.12.4.1.7498 (Source of information) ) is defined as “ **options to specify the source of data for an allergic reaction** ” (my translation).
Which means this is specific for adverse reaction – and I think it should be archetyped to model this requirement. If it is only national then there should be some extension slots in the Archetypes – or we need some specialization of the Archetype to handle this.
Vennlig hilsen
Bjørn Næss
Produktansvarlig
DIPS ASA
Mobil +47 93 43 29 10
---
### Author:  [@system](https://discourse.openehr.org/u/system)
#### Post date: [18 January 2017 11:33 UTC](https://discourse.openehr.org/t/use-of-rm-provider/15475/8 "2017-01-18T11:33:12Z")
Hi Bjorn,
Thanks - it makes much more sense in the context of Adverse reaction but TBH I still doubt very much if this ‘provenance’ source metadata is captured or known reliably. I asked a couple of UK GP colleagues and they agreed. I would argue that this data a) often not available b) unreliable c) a pain in the neck to manage and d) not something you ever want to do decision support on.
If an allergy has been asserted, it needs to be regarded as positive and kick decision support into life, no matter how vague the provenance or potentially unreliable the witness.
But hey, that’s what the extension slot in the archetype is for 🙂
Ian
---
### Author:  [@siljelb](https://discourse.openehr.org/u/siljelb)
#### Post date: [18 January 2017 15:03 UTC](https://discourse.openehr.org/t/use-of-rm-provider/15475/9 "2017-01-18T15:03:52Z")
Hi Bjørn and Ian!
The original use case for this code set was adverse reactions, but with the development of the national summary records, it’s been expanded for use with other kinds of critical information as specified below. This means that it needs to be used across the adverse reaction risk, problem/diagnosis and precaution archetypes, and I think therefore making a new cluster for use in the Extension slots is the best way to go forward.
@Ian: Yes, this information is mandatory for each entry. They’re generally few and seldom changed, so I don’t think there’s too much overhead from this:
Regards,
**Silje**