Open Source EHR at the Americal Academy of Family Physicians ...

Technical (archive)
1

Forgive me for a late comment, but I would like to clear the record at
little concerning hL7. HL7 is an accredited ANSI body, and follows ANSI
rules. Itis an organization that is not funded by any outside group. As
such, the organization deopends on membership for dues to support the
organization. I agree that totally free standards would be the best, and I
have argued that position within the HL7 Board. The best I have been able
to do is to get the draft available free. At the same time, release of a
new standard is always accompanied by an increase in membership. Also, I
must point out, that I think it is reasdonable for anyone who will gain
advantage from the standard to contribute. The proce for the standard is
just slightly greater than membership dues. Also, I point out that ISo and
ANSI sells standards - actually for more than HL7.

I hope you all will continue to support HL7 and its work. I always find it
interesting when people talk about HL7 as if it was a them and us. I hope
it is just us, and we struggle to support the tremendous cost of producing
the standard. I think the US may be the only country whose government does
not support the creation of standards - but at the same time, I don't think
that is all wrong.

We need the support of all the poeple who understand the value of standards
to work together and get the appropriate standards out there while the
stars are aligned. All of you are making important contributions.

Ed H

David Forslund <dwf@lanl.gov>@openehr.org on 08/20/2003 09:37:41 AM

Please respond to David Forslund <dwf@lanl.gov>

Sent by: owner-openehr-technical@openehr.org

cc:

       Physicians ...

2

Thank you Ed for clearing stating the HL7 position, which has always been clear, in my opinion.
It is understandable, but, as you suggest, it would be desirable that the standards once
they are complete be free. I'm not clear on what the licensing issues are in terms of redistribution
of those standards, however. I also agree that HL7 isn't alone in charging a fee and is probably
on the lower end of cost for those who do charge.

There are other models, of course, for handling standards within a non-profit organization, and it might be useful for HL7 to
explore those. I don't want to suggest that people not support HL7 and its important work because of the fee.

Thanks,

Dave

3

The recent agreement between the Health and Human Services and the College of American Pathologists about integrating SNOMED into UMLS, and making it available for free to everyone in USA, was a landmark.

Is there a thought process within HL7 that is exploring such opportunities? If HHS agrees to support HL7 to allow it to make its standards available for free, it will hasten its adoption and development while it serves the goals of the federal government too.

There is a need to bring into sync UMLS and HL7 at some level. To my mind Semantic Network and HL7 V3 RIM have to be reconciled. This will facilitate reuse in an object oriented way while retaining semantic validity. We can then have a true unified health information infrastructure.

Regards,

Hemant

4

Dear colleagues,

5

I agree with Gerard that we need to be careful. However, that does not
mean that we go to the lowest denominator. IF we think SNOMED is the best
solution, then we need to spend our time and energy on finding how to make
SNOMED available to the rest of the world. We have a debate in our school
system in Durham. The poorer kids do not have access to the Internet and
to laptops. The debate is whether to prohibit the use of computers and
Internet for school work or to try to find methods that will provider
laptops and Internet access to the poorer kids. I think the answer is
simple.

However, I do think it is important to make sure that SNOMED is the answer
and will be acceptable before we move aggressively.

Ed Hammond

6

In een bericht met de datum 25-9-2003 15:10:09 West-Europa (zomertijd), schrijft hammo001@mc.duke.edu:

I agree with Gerard that we need to be careful. However, that does not
mean that we go to the lowest denominator. IF we think SNOMED is the best
solution, then we need to spend our time and energy on finding how to make
SNOMED available to the rest of the world. We have a debate in our school
system in Durham. The poorer kids do not have access to the Internet and
to laptops. The debate is whether to prohibit the use of computers and
Internet for school work or to try to find methods that will provider
laptops and Internet access to the poorer kids. I think the answer is
simple.

However, I do think it is important to make sure that SNOMED is the answer
and will be acceptable before we move aggressively.

Ed Hammond

If you have any questions about using this list,
please send a message to d.lloyd@openehr.org

I agree with Ed in that if we can make this resource available, we need to work on that.
I think it is OK that HL7 uses SNOMED as preferred terminology.

However, I would be very dissapointed if this would become the only terminology that the current v3 RIM and derivates could handle. I believe also local, or specialty or situation specific terminologies / vocabs etc. should be allowed in messages.

But maybe I am overreacting, I did not hear / read that this would not be the case.

William Goossen

7

William,

I do not think you are over reacting. I agree with you. My only point is
that we should be driven by what is best and what is a true solution, and
not by the wrong reasons. I would be most interested in seeing us compile
a list of candidates for terminologies that should be considered and a
process by which we could blend the terminologies.

I don't know what the best method might be and what organization(s) might
be best for doing the work and distributing the product. What is the level
of trust for the NLM around the world?

Ed

8

Hi all,
Just wanted to let folks know that at the “Computational Level” the work of the OMG Healthcare Domain Task Force’s Lexicon Query Service (LQS) address the issue of multiple terminologies and their use in IT settings. This technology is being used today as a mediator between systems that utilize different terminologies which makes it possible to use legacy and new terminologies. The concept is that there will never be 1 terminology system in the world due to human, technological and legacy systems but you can narrow the focus and then utilize technology (LQS) to assist you. The specification was designed and development in an international setting with many of the prominent terminology experts as contributors. As HL7 moves forward at the “Informational Level” maybe they can glean some insights from LQS in terms of working with multiple terminologies.

Tom

9

Hi Gerard,

Appreciate your post. It confirms a suspicion of mine that a workable global solution, regardless of topic, is one where compatibility, interoperability and usability are prime concerns, e.g., the drive for globalization is modified so that common goals and objectives with workable interfaces are targets.

To justify this recall that in the US we are still on the English system of measurement rather that the metric system simply because a majority of the populace considers metric measurement more difficult and a cheat.

One can purchase a set to tools for the the automobile in English or Metric, and perhaps both. The various legislatures in the US have backed off many efforts to drive one of the other.

H7 is a good effort. However, considerable time, effort and resources can be wasted attempting to derive a common standard. As long as the different systems interface well, why bother.

Our politicians are still attacking those countries, cultures, people, etc that opposed the invasion of Iraq. Efforts to get the population to accept 'French Fries' are still failing. Heard a business brief that indicates that sales of these food items are still down from prior levels.

Charge ahead with the European approach and develop good interfaces.

-Thomas Clark

Gerard Freriks wrote:

10

Hi William,

A small addition:

1)It seems unlikely that Practitioners, Patients and associated parties would be UNABLE to effectively communicate without speaking SNOMED.

2)It seems likely that a single, dominating system for handling day-to-day Healthcare-related activities it become efficient on a global scale.

3)Rigorous testing, including scalability, of SNOMED seems to be sparse:

PERFORMANCE; Google search: "SNOMED performance |"
http://etbsun2.nlm.nih.gov:8000/publis-ob-offi/pdf/2000-tal-ob-Ft.pdf
(1 hit)

SCALABILITY: Google search: "SNOMED scalability |"
(no hits)

EFFECTIVENESS: Google search: "SNOMED effectiveness |"
(no hits)

RELIABILITY: Google search: "SNOMED reliability |"
(no hits)

AVAILABILITY: Google search: "SNOMED availability |"
http://quickstart.clari.net/qs_se/webnews/wed/bx/Bga-mckesson-info-sols.Rn1s_Dl9.html
(1 hit); DIFFERENT KIND OF 'availability', i.e., availabile for use

COMPLAINTS: Google search: "SNOMED complaint |"
(no hits)

ERRORS: Google search: "SNOMED error |"
(no hits)

SUSTAINABILITY: Google search: "SNOMED sustain |"
(no hits)

OK! I give up!

SNOMED, it appears, has never been subjected to any kind of analysis. It appears to be in the same category as home repair contractors who provide an on-the-spot 'tail-light' warranty.

To roll on this one and push it on the global healthcare community needs some justification I can't provide.

-Thomas Clark

Williamtfgoossen@cs.com wrote:

11

Ed,

I agree with you.
Today I had an discussion with Diane Ashman on this topic.
She is very willing to think along those lines.
But we all must move with caution, think of the many consequences and find
the proper balance.

Gerard

-- <private> --
Gerard Freriks, arts
Huigsloterdijk 378
2158 LR Buitenkaag
The Netherlands

+31 252 544896
+31 654 792800

12

Only a tiny percentage of the biomedical literature is accessible to
Google - you need to search PubMed - see
http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed

A completely naive search on "SNOMED evaluation" yielded 33 hits, most
of which we on topic. More extensive searches would flush out a lot more
papers, I'm sure.

BTW, please talk about "SNOMED CT" which is the original SNOMED
classification combined with a clinical terms, including the Read codes.

Personally I think that SNOMED CT is far from perfect (as has been
discussed on this list in the past), but it is the best show in town for
the near and medium future. The fact that there is a universal license
to use it in the US, and some form of NHS-wide license in the UK, makes
it attractive to software developers. I understand that serious
consideration is being given to negotiating a national license for
SNOMED CT for Australia, but there is some due process to be gone
through first before a decision is made.

Tim C

13

Hi Tim,

Pieces of the 33 hits are included below:

-Sarcomatoid carcinoma of the cervix
-An evaluation of the usefulness of two terminology models for integrating nursing diagnosis concepts into SNOMED Clinical Terms
-Improved coding of the primary reason for visit to the emergency department using SNOMED
-Which coding system for therapeutic information in evidence-based medicine
-Automating SNOMED coding using medical language understanding: a feasibility study
-An evaluation of the utility of the CEN categorical structure for nursing diagnoses as a terminology model for integrating nursing diagnosis concepts into SNOMED
-Semantic features of an enterprise interface terminology for SNOMED RT
-Evaluation of a method that supports pathology report coding
-Evaluation of SNOMED3.5 in representing concepts in chest radiology reports: integration of a SNOMED mapper with a radiology reporting workstation
-Representation by standard terminologies of health status concepts contained in two health status assessment instruments used in rheumatic disease management
-An evaluation of ICNP intervention axes as terminology model components
-[Medical data in pathology--evaluation of a large collection. (530,000 diagnoses coded in SNOMED II)]
-Scalable methodologies for distributed development of logic-based convergent medical terminology
-The role of peer review in internal quality assurance in cytopathology
-Evaluation of a "lexically assign, logically refine" strategy for semi-automated integration of overlapping terminologies
-Phase II evaluation of clinical coding schemes: completeness, taxonomy, mapping, definitions, and clarity. CPRI Work Group on Codes and Structures
-The surgical pathologist in a client/server computer network: work support, quality assurance, and the graphical user interface
-Comparison of the reproducibility of the WHO classifications of 1975 and 1994 of endometrial hyperplasia
-Planned NLM/AHCPR large-scale vocabulary test: using UMLS technology to determine the extent to which controlled vocabularies cover terminology needed for health care and public health
-Mass screening for cervical cancer in Norway: evaluation of the pilot project
-The LBI-method for automated indexing of diagnoses by using SNOMED. Part 2. Evaluation
-Representing HIV clinical terminology with SNOMED
-The LBI-method for automated indexing of diagnoses by using SNOMED. Part 1. Design and realization
-A comparison of four schemes for codification of problem lists
-Can SNOMED International represent patients' perceptions of health-related problems for the computer-based patient record?
-Extraction of SNOMED concepts from medical record texts
-Terms used by nurses to describe patient problems: can SNOMED III represent nursing concepts in the patient record?
-[Descriptive epidemiology from autopsies at the Ospedale Maggiore di Milano from 1986 to 1987]
-[Development of a findings and results data system for forensic medicine autopsy cases]
-Medical linguistics: automated indexing into SNOMED
-Evaluation of the CAP microcomputer-based SNOMED encoding system
-[A new microglossary for biopsy pathology]

None of these hits can be related in any significant way to to the implementation and deployment of a system with SNOMED functionality, i.e., based wholly on SNOMED or integrating it as a plug-in or an integral function.

My original posting included some major review topics typically encountered in a software product design (the focus immaterial).

There is an old saying where I come from:
Quiting playing with the design and produce something before the competition does.

Design, develop, deploy sustain and upgrade later.

The motivation to charge for SNOMED may well prompt competition to action . Right now, in my opinion, SNOMED needs relevant Google/developer entries.

Additional comments in your text.

Thanks!

-Thomas Clark

Tim Churches wrote:

3)Rigorous testing, including scalability, of SNOMED seems to be sparse:

PERFORMANCE; Google search: "SNOMED performance |"
http://etbsun2.nlm.nih.gov:8000/publis-ob-offi/pdf/2000-tal-ob-Ft.pdf
(1 hit)

SCALABILITY: Google search: "SNOMED scalability |"
(no hits)

EFFECTIVENESS: Google search: "SNOMED effectiveness |"
(no hits)

RELIABILITY: Google search: "SNOMED reliability |"
(no hits)

AVAILABILITY: Google search: "SNOMED availability |"
http://quickstart.clari.net/qs_se/webnews/wed/bx/Bga-mckesson-info-sols.Rn1s_Dl9.html
(1 hit); DIFFERENT KIND OF 'availability', i.e., availabile for use

COMPLAINTS: Google search: "SNOMED complaint |"
(no hits)

ERRORS: Google search: "SNOMED error |"
(no hits)

SUSTAINABILITY: Google search: "SNOMED sustain |"
(no hits)

OK! I give up!

SNOMED, it appears, has never been subjected to any kind of analysis. It appears to be in the same category as home repair contractors who provide an on-the-spot 'tail-light' warranty.
   
Only a tiny percentage of the biomedical literature is accessible to
Google - you need to search PubMed - see
http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed

A completely naive search on "SNOMED evaluation" yielded 33 hits, most
of which we on topic. More extensive searches would flush out a lot more
papers, I'm sure.

BTW, please talk about "SNOMED CT" which is the original SNOMED
classification combined with a clinical terms, including the Read codes.

None of the 33 hits listed above refers to SNOMED CT. This looks like an issue that should be resolved.

Personally I think that SNOMED CT is far from perfect (as has been
discussed on this list in the past), but it is the best show in town for
the near and medium future. The fact that there is a universal license
to use it in the US, and some form of NHS-wide license in the UK, makes
it attractive to software developers. I understand that serious
consideration is being given to negotiating a national license for
SNOMED CT for Australia, but there is some due process to be gone
through first before a decision is made.

Tim C

Have the current set of licensees subjected SNOMED CT to an IT WORLD review, analysis and performance/evaluation? If so, did they publish?

-Thomas Clark

14

William E Hammond wrote:

However, I do think it is important to make sure that SNOMED is the answer
and will be acceptable before we move aggressively.

Ed - how will this happen - what process can be followed to do this? Do you mean "clincally" acceptable, or acceptable in terms of licencing, $ conditions?

I personally have great doubts that any one refernce terminology can be the "one answer" to everything. All the work going on with archetypes, RMIMs etc at the moment shows quite clearly that the meaning of any term in a specfic context is often (usually) not the meaning of the same work in a reference terminiology (which by definition almost, must have a kind of compromise definition of its meaning). So even if all the licencing and access issues are sorted out to everyone's satisfaction, I don't believe that the final solution has been reached. This comment is not specific to Smomed of course - it is a general principle.

- thomas beale

15

Williamtfgoossen@cs.com wrote:

In een bericht met de datum 25-9-2003 15:10:09 West-Europa (zomertijd), schrijft hammo001@mc.duke.edu:

I agree with Ed in that if we can make this resource available, we need to work on that.
I think it is OK that HL7 uses SNOMED as preferred terminology.

However, I would be very dissapointed if this would become the only terminology that the current v3 RIM and derivates could handle. I believe also local, or specialty or situation specific terminologies / vocabs etc. should be allowed in messages.

I agree - I would state even more strongly - I don't think it can be any other way. Recently, Sam did a review of our models of "Apgar result" (your favourite;-) and discovered that the terms used for various things on US and UK websites were different (e.g. the terms used for the 0,1,2 values for each of the 5 input variables). No single global terminology can deal with this problem - only capsule terminologies which are strongly bound to particular concepts can.

- thomas beale

16

William E Hammond wrote:

William,

I do not think you are over reacting. I agree with you. My only point is
that we should be driven by what is best and what is a true solution, and
not by the wrong reasons. I would be most interested in seeing us compile
a list of candidates for terminologies that should be considered and a
process by which we could blend the terminologies.

I don't know what the best method might be and what organization(s) might
be best for doing the work and distributing the product. What is the level
of trust for the NLM around the world?

Peter Elkin (Mayo) claims to have identified about 40 or so candidate terminologies for use in an open terminology system, according to his paper in MIE 2003.

But I don't believe the correct methodology in this area has yet surfaced. It will start to when small, targetted knowledge models start being used more widely, and terminologists start to see that there is no solution based on the idea of a "single , perfect holy grail terminology". It just doesn't work like that. There are capsules of meaning everywhere which link back into ontologies, and I think that a theory and methodology based on this idea will begin to surface in the next few years. Snomed-ct will be then seen as a best effort without this theory, and may end up being the biggest single resource for re-enginering into a new typology of terminologies / ontologies / small knowledge models (archytpes, HL7 models, guidelines etc).

Right now I really think people need to understand that there is still a lot of intellectual work ot go in this area, and that finalising licencing situations will not particularly change things.

- thomas beale

17

I basically agree. I think I mean both clinical and economical. What I am
hoping for is that we can create a single process in which all the
appropriate terminologies can be blended, overlaps and mapping, and
distribution made common. Do it once not each institution or even each
country.

I would like to establish a core terminology group that is international
that works toward this goal.

Ed

18

And theoretical health informaticists need to understand that the
absence of a widely available termonology/classification is badly
hurting real-life efforts to improve and protect health, right now. I
don't particularly like SNOMED CT - its bulky and inelegant (although
fairly comprehensive), and as Thomas points out, uses way too much
pre-coordination. But from where I sit, as a practicing epidemiologist
who works with practicing clinicians, we need a terminology now. As I
said, SNOMED CT seems to be the best bet, at least for English-speaking
countries, and the license costs at the national level - US$32 million
for the whole US for 5 years, presumably rather less for, say, all of
Australia - are not unsustainable, and at least SNOMED is essentially a
non-profit organisation, not a rapacious multinational corporation.

But efforts on open terminologies, both niche and global, should
definitely continue. Hopefully SNOMED CT can then be replaced in a
decade or so with a free, global alternative.

19

Dear Ed,
I fully agree with you. We do need an international vocabulary. We need to
make translations to other languages and it is no so easy to convince the
ones who pay the bill that to translate SNOMED ( for example) to Portuguese
should be done. If this is an international effort with many other countries
aligning maybe we can try to find funds together. The sooner the better. At
the moment we are defining a new vocabulary for health procedures - sort of
Brazilian CPT...
Best regards,
Beatriz

20

Tim

But efforts on open terminologies, both niche and global,
should definitely continue. Hopefully SNOMED CT can then be
replaced in a decade or so with a free, global alternative.

We shouldn't wait for a decade or so. We need it now. In the UK, SNOMED-CT
is a done deal, and it is the door opener to Europe. These things have
their own momentum, and before we know it, the SNOMED juggernaut will roll
over everywhere, and the not-for-profit (who said it was?) organisation will
become just as dominating as any other 'globlaisation-embracing' full
blooded organisation.

We need this open terminology now. What Gerard descibed many years ago as
"MedSpeak".

Ahmad Risk
http://hi-europe.info