# post-coordinated code on DV\_CODED\_TEXT?

**URL:** https://discourse.openehr.org/t/post-coordinated-code-on-dv-coded-text/16324
**Category:** Technical (archive)
**Created:** [21 March 2011 10:42 UTC](https://discourse.openehr.org/t/post-coordinated-code-on-dv-coded-text/16324 "2011-03-21T10:42:18Z")
**Posts on this page:** 4
**Page:** 1

<div class="post-metadata">

### Author: ![yampeku](https://discourse.openehr.org/user_avatar/discourse.openehr.org/yampeku/32/25_2.png) [@yampeku](https://discourse.openehr.org/u/yampeku)
#### Post date: [21 March 2011 10:42 UTC](https://discourse.openehr.org/t/post-coordinated-code-on-dv-coded-text/16324/1 "2011-03-21T10:42:18Z")

</div>

Hello,

So I was looking to the Text package in the data\_types specification  
and I was wondering how is supposed a DV\_CODED\_TEXT to handle a  
post-coordinated term (from for example SNOMED). If I understand  
correctly is something that relies in a supposed terminology service?  
isn't doing that making the system dependent of a technology or how a  
system understands some terminology queries? Can someone provide a XML  
snippet for the example “foot has-laterality left”?

---

<div class="post-metadata">

### Author: ![thomas.beale](https://discourse.openehr.org/user_avatar/discourse.openehr.org/thomas.beale/32/35_2.png) [@thomas.beale](https://discourse.openehr.org/u/thomas.beale)
#### Post date: [21 March 2011 11:18 UTC](https://discourse.openehr.org/t/post-coordinated-code-on-dv-coded-text/16324/2 "2011-03-21T11:18:33Z")

</div>

Diego,

that’s what the code\_string attribute is for - it carries a single concept code or a post-coordinated code-string, or ‘code phrase’, generated by the terminology service. It is not the job of an openEHR system to ‘understand’ these strings, just to store them faithfully. Currently (AFAIK) only SNOMED has a properly defined syntax for post-coordinations, but the code\_string could carry any string expression expressing a post-coordination from any terminology.

- thomas

---

<div class="post-metadata">

### Author: ![yampeku](https://discourse.openehr.org/user_avatar/discourse.openehr.org/yampeku/32/25_2.png) [@yampeku](https://discourse.openehr.org/u/yampeku)
#### Post date: [21 March 2011 11:40 UTC](https://discourse.openehr.org/t/post-coordinated-code-on-dv-coded-text/16324/3 "2011-03-21T11:40:24Z")

</div>

But then if we want to communicate this coded string outside from our  
openEHR system would be impossible to know 100% what it is. I don't  
get the point of having 'mappings' (which could also be done by a  
terminology service) and not having a 'qualifiers' or whatever.

Doesn't Loinc also allow post-coordination?

---

<div class="post-metadata">

### Author: ![thomas.beale](https://discourse.openehr.org/user_avatar/discourse.openehr.org/thomas.beale/32/35_2.png) [@thomas.beale](https://discourse.openehr.org/u/thomas.beale)
#### Post date: [21 March 2011 14:22 UTC](https://discourse.openehr.org/t/post-coordinated-code-on-dv-coded-text/16324/4 "2011-03-21T14:22:57Z")

</div>

> ```auto
> But then if we want to communicate this coded string outside from our
> openEHR system would be impossible to know 100% what it is.
> 
> ```

well don’t forget that the terminology\_id attribute is telling you what terminology it is, so it should always be the case that the code\_string is interpretable given the terminology. Now, if we start having multiple syntaxes or ways of expressing a post-coordination for a given terminology, something else is going to be needed.

> ```auto
> I don't
> get the point of having 'mappings' (which could also be done by a
> terminology service) and not having a 'qualifiers' or whatever.
> 
> ```

A mapping can be applied to a coded or uncoded data item, e.g. an ICD10 billing code attached to a much more specific SNOMED code, or just to a piece of text. Don’t forget, ‘coding’ is still largely done as a post hoc exercise in hospitals by special people who do only that, and who are essentially responsible for choosing the ‘right’ codes so that the hospital gets properly remunerated for the procedure done. So this is openEHR’s way of representing that. This kind of mapping can’t be done by a terminology service, because it is not a definitive mapping in the sense that you would find inside the terminology service - the latter kind of mapping might map the SCT concept ‘fracture of femur’ to an ICD code meaning the same thing, or as close as possible. But the ‘mapping’ attribute in openEHR allows the coders within a hospital (or the original consultant) to put a code like ‘long bone fracture’ chosen from ICD, which has been determined as the correct billing code.

> ```auto
> Doesn't Loinc also allow post-coordination?
> 
> ```

well LOINC is a kind of post-coordination system, with 6 axes. A LOINC code and a LOINC ‘post-coordination’ are more or less the same thing (you can have wildcards etc in LOINC).

hope this clarifies

- thomas
