Generic observation?

[Charles Thompson]

I'm developing an HTML form and I need an concept that takes a coded response that I can answer with any other concept. Does OpenMRS have something like this?

[Rowan Seymour]

The idea of a coded concept is that the concept defines a set of valid answers. I think (could be wrong) that the API doesn't actually enforce that so you can actually save any concept as the answer to an coded question. But that's not good practice

What kind of question can be answered by any other concept? Perhaps you need a free text obs?

On 28 February 2013 15:00, Charles Thompson <char...@gmail.com> wrote:

I'm developing an HTML form and I need an concept that takes a coded response that I can answer with any other concept. Does OpenMRS have something like this?

--
OpenMRS Implementers: http://go.openmrs.org/implementers
Post: implem...@openmrs.org
Unsubscribe: implementers...@openmrs.org
Manage your OpenMRS subscriptions at https://id.openmrs.org/
 
 

--
Rowan Seymour
tel: +250 783835665

[Paul Biondich]

Hi Charles. :)

Yes.  In fact, this feature is one of the primary foundations of OpenMRS.

Create a "question" concept, give it a "coded" datatype, and it's answer can in fact be any other concept.

So, for example:  one concept called "urine color" can take coded concept answers of "clear", "serosanguinous", etc.

Hope this helps,

-Paul

On Thu, Feb 28, 2013 at 8:00 AM, Charles Thompson <char...@gmail.com> wrote:

I'm developing an HTML form and I need an concept that takes a coded response that I can answer with any other concept. Does OpenMRS have something like this?

--

[Paul Biondich]

Actually, Rowan:  specifying no codes allows all codes to be valid.

-Paul

[Charles Thompson]

In my paper form, I have a field labeled "Adverse effects of medication" where the provider can indicate coded response such as Nausea, Diarrhea, or Fatigue. Is there already a concept for this?

[Rowan Seymour]

@Charles we're using the CIEL dictionary and I'm not seeing a concept with that exact meaning

@Paul it looks like that that even for concepts that do specify a coded list of answers, you can save any concept as the answer. That seems a little odd

[Paul Biondich]

Yes, not sure why that behavior is there.  Pretty sure I intended the code to restrict answers to the predefined set when it was defined, and left it open ended when no concepts were defined.

@Burke:  does this make sense to you?  We probably should file a ticket if this is happening.

-Paul

[Michael Seaton]

The ship may have sailed on this, if you intend to file a ticket.  Many, many forms have been written that contain the desired coded answers to questions, but for which the concept dictionary was never modified to contain those answers in the answer list.  So this will introduce lots of incompatibility, similar to the validation issues that have plagued implementations in the 1.9 upgrade.

Personally, I'd like to see us think more about our concept answer approach as a whole.  For example, thinking about whether it makes sense to use the existing Concept Set construct to build re-usable answer sets, and then to be able to associate one or more "answer sets" with a concept.  Then on your form you would choose which answer set you want to use, rather than have to continually pick a few answers that are relevant out of the world of possible answers.

Mike

[Burke Mamlin]

We don't have very good validation for observations and, as a result, many implementations have succumb to entropy and will have messy data.

Our intent was to eventually support answer_set & answer_class, which would allow concept answers to be constrained by set(s) or class(es) along with specific concepts.

-Burke

[Darius Jazayeri]

Charles,

To answer you original question, yes.

See https://wiki.openmrs.org/display/docs/HTML+Form+Entry+Module+HTML+Reference#HTMLFormEntryModuleHTMLReference-Coded

-Darius