XDW classCode and typeCode

46 views
Skip to first unread message

Dave Franken

unread,
Dec 12, 2014, 7:23:10 AM12/12/14
to ihe-xdw-im...@googlegroups.com
Is there any update on the LOINC codes for XDW?

Is this in motion yet?

Regards,

Dave

Dave Franken

unread,
Dec 19, 2014, 2:26:09 AM12/19/14
to ihe-xdw-im...@googlegroups.com
An update for other interested parties.

Mauro Zanardini has contacted me about this and started looking at it again, putting some wheels in motion.
If we're lucky, we'll have something in a few weeks.

I'm still a bit surprised this hasn't been done yet...

Parisot, Charles (GE Healthcare)

unread,
Dec 21, 2014, 7:09:23 PM12/21/14
to dfra...@meddex.nl, ihe-xdw-im...@googlegroups.com

I am wondering if we did not make a mistake in placing this code into the ClassCode.

 

I believe that this Workflow document should rather be in the DocumentType Code.

 

Two reasons:

 

LOINC codes are always used for Doc Type Codes in all other profiles and mandated by CDA.

 

The class code should speak to workflow as a class of documents that may contains several Doc Types:

1 – XDW docs of Type Code (to be given by LOINC)

2 – Refrenced document that are not clinical in nature, but workflow such as referral request letters, referral responses, request for additional information, etc.

 

What do people think of this approach ?

What would be the consequences for the existing implementations ?

Should I prepare a CP ?

 

Charles

--
You received this message because you are subscribed to the Google Groups "ihe-xdw-implementors" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ihe-xdw-implemen...@googlegroups.com.
To post to this group, send email to ihe-xdw-im...@googlegroups.com.
Visit this group at http://groups.google.com/group/ihe-xdw-implementors.
For more options, visit https://groups.google.com/d/optout.

Mark Sinke

unread,
Dec 22, 2014, 2:42:42 AM12/22/14
to Charles Parisot, dfra...@meddex.nl, ihe-xdw-im...@googlegroups.com
Hi folks,

I think putting it into class code is correct here. Workflow documents are of a very different nature than referral letters etc. They have little significance to the end user directly, whereas the referral letters etc. do. So Charles' argument about these being the same "class" does not hold IMHO.

From a practical perspective, it is pretty useful to be able to determine XDW documents from the rest of the documents, regardless of the specific type of XDW document, and still to be able to see the concrete type of document based on the typeCode (although formatCode should indicate that as well, if we used typeCode to hold the XDW code).

I do however have a strong opinion about the implication of using typeCode: IHE should then also specifiy the classCode. I do not think it makes sense to specify a more specific code without specifying the coarser-grained code.

In fact, that makes us circle back to the beginning of the argument; only we need two codes now, instead of one :-).

Regards, Mark 

PS As XDW is still in trial implementation, practical implementations will be able to adopt.


Mark Sinke

CTO


m +31 6 46 255 635

p  +31 30 699 19 30

e  mark....@forcare.com

w www.forcare.com

forcare-logo.png


Parisot, Charles (GE Healthcare)

unread,
Dec 22, 2014, 3:11:01 AM12/22/14
to Mark Sinke, dfra...@meddex.nl, ihe-xdw-im...@googlegroups.com

Mark,

 

We need to a more serious discussion.

Your analysis is way to centered on XDW.  Class Code is so critical that it has to be designed having a long-term and braod perspective on all types of documents that will eventually have to be managed in an XDS Registry.

Looking at it, one use case at a time has resulted into many projects around the world into having to redesign entirely their Class Code strategy.

 

Here is the list of class codes that are slowly but surely emerging:

Reports

Summaries

Images

Prescribed Treatments and Diagnoses

Dispensations

Treatment Plan or Protocol

Health Certificates and Notifications

Patient Expression and Preferences

Workflow Management

 

WE need to remember that the power of an element of metadata reduces its efficacy, when the classification becomes fuzzy.

That is why there is only 9 values in this example.  Having more than 15-18 values is a top maximum.

 

So it might not be very wise to follow your route of such a fine grained distinction: putting one class code only for XDW documents.

In the class code of workflow management, of course XDW documents will be placed, but there are several other workflow management related documents.

 

Having the need to segregate XDW document is an important need, but we should use the right tool for that:

- either a document type code

- or the format code: that one will be unique for all XDW documents.è that is the one I would recommend to support your need.

 

In any case, this is a very important discussion, whose time has come.

 

Charles

Boone, Keith W (GE Healthcare)

unread,
Dec 22, 2014, 10:15:25 AM12/22/14
to <mark.sinke@forcare.com>, Parisot, Charles (GE Healthcare), dfra...@meddex.nl, ihe-xdw-im...@googlegroups.com
From recentbwork that I've doneit feels right for an AFFINITY DOMAIN to define a class code to organize its documents (e.g., clinical vs, workflow) but I would not feel good about IHE enforcing classCode AT ALL. That has always been under affinity domain control.

Keith

On Dec 22, 2014, at 2:42 AM, "Mark Sinke" <mark....@forcare.com<mailto:mark....@forcare.com>> wrote:

Hi folks,

I think putting it into class code is correct here. Workflow documents are of a very different nature than referral letters etc. They have little significance to the end user directly, whereas the referral letters etc. do. So Charles' argument about these being the same "class" does not hold IMHO.

From a practical perspective, it is pretty useful to be able to determine XDW documents from the rest of the documents, regardless of the specific type of XDW document, and still to be able to see the concrete type of document based on the typeCode (although formatCode should indicate that as well, if we used typeCode to hold the XDW code).

I do however have a strong opinion about the implication of using typeCode: IHE should then also specifiy the classCode. I do not think it makes sense to specify a more specific code without specifying the coarser-grained code.

In fact, that makes us circle back to the beginning of the argument; only we need two codes now, instead of one :-).

Regards, Mark

PS As XDW is still in trial implementation, practical implementations will be able to adopt.


________________________________

Mark Sinke

CTO


m +31 6 46 255 635

p +31 30 699 19 30

e mark....@forcare.com<mailto:mark....@forcare.com>

w www.forcare.com<http://www.forcare.com>


[forcare-logo.png]



On Mon, Dec 22, 2014 at 1:09 AM, Parisot, Charles (GE Healthcare) <charles...@med.ge.com<mailto:charles...@med.ge.com>> wrote:
I am wondering if we did not make a mistake in placing this code into the ClassCode.

I believe that this Workflow document should rather be in the DocumentType Code.

Two reasons:

LOINC codes are always used for Doc Type Codes in all other profiles and mandated by CDA.

The class code should speak to workflow as a class of documents that may contains several Doc Types:
1 – XDW docs of Type Code (to be given by LOINC)
2 – Refrenced document that are not clinical in nature, but workflow such as referral request letters, referral responses, request for additional information, etc.

What do people think of this approach ?
What would be the consequences for the existing implementations ?
Should I prepare a CP ?

Charles

From: ihe-xdw-im...@googlegroups.com<mailto:ihe-xdw-im...@googlegroups.com> [mailto:ihe-xdw-im...@googlegroups.com<mailto:ihe-xdw-im...@googlegroups.com>] On Behalf Of Dave Franken
Sent: Friday, December 19, 2014 8:26 AM
To: ihe-xdw-im...@googlegroups.com<mailto:ihe-xdw-im...@googlegroups.com>
Subject: [ihe-xdw-implementors:65]] Re: XDW classCode and typeCode

An update for other interested parties.

Mauro Zanardini has contacted me about this and started looking at it again, putting some wheels in motion.
If we're lucky, we'll have something in a few weeks.

I'm still a bit surprised this hasn't been done yet...

On Friday, December 12, 2014 1:23:10 PM UTC+1, Dave Franken wrote:
Is there any update on the LOINC codes for XDW?

Is this in motion yet?

Regards,

Dave
--
You received this message because you are subscribed to the Google Groups "ihe-xdw-implementors" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ihe-xdw-implemen...@googlegroups.com<mailto:ihe-xdw-implemen...@googlegroups.com>.
To post to this group, send email to ihe-xdw-im...@googlegroups.com<mailto:ihe-xdw-im...@googlegroups.com>.
Visit this group at http://groups.google.com/group/ihe-xdw-implementors.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "ihe-xdw-implementors" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ihe-xdw-implemen...@googlegroups.com<mailto:ihe-xdw-implemen...@googlegroups.com>.
To post to this group, send email to ihe-xdw-im...@googlegroups.com<mailto:ihe-xdw-im...@googlegroups.com>.
Visit this group at http://groups.google.com/group/ihe-xdw-implementors.
For more options, visit https://groups.google.com/d/optout.


--
You received this message because you are subscribed to the Google Groups "ihe-xdw-implementors" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ihe-xdw-implemen...@googlegroups.com<mailto:ihe-xdw-implemen...@googlegroups.com>.
To post to this group, send email to ihe-xdw-im...@googlegroups.com<mailto:ihe-xdw-im...@googlegroups.com>.

Mauro Zanardini

unread,
Dec 22, 2014, 12:08:35 PM12/22/14
to Keith Boone, <mark.sinke@forcare.com>, Parisot, Charles (GE Healthcare), dfra...@meddex.nl, ihe-xdw-im...@googlegroups.com
I think that we should start from two points: 

1) classCode should be the most generic content classification of the document. We should discourage the definition of a bunch of codes at this level. 

2) we still need to take in consideration the use-cases approach. formatCode is correctly fixed to the urn:ihe:iti:xdw:2011:workflowDoc value. This allows a consumer to identify the format of the document (XDW schema). An XDW Content Consumer should be able to query for every Workflows related to a patient (e.g. a system that wants to follow clinical paths for research purposes OR for monitoring purposes…). Another XDW Content Consumer would participate in a specific workflow (e.g. a Consultant). Both the use-cases need to be supported by XDW. 

I think that the choice done when we drafted XDW was the right one: 

classCode: code that identifies a generic Workflow Document

typeCode: code that identifies the specific clinical Workflow described in the Workflow Document. This should be related to the workflow Definition profile supported by XDW actors involved in the clinical workflow. 

LOINC could be the right place to formalize these classifications...

I do not see reasons to leave these codes to Affinity Domain definition…

Mauro

To unsubscribe from this group and stop receiving emails from it, send an email to ihe-xdw-implemen...@googlegroups.com.
To post to this group, send email to ihe-xdw-im...@googlegroups.com.



--
Dott. Mauro Zanardini
Project Engineer

cell. +393346482818

Consorzio Arsenàl.IT
Centro Veneto Ricerca e Innovazione per la Sanità Digitale
Viale Oberdan, 5
31100 Treviso

E-mail: mzana...@consorzioarsenal.it

Boone, Keith W (GE Healthcare)

unread,
Dec 22, 2014, 12:46:11 PM12/22/14
to Mauro Zanardini, <mark.sinke@forcare.com>, Parisot, Charles (GE Healthcare), dfra...@meddex.nl, ihe-xdw-im...@googlegroups.com

Your affinity domain and mine differ on what the code should be.  How do I know?  Because I have one affinity domain already that has defined a class code for this using a local vocabulary.  They put all workflow related documents with the same  class code.  But some of these documents I know of already another affinity domain would use a clinical code for (e.g., referral request).  I would be very careful on specifying this too hard.  The document type should certainly be IHE specified, but I would leave class code for the affinity domain.

 

                Keith

Dave Franken

unread,
Jan 3, 2015, 7:13:15 AM1/3/15
to ihe-xdw-im...@googlegroups.com, mzana...@consorzioarsenal.it, mark....@forcare.com, charles...@med.ge.com, dfra...@meddex.nl
XDS-I complicates this discussion, for KOS objects it reserves the type code to describe the procedure (using an affinity domain specific vocabulary), meaning we can only put the LOINC code for KOS objects in class code or choose to not put it anywhere.

From: ihe-xdw-im...@googlegroups.com<mailto:ihe-xdw-implementors@googlegroups.com> [mailto:ihe-xdw-im...@googlegroups.com<mailto:ihe-xdw-implementors@googlegroups.com>] On Behalf Of Dave Franken
Sent: Friday, December 19, 2014 8:26 AM
To: ihe-xdw-im...@googlegroups.com<mailto:ihe-xdw-implementors@googlegroups.com>
Subject: [ihe-xdw-implementors:65]] Re: XDW classCode and typeCode

An update for other interested parties.

Mauro Zanardini has contacted me about this and started looking at it again, putting some wheels in motion.
If we're lucky, we'll have something in a few weeks.

I'm still a bit surprised this hasn't been done yet...

On Friday, December 12, 2014 1:23:10 PM UTC+1, Dave Franken wrote:
Is there any update on the LOINC codes for XDW?

Is this in motion yet?

Regards,

Dave
--
You received this message because you are subscribed to the Google Groups "ihe-xdw-implementors" group.

To unsubscribe from this group and stop receiving emails from it, send an email to ihe-xdw-implementors+unsub...@googlegroups.com<mailto:ihe-xdw-implementors+unsub...@googlegroups.com>.
To post to this group, send email to ihe-xdw-i...@googlegroups.com<mailto:ihe-xdw-implementors@googlegroups.com>.


Visit this group at http://groups.google.com/group/ihe-xdw-implementors.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "ihe-xdw-implementors" group.

To unsubscribe from this group and stop receiving emails from it, send an email to ihe-xdw-implementors+unsub...@googlegroups.com<mailto:ihe-xdw-implementors+unsub...@googlegroups.com>.
To post to this group, send email to ihe-xdw-i...@googlegroups.com<mailto:ihe-xdw-implementors@googlegroups.com>.


Visit this group at http://groups.google.com/group/ihe-xdw-implementors.
For more options, visit https://groups.google.com/d/optout.


--
You received this message because you are subscribed to the Google Groups "ihe-xdw-implementors" group.

To unsubscribe from this group and stop receiving emails from it, send an email to ihe-xdw-implementors+unsub...@googlegroups.com<mailto:ihe-xdw-implementors+unsub...@googlegroups.com>.
To post to this group, send email to ihe-xdw-i...@googlegroups.com<mailto:ihe-xdw-implementors@googlegroups.com>.

Visit this group at http://groups.google.com/group/ihe-xdw-implementors.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "ihe-xdw-implementors" group.

To unsubscribe from this group and stop receiving emails from it, send an email to ihe-xdw-implementors+unsub...@googlegroups.com.
To post to this group, send email to ihe-xdw-i...@googlegroups.com.

Parisot, Charles (GE Healthcare)

unread,
Jan 5, 2015, 8:21:04 AM1/5/15
to dfra...@meddex.nl, ihe-xdw-im...@googlegroups.com, mzana...@consorzioarsenal.it, mark....@forcare.com

Dave, Mauro,

 

You do not need a LOINC document type code in XDS Metadata for all use cases.

Having in XDS-I a Radiology Procedure Code is more precise: “radiology report for procedure XXX” and important to select relevant prior.

 

I like Mauro’s approach, in term of requirements:

I think that we should start from two points: 

 

1) classCode should be the most generic content classification of the document. We should discourage the definition of a bunch of codes at this level. 

 

2) we still need to take in consideration the use-cases approach. formatCode is correctly fixed to the urn:ihe:iti:xdw:2011:workflowDoc value. This allows a consumer to identify the format of the document (XDW schema). An XDW Content Consumer should be able to query for every Workflows related to a patient (e.g. a system that wants to follow clinical paths for research purposes OR for monitoring purposes…). Another XDW Content Consumer would participate in a specific workflow (e.g. a Consultant). Both the use-cases need to be supported by XDW.

  

classCode: code that identifies a generic any Workflow related Documents (XDW being one of those).

 

typeCode: code that identifies the specific clinical Workflow described in the Workflow Document. This should be related to the workflow Definition profile supported by XDW actors involved in the clinical workflow. 

 

LOINC could be the right place to formalize these classifications (TypeCodes for XDW workflow documents).  Not sure the Workflow Class Code should be defined by LOINC.  This is a class of documents, not a doctype.

 

I would like to explain my point, that there should NEVER be  a ClassCode reserved for a specific document. 

This seems to contradict the  principle 1) above.

However, it is important to support the function to query for a specific patient for all XDW workflow documents:

This should be done by querying for :

·         DocClass=            Workflow Management

·         FormatCode:     urn:ihe:iti:xdw:2011:workflowDoc

 

By supporting the query only on DocClass =  Workflow Management, one gets all active XDW documents and associated referenced document that are workflow related (e.g. appointment, workflow management responses).

 

How does that sound ?

 

Charles

From: ihe-xdw-im...@googlegroups.com<mailto:ihe-xdw-im...@googlegroups.com> [mailto:ihe-xdw-im...@googlegroups.com<mailto:ihe-xdw-im...@googlegroups.com>] On Behalf Of Dave Franken
Sent: Friday, December 19, 2014 8:26 AM
To: ihe-xdw-im...@googlegroups.com<mailto:ihe-xdw-im...@googlegroups.com>
Subject: [ihe-xdw-implementors:65]] Re: XDW classCode and typeCode

An update for other interested parties.

Mauro Zanardini has contacted me about this and started looking at it again, putting some wheels in motion.
If we're lucky, we'll have something in a few weeks.

I'm still a bit surprised this hasn't been done yet...

On Friday, December 12, 2014 1:23:10 PM UTC+1, Dave Franken wrote:
Is there any update on the LOINC codes for XDW?

Is this in motion yet?

Regards,

Dave
--
You received this message because you are subscribed to the Google Groups "ihe-xdw-implementors" group.

To unsubscribe from this group and stop receiving emails from it, send an email to ihe-xdw-implemen...@googlegroups.com<mailto:ihe-xdw-implemen...@googlegroups.com>.
To post to this group, send email to ihe-xdw-i...@googlegroups.com<mailto:ihe-xdw-im...@googlegroups.com>.


Visit this group at http://groups.google.com/group/ihe-xdw-implementors.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "ihe-xdw-implementors" group.

To unsubscribe from this group and stop receiving emails from it, send an email to ihe-xdw-implemen...@googlegroups.com<mailto:ihe-xdw-implemen...@googlegroups.com>.
To post to this group, send email to ihe-xdw-i...@googlegroups.com<mailto:ihe-xdw-im...@googlegroups.com>.


Visit this group at http://groups.google.com/group/ihe-xdw-implementors.
For more options, visit https://groups.google.com/d/optout.


--
You received this message because you are subscribed to the Google Groups "ihe-xdw-implementors" group.

To unsubscribe from this group and stop receiving emails from it, send an email to ihe-xdw-implemen...@googlegroups.com<mailto:ihe-xdw-implemen...@googlegroups.com>.
To post to this group, send email to ihe-xdw-i...@googlegroups.com<mailto:ihe-xdw-im...@googlegroups.com>.

Visit this group at http://groups.google.com/group/ihe-xdw-implementors.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "ihe-xdw-implementors" group.

To unsubscribe from this group and stop receiving emails from it, send an email to ihe-xdw-implemen...@googlegroups.com.


To post to this group, send email to ihe-xdw-i...@googlegroups.com.
Visit this group at http://groups.google.com/group/ihe-xdw-implementors.
For more options, visit https://groups.google.com/d/optout.



 

--

Dott. Mauro Zanardini

Project Engineer

 

cell. +393346482818

Consorzio Arsenàl.IT
Centro Veneto Ricerca e Innovazione per la Sanità Digitale
Viale Oberdan, 5
31100 Treviso

E-mail:
mzana...@consorzioarsenal.it

--

You received this message because you are subscribed to the Google Groups "ihe-xdw-implementors" group.

To unsubscribe from this group and stop receiving emails from it, send an email to ihe-xdw-implemen...@googlegroups.com.
To post to this group, send email to ihe-xdw-im...@googlegroups.com.

Mauro Zanardini

unread,
Jan 7, 2015, 6:50:22 AM1/7/15
to Parisot, Charles (GE Healthcare), dfra...@meddex.nl, ihe-xdw-im...@googlegroups.com, mark....@forcare.com
Charles, 

I like your idea. 

So, i propose to proceed in this way: 

-I'll ask LOINC cmte for a set of codes that shall be used to classify (typeCode) Workflow Documents related to specific Workflow Definition Profiles in TI: Telemonitoring , eReferral, Tumor Board,  etc.

-Done that,  I'll submit a CP for XDW and for related workflow definition profiles. 

Do you Agree ? 

Mauro

 
Reply all
Reply to author
Forward
0 new messages