Sources of information on HL7 EHR/OpenEHR

In een bericht met de datum 15-9-2006 22:21:23 West-Europa (zomertijd), schrijft gfrer@luna.nl:

snip snip:

I agree.
There is only one patient, with one problem that needs our unified attention and devotion.
So we have to co-operate.
But we have to continue to discuss and provide arguments and listen to the arguments given.
Instead of attacking persons, as I have been able to observe several times it to happen in the Netherlands.

Lets start the real debate.
Patients and healthcare providers need real solutions that empower them.

Gerard

I agree with several comments on HL7 v3, there are solutions for that underway.
I agree that CEN 13606 / Open EHR and HL7 v3 have their advantages and disadvantages.

I agree that we can work together
I agree that there is a lot to be done
I agree that both approaches have led to implementations and also to the determination of problems in the technological approach.

I disagree that we should discuss this from a view point of superiority of one approach against the other. That is the WW 1 approach.
That WW1 ‘for or against’ approach will not lead to working solutions.

My comments are based on believing that your points 1-6 are sensible to discuss and find solutions.

Your comment 7 is not useful and I would suggest you to refrain from the pro - con discussion.

For the one patient with always more problems (there is no situation in health care where a patient has only one problem, he might have one disease, but that will always have more than one interelated problems) that need our dispersed attention to tackle this in a holistic way.

So the ‘standard’ approach is:

one patient with
one or more diseases
and none to many associated problems / complaints
and none to many associated nursing diagnoses / allied health problems
and one to many activities for the disease (s)
and one to many activities for the associated problems (the minimum, so always one has to be there is to monitor the onset of a potential problem)
and one to many activities for the associated nursing diagnoses / allied health problems.

Definitely this is is with respect to workflow a difficult non-linear process of care delivery and of changes of the disease(s) and the different problem situations.
Thus we need to be able to have a system that can track what these changes are, and what the status of this set of mixed activities is. We must also be able to exchange this in the multitude of systems available in health care that will last for several decades.

I am pretty sure that I can deal with these complex situations in HL7 v3 speak. I have not seen examples of these complex care situations in 13606 speak since that is missing examples from health care with this respect.

Perhaps we need to work more from the clincial perspective and determine the requirements and from there to the technical bits.

William

In een bericht met de datum 15-9-2006 23:15:45 West-Europa (zomertijd), schrijft randy.neall@veriquant.com:

Since I’m not sufficiently acquainted with all the details of either system I could not follow each nuance of the argument. But among the things I’m taking away from this is that HL7 involves a complex and ever-changing schema at the DB storage level, something that worries me. I did not hear a rebuttal of this point from the HL7 side.

I am not a specialist on this, but I believe it is possible to have a quite simple DB schema based on HL7 v3.

Difficult parts include the Entity parts (person, patient, organisation) because the entries in fields will change for some and not for other (e.g. changing addresses, phones etc.

relationships between entities are handled via roles: that is not changing.

participation from a role in activities are not changing. Of course there are different kinds of participation that change, but that is doable with coding

then there are acts that have relationships with other acts, identifiers, moodcode and status codes, as are codes, times and where the act is an observation: there is a value.

These parts of the RIM and D-MIMs have been stable for quite a while.

Key of Act storage is to store the code for the act, the time, the author (via participation / role relationships), the mood (request, promise plan, carried out) and the status (activive, obsolete, competed).

These things have not changed in principle.

However, the D-MIMs derived from RIM have changed in those domains where development is ongoing. For some implementers it means for instance that the condition class cannot be used anymore. Instead we work with a problem list allowing activities to be associated on problem level and not on ‘condition’ level. Condition is changed to an observation.

Part of this is because not all components of the HL7 v3 standard are ready now. Early implementers of work in progress thus will face changes that affect their developments.
For the Netherlands national ICT project we include the industry in the work in progress and discuss this situation with them before starting and during development and balloting of standards.

To prevent too many changes we have now the draft standard for trial use DSTU that allows a couple of years fixation of the standard, testing, finding out the problems, revising and then develop the final standard and ballot that.

William Goossen

Heath

Are there people interested in providing clinical scenarios and data to assist in the requirements and content gathering process to be used in clinical workflow simulations? How can we initiate and progress this kind of activity and investigation?

Try contacting NICS - the National Institute of Clinical Studies http://www.nicsl.com.au/asp/index.asp
They may be interested in assisting with this work.

Gordon Tomes
Projects Director
Acute Care Division
Department of Health and Ageing (MDP 63)
PO Box 9848, Canberra ACT 2601
Ph 02 6289 5081 | Mobile 0423 024 922 | Fax 02 6289 7630

2006 Biennial Health Conference
Exploring and debating acute care provision
14-16 November, Sydney, Australia

for the latest information visit http://www.healthconference.com.au/

“Heath Frankel” heath.frankel@frankelinformatics.com
Sent by: openehr-technical-bounces@openehr.org

15/09/2006 12:04 PM
Please respond to For openEHR technical discussions


|
To: “‘For openEHR technical discussions’” openehr-technical@openehr.org
cc:
Subject: RE: Sources of information on HL7 EHR/OpenEHR |

  • | - | - |

Gerard,
Interesting you raise this topic as it is becoming an interest of mine to start investigating the use of openEHR instructions to support the documentation requirements of clinical workflows such as medication prescriptions, dispense and administration, and referrals. The existing work of ContSys could certainly assist in this but being a techo, I need some clinicians to assist in developing these requirements and my Ocean clinician colleagues are already over extended. As you know, Ocean has and continues to develop the tools to support a simulation of these kind of clinical workflow scenarios and are looking for ways to gather more and varied clinical content to populate and test the OceanEHR suite. Are there people interested in providing clinical scenarios and data to assist in the requirements and content gathering process to be used in clinical workflow simulations? How can we initiate and progress this ki! nd of activity and investigation?

Regards

Heath

Heath Frankel
Product Development Manager
Ocean Informatics
Ground Floor, 64 Hindmarsh Square
Adelaide, SA, 5000
Australia

ph: +61 (0)8 8223 3075
fax: +61 (0)8 8223 2570
mb: +61 (0)412 030 741
email: heath.frankel@oceaninformatics.biz

Williamtfgoossen@cs.com wrote:

So the 'standard' approach is:

one patient with
one or more diseases
and none to many associated problems / complaints
and none to many associated nursing diagnoses / allied health problems
and one to many activities for the disease (s)
and one to many activities for the associated problems (the minimum,
so always one has to be there is to monitor the onset of a potential
problem)
and one to many activities for the associated nursing diagnoses /
allied health problems.

Definitely this is is with respect to workflow a difficult non-linear
process of care delivery and of changes of the disease(s) and the
different problem situations.
Thus we need to be able to have a system that can track what these
changes are, and what the status of this set of mixed activities is.
We must also be able to exchange this in the multitude of systems
available in health care that will last for several decades.

I am pretty sure that I can deal with these complex situations in HL7
v3 speak. I have not seen examples of these complex care situations in
13606 speak since that is missing examples from health care with this
respect.

Hi William,

the workflow aspects of the above need workflow languages and service
models to support multi-enterprise choreography. There are guideline and
workflow languages (not provided by HL7 or openEHR), and the beginnings
of models for choreography coming from WfMC and other places. I can't
think of much HL7 provides in this area. But the main question with
respect to workflow is: what needs to be automated? Humans are far
better at performing real-world tasks than machines, due to their
ability to handle exceptional cases. Mostly what we need to solve right
now is:
- clinical content
- simple workflow around medication prescription and management
- some basic workflow around admission, discharge, referral

Other workflows are wonderful problems for computer science theses, but
mostly are done better and more economically by human beings at the
moment. If we just get clinical content solved, that will be a huge step.

Perhaps we need to work more from the clincial perspective and
determine the requirements and from there to the technical bits.

exactly...

- thomas

Hello,
I am completely okay with T Baele.
We have in France a comparable debate with the DMP project (Personal Medical
record).
The physician needs information under shape of documents.He/it wants them
quickly and sure being that the content is reliable to be able to take a
decision or to prescribe a treatment.
For he/it is asked to the physician to be the supplier of data to put medicine
in a system of information for decision-makers, payers or political. It is
another probléme.....

Dr R LONJON france

Selon Thomas Beale <Thomas.Beale@oceaninformatics.biz>:

In een bericht met de datum 18-9-2006 10:45:04 West-Europa (zomertijd), schrijft Thomas.Beale@OceanInformatics.biz:

There are guideline and
workflow languages (not provided by HL7 or openEHR), and the beginnings
of models for choreography coming from WfMC and other places.

I have looked into the WfMC materials, and the basic process flow descriptions are currently met with the HL7 v3 Care Plan. (This is not a point if HL7 can do, it is the point that it is possible to define the clinical process using a standard, I think it is transferable to OpenEHR archetype as well).

The key here is the use of the following mood codes:
definition will tell you wat according to best practice or evidence base should be done for a patient with problem x. (including monitoring of observations, tests, meds etc).

The OpenEHR template specification that links archetypes could perhaps do similar things.

intent mood helps the clinician to carry over from guideline into the care plan what is necessary for individual patient P.
Thus the set of data required can be determined, and it can be justified why items are not carried from guideline to plan. (E.g. you do not female things for a male patient).

Then if some professional wants to order a observation this can be done with request. e.g. the doctor askes the nurse to measure the blood pressure 4 times a day.

In the Goal mood, the expected value can be set, e.g. the expected value of BP in a week should best be 130/90.

the observatoin is carried out say 7 days 4 times a day leading to 7 x 4 = 28 observations in event mood.

The statement collecter allows to trend this.

The comparison of goal versus the event(s) trends, or the last value of day 7 allows to determine if the goal is met (conclusion being then the 29th observation).

The derivation method allows to specify also workflow rules like:
do BP measurements until 4 x < 130/90 or similar as a criterion for the do X until Y workflow standard.

I am not telling this is best handled in HL7 v3, I just want to say that a] it is possible to express clinical meaningful workflows, that at EHR level are pretty handy for a nurse to pop up on the worklist every 6 hours, and that it is possible to exchange the semantics of such a workflow / careplan via a message.

Yes, this is interesting stuff and needs a lot of work.

William

this is true. What is there now:
* CEN HISA
* emerging openEHR service models for EHR, demographics, terminology access and
archetype access
* state-based process management for Instructions, i.e. medications, orders,
procedures.
* high-level HL7 HSSP specifications like RLUS etc
* older and probably undervalued Corbamed specifications, like PIDS
    

and the HL7 v2 and HL7 v3 event models.

sorry, I should have included those as well. I have a reservation that
they seem too pre-defined, and the real world just constantly throws up
exceptions, but I will be guided by others with more experience with
these models.

yep, they are. but 90/10 rule, they are very useful and they
are proven to cover 90% so worth considering.

Grahame

William,
There are still a few missing links in the V3 Care Plan model in regard to work flow modelling, some of which are supported in openEHR and others that are left to more dedicated work flow modelling and tooling, deliberately.

The key issue in the V3 Care Plan is the lack of work flow decision points. The event criterion (EVN.CRT) mood is only supported on the Observation Act in the Clinical Statement and hence the Care Plan model, therefore you can not model decision points that are not Observations like Procedures and Substance Administrations, which I would have thought to be rather key. This is not to downgrade HL7, just to correct your statement that these can be done in HL7. It is true that it could be done but not with the current Care Plan Clinical Statement model.

The second issue is the use of derivation expression. This is an interesting attribute that is just a catch all that is not already supported. So we can do some modelling using RMIMs but then what we can’t say we put into derivation expression but then what is the language to use here? HL7 doesn’t have an endorsed constraint language (although it has a standard for one, GELLO) and it also doesn’t have work flow language. So to actually make what you suggest can be done in HL7 you need to invent your own rule expressions.

My main issue here is not that they don’t have these, but that you have more than one way of representing these. To build software on this you need to be able to process the HL7 event criterion and the derivation expressions in what ever invented language you used. That doesn’t seem to be overly interoperable. As you said it’s not perfect.

openEHR has drawn the line and said, we will not reinvent the wheel on Work Flow. openEHR will provide the data required, record the states and the recommendations but not the actual rules used by software. It is intended to use existing work flow engines in conjunction with an openEHR repository. Although, I understand that there might be a future effort to develop an abstract model of work flow that work with archetypes that could be mapped into the existing Work Flow models and engines.

Now this needs to be proven and worked through and as the original Author of the HL7 V3 Care Plan I am very interested in working this through in openEHR. The HL7 V3 Care Plan model is just to cumbersome even though I agree that the mood code did allow for a very powerful representation of a Guideline and Care Plan. I just don’t think we can reliably write software to process those models within the development budgets that we have in Healthcare, even in the UK, and certainly not in Australia.

Considering the amount of money spent in the UK (I don’t know the Netherlands budget) on building an EHR repository which is nothing more than a message store using V3 compared with what we have achieved with a very small team to build an openEHR repository, I suspect we can extrapolate this into clinical work flow processing. It is this cost of implementation (not that it can’t be done, but how much it costs to do it) that has turned me away from HL7 V3 whereas the more I do with openEHR the more I am amazed how easy things are to implement.

Again, I didn’t want to beat up on HL7 V3, but I did want to present my experience being someone that is intimate with both sides of the fence and up until recently sat on the fence but I am certainly getting blown to one side where the implementation experience is more favourable. And on a final note, the CEN 13606 based RMIM I developed that was the cyclone that hit me considering you need an A3 page to represent something that fits on an A4 page in UML, just imagine if we represented HL7 V3 models in UML :>. And there are so many issues that HL7 still have not solved around versioning, attestation, templates, term bindings and queries.

I really do hope that we can collaborate in a two-way exchange. As Gerard mentioned, there are agreements to collaborate with HL7 but that presently appears to be one-directional and it would be a nice change if HL7 parties begin actively participating the development of knowledge artefacts in openEHR/CEN space. Now I know there is an initiative in the US to trial this but I wonder if the likes of yourself could start getting deeper into the Instruction model and how it will work for care planning and make an objective assessment between the two approaches. We don’t have enough people that are fluent in both and those that are get caught up in time wasting religious debates. I think you said it a couple of time, we need to determine what the requirements first, but we never get to do this because we argue about it it can or can’t be done in a particular technology.

I know the archetype and template tools don’t support the requirements gathering you require but we should have an archetype repository that is able to store additional meta-data fairly soon. There is nothing wrong using Word and Excel as you are now. What we need is equivalent openEHR archetype for each of your Care Statement RMIMs and in your mapping spreadsheet a couple of columns for the openEHR archetype mappings. Once we get the process right we can then develop the tools to support it.

BTW, a member of my development team (who was a obstetrician) is going through the process of developing a pregnancy clinical scenario (mega storyboard) and mapping the data element and sample data into archetypes. I wonder of you would be interested in working with her or at least sharing your experiences and current process?

Regards

Heath

Heath Frankel
Product Development Manager
Ocean Informatics

Ground Floor, 64 Hindmarsh Square
Adelaide, SA, 5000
Australia

ph: +61 (0)8 8223 3075
fax: +61 (0)8 8223 2570
mb: +61 (0)412 030 741
email: heath.frankel@oceaninformatics.biz

ok, we have real convergence here.

OpenEHR works exactly like HL7 - define a reference model
with all the needed semantics, and then refine things away
in constraint models (and use the refinements as a basis for
composition). So the principle is the same.

We can generate [class models|schemas|wire formats] from
constraint patterns. Doing so has benefits and costs. The
same benefits and costs in either HL7 or OpenEHR. And one
of the clear costs relates to persistence. I think we need
to search for a better way, but that's not going to happen
in the short term.

But the OpenEHR & HL7 reference models are quite different.
In most parts, the HL7 reference model is more abstract
(Which is way Act gets 22 attributes). So harmonising between
the reference models is going to require actual change rather
than adroitly altering perspectives, as with data types and
the constraint model things.

This will be hard, and painful. And it must involve compromise,
so this is when we find out who really values collaboration.

And we don't want to take away the real benefits that OpenEHR
has in the process (same for HL7)

Grahame

Grahame,

One of the theoretical difficulties of the debate that you are contributing
so eloquently to is to separate methods proper to computing from methods
defined by domain specificity.
Object-oriented programming, modular design, abstract data types, etc. are
fundamental principles whose existence does not follow from any particular
domain features. They lead to better information modelling in terms of
architecture, computational semantics and have been adopted as the best
engineering methodology to create information processing systems.
The abstractions that bridge the gap between computing and domain-driven
modelling have to be able to represent in machine processable form the
business objects we are dealing with. What you are doing in this case is
less important than what you are doing it to. The marriage of the two
strands does not deny their individuality but produces an offspring that
blends their essential features in an organic way.

openEHR contains RECORD in its name as the basic abstraction. The RECORD is
a container structure that has well defined computational semantics - data
entry is structured in the form of compositions that may be of different
type (instruction, observation, etc.) because of differences in computing
requirements - data representation, etc.
Thus we have a model of a record (document) management system which at its
highest level is generic. However, its type system is geared towards the
faithful representation of information in the health care domain with the
goal of producing a shared longitudinal patient record.

HL7 as the name attests is based on the messaging paradigm. The MESSAGE has
been historically the modelling abstraction that defines HL7. Thus, we have
to ask ourselves, which root abstraction fits better the twin criteria of
clear and purposeful computational semantics and faithful representation of
domain specifics.

You said:

ok, we have real convergence here.
OpenEHR works exactly like HL7 - define a reference model
with all the needed semantics, and then refine things away
in constraint models (and use the refinements as a basis for
composition). So the principle is the same.

(I have the feeling that you are thinking of HL 7 V4.0)
Convergence is a really strong statement. I understand your motivation and
your honest desire to overcome the gap between openEHR and HL7. However, one
has to be equally honest in appraising the
two approaches and their results.

openEHR doesn't "work exactly like HL7". A brief look at the history of the
two would show that while openEHR has always been based on sound software
engineering and knowledge modelling principles, HL7's effort to a reference
model to its messaging structures is not necessarily a product of organic
development, hence th gap between version 2 and version 3.

openEHR does define a reference model with all the needed computational
semantics but leaves domain-specific knowledge modelling to the second level
of its methodology - archetype modelling.
Implementation efforts have clearly shown openEHR's ability to produce clean
clean computational models that enhance domain-specific semantic
interoperability first and foremost throught the strict enforcement of the
separation of concerns between information modelling and knowledge
representation. The two sides click together at run-time in a very efficient
and effective way.

This is not refining things away in constraint models.
The fundamental difference is how one deals with modelling or how one uses
abstraction.

"Abstraction means to strip away the superficial or incidental aspects of a
thing and to reveal its most important aspects, or essence, aka theory,
hypotheses. Abstraction is important and can lead to deeper (beyond surface
understanding ) of things.
Abstraction can also go wrong. There are different ways to abstract. How do
we test whether our abstractions (theories, hypotheses) are right or wrong?"
(to quote a good blog:
http://billkerr2.blogspot.com/2006/08/ascending-from-abstract-to-concrete.html).

The key to using abstraction in theoretical modelling is to understand how
one can "ascend from the abstract to the concrete." I don't think
constraining and refining actually play such an important role in the design
process.

The same blog:

"Marx talked of "ascending from the abstract to the concrete" which on the
surface can seem rather absurd.
How can a concrete view of reality be superior to a more abstract view? We
all know that science takes us beyond the surface appearance of things and
proposes deeper explanations that are not immediately apparent. And anyone
who has studied child development knows that initially children view the
world in a "concrete" way and as they become able to think better they
become capable of thinking more abstractly. So isn't abstract thinking more
advanced than concrete thinking?
The difficulty arises from confusing the Marxist idea of "the concrete" with
the idea that what is concrete is that which is easily and immediately
perceived via the sense organs - ie the surface appearance of things. (note:
the whole idea of anything being "immediately perceived" is in any case
incorrect. Perception always involves some cognitive processing ).
However what Marx meant by "concrete" was not what is meant by the
pedagogical distinction between concrete and abstract (or formal) thinking.
Marx's use of the word "concrete" has to do with the notion of truth and is
not related to the idea of concrete thinking as child-like and primitive.
To ascend from the abstract to the concrete means to move from the initial
ability to abstract away from surface appearance (via everyday and
relatively easy generalisations and simple concepts) toward a richer and
more accurate view of concrete reality. This does involve abstraction - but
it is abstraction that is on its way to a richer and more concrete
(truthful) world view."

You don not find the richer and more truthful world view through
constraining and refining. You have two processes at work: the first one is
maintaining valid and verifiable data structures that exist within the type
system of your information model; the second one deals with concrete
instances of data and information from the clinical world that live in your
software system, their machine processable nature is defined by computer
semantics and platform specificity but their meaning is not.

The concrete clinical notions modelled via archetypes are thus something
more than just a constrained pattern. These produce concrete instances,
objects much richer in content than the abstract model whose elements they
represent.

The trick is to accomplish this ascent seamlessly within your defined
universe on the basis of a stable single view of that universe. Your
knowledge model represents the domain-specific semantics your are dealing
with and it will exist with or without an underlying computational system.
But machine processing requires that we have a working system even when our
knowledge acquisition process is ongoing.

To sum up - information models are based on abstractions and data types
chosen because of their computational semantics and suitability to
faithfully represent instances of domain-specific business objects. The
latter derive their meaning from the theoretical (clinical) body of
knowledge and the only information processing constraint is that they be in
a machine computable form, that is, they do not break information model and
the sharing of data between system.

Archetypes represent concrete reality as machine processable entities that
co-exist in the same knowledge universe. This universe is open to future
study and representation. What is important in openEHR's view of the world
is to abide by the open/closed principle which means that our models should
be closed in a sense that they can work today while leaving open the system
boundaries for future extension (sic!).

We can generate [class models|schemas|wire formats] from
constraint patterns. Doing so has benefits and costs. The
same benefits and costs in either HL7 or OpenEHR. And one
of the clear costs relates to persistence. I think we need
to search for a better way, but that's not going to happen
in the short term.

Persistence is orthogonal to design and modelling. Persistence closure
requires that all object references be stored and then reproduced when
necessary (it doesn't need to know about actual object creation). It also
involves efficient query mechanisms but again, we have two sides of the same
story - generating class models, schemas, wire formats is a case of clean
computational semantics while creating, storing and quering of knowledge
structures (archetypes) has to do with domain-specific knowledge. I cannot
comment on the costs of persistence but recent trends in database theory
point in the direction of knowledge-aware persistence mechanisms.

But the OpenEHR & HL7 reference models are quite different.
In most parts, the HL7 reference model is more abstract
(Which is way Act gets 22 attributes). So harmonising between
the reference models is going to require actual change rather
than adroitly altering perspectives, as with data types and
the constraint model things.

I don't think that one can discuss whose model is more abstract (or bigger).
What interests me is whether the two models have the right abstractions for
the task, and whether principles of information and knowledge modelling have
been put to good use. Actual change should contribute to the goal of greater
expressiveness and focused coverage rather than to the elusive goal of
harmonisation.

This will be hard, and painful. And it must involve compromise,
so this is when we find out who really values collaboration.

Blood, sweat and tears? In the name of what? What do you mean by
compromise?
What is the nature of compromise in scientific research? This is a term that
belongs in areas such as politics and conflict resolution. In research, the
middle ground, as Goethe would say, is where conflict begins (my apologies
for the imprecise quote).

Who really values collaboration? Again, strong words, Grahame.
What is the nature, program and goals of such collaboration? Is it the
creation of an open Electronic Health Record based on computational and
domain-specific semantic interoperability?
What is the mechanism of this collaboration? I'd be interested to see the
specific research program and see how emerging joint efforts lead to
uncompromising collaboration in the name of a clearly defined final goal.

And we don't want to take away the real benefits that OpenEHR
has in the process (same for HL7)

I agree,

Ogi Pishev

Graham,

Isn 't is a matter of scope, as the thing we need to have agreement on first?
Isn 't is a matter of requirements, as the thing we need to have agreement on second?
Will the rest of the harmonisation process follow from that?

I think that harmonising two things with substantial different scopes and requirements is impossible.

Gerard

– –

Gerard Freriks, arts

Huigsloterdijk 378

2158 LR Buitenkaag

The Netherlands

T: +31 252 544896

M: +31 653 108732

William
I think targets are more complex than setting a mood code, but I might be wrong. The Goal is 130/90 - is it greater than this or less than this? What if the BP is 40/20 - does this meet the goal - how does the computer know?
It is hard to get this right - but very important if we are to make use of computers.
Cheers, Sam