ISO/TC215 Health Informatics Working Group 1
Health Records and Modelling Coordination

Sydney, 21-22 January 1999

Report by Michael Mair FRACS, Delegate for New Zealand

Contents:


back to contents page

Introduction

This report should be read in conjunction with the first report on the ISOTC/215 process. It is about the first 'modelling' work group meeting of that committee. The reader may recall that at the earlier Orlando meeting in August 1998, the 'territory' of the global electronic medical record standard was divided up into four main work groups, and the convenor role for them allocated. These work groups were:

In addition, informal or 'ad hoc' work groups for Health Cards and Imaging were formed. The next full session of the ISO/TC215 committee is in Berlin in April 1999, and will be immediately preceded by meetings of the work groups, which will report to the plenary session. Already, the next WG1 meeting after Berlin has been agreed for London on 16-17 September 1999, and the following plenary session is in Japan in November 1999. From there, the committees and work groups continue their exotic global perambulation.

There was no advance documentation for this meeting beyond the minutes of the previous meeting, a repeat of the first cut at the 'scope statement' previously outlined in Orlando, an agenda, and an invitation to contribute 'standards' (see Addendum 1). We had already made a submission to the original committee offering our own pre standard (The New Zealand Electronic Medical Record Standard Draft v 1.06, 25 Feb 1998). I made a further brief submission (Addendum 2 below) to Peter Williams who convenes the WG1 committee. The invitation was to make a general comment, to think big, to go for the grand concept (or at least I interpreted it that way). On preparing my submission, I circulated it to the others in our 'Delegates' Groups and received one comment from David Menkes complaining appropriately about the grammar. Nobody suggest I write off to Australia to alter or cancel it so it went ahead. It was available at the work group meeting in the modest documentation provided, accompanied by the whole of the New Zealand draft standard version 1.06.

We had prepared for our Work Group Role by assembling a 'Delegates Group'. These included: Mike Mair (Ophthalmologist, Timaru); David Menkes (Psychiatrist, Otago); Ashwin Patel (GP, Auckland); Dick Whiddett (Information Technology, Massey); Lindsay Stewart ('Change Controller' MoH); Patricia Kerr (Chair, NZ Health Informatics Foundation). Ian Nottage, Standards Director for the NZHIF, has also effectively joined the group now. As it turned out, we have more 'delegates' for this work group than any other nation. However, our circumstances are quite different from any other nation and our involvement in the ISOTC/215 process may come to have a special significance for us in that we may use it as the vehicle to achieve our own national standard health network.

We had discussed what line to take at the modelling work group and had severally and individually been sufficiently persuaded of the GEHR approach to feel we could and should support it. A recent airline advert could be modified to express the stance we had agreed to take. I was to say of the New Zealand position "Oh, we're with GEHR". This would have been inappropriate behaviour, as it turned out.

I had hoped that Dick Whiddett could go to the Sydney meeting. He has made the major contribution to the New Zealand standards project in this area. In the end, I alone went, but I missed Dick's company and guidance, and very much hope he can be involved in future meetings.

top of page

The Venue, the Ambience, the Participants

The meeting was held in Australian Business Limited, 140 Arthur Street, North Sydney, in a modestly appointed smallish conference room, and throughout the meeting , it poured with rain. It was rain which flooded the streets, clamoured at the windows, appeared to mingle with the tangled thought paths in the room. There was a great wash of concepts and rather libertarian interactions that melded well with the tumult from outside. Coming from New Zealand in our present drought, it was a wondrous sight and sound.

There were two days of 'head down' meetings interspersed with sandwiches and coffee, and a dinner for the delegates with the Director General of NSW Health. The mood was informal, the humour ribald on occasions. It was mainly about trying to put a truly global form on the territory, but the globe was only represented by two Americans, four Australians (including Peter Tresseder, who is the overall Chair of ISO/TC215), three British people, three representatives from Japan, and myself. Tom Marley and Peter Kay from the UK are also key people in the CEN251 committee which has a parallel project to the ISO committee. The third Briton was Ian Sheppard of the British Pharmacological Society. The three Japanese delegates, Mihoko Okada, Ken Toyoda, and Arther Fujie all have university or high-ranking technical posts in Health Informatics. Peter Wagerman led the US delegation, and he is one of the major figures in the field in the US, and runs the TEPHR organisation ('Toward an Electronic Health Record'), which among other things organises trailblazing conferences yearly in both the US and Europe. He is a German American.

The other American representative was Ralph Korpman, health record entrepreneur and major tycoon in this area who informed us that his medical systems supplied in excess of 29 million clients, including the entire city of New York. He suggested that his own modelling model was so good that we might just as well adopt that. His ebullient style made it impossible to take offence at his hyperbole, and we quickly persuaded him that although that might indeed happen, it had not happened yet, and therefore there was material to discuss. I had been to one of his talks in London that had made me nervous because he appeared to be getting it right.

The Australians included Peter Williams, the convener who directs information management at NSW Health, Michael Toohey, a health information manager from Perth who is having some success persuading doctors to adopt IT, and Peter White of the Australian National Insitution for Health and Welfare who is responsible for the Australian Data Model which is having considerable profile world wide. It has recently been adopted by the US Health Care Funding Agency. Peter Tresseder, the overall chairman of the ISO committee, was also present. This was a very sophisticated groups of people from diverse backgrounds.

I was the only practicing clinician, although Ralph Korman said he worked as a haematologist for one week each month, and Peter White said he used to be a pathologist. I felt some responsibility for representing doctors, and thus spoke my mind where I could. These are processes which will affect all doctors everywhere. Life and Death are of Supreme Importance, and attending to them is our vocation. In all, there was a dedication by the participants to 'make it work'.

top of page

The Proceedings

Peter Williams' welcoming address set the tone for the meeting. He explained why the documentation assembled for the meeting was so limited. We each had a modest folder of documents including the agenda, draft scope documents, contributions to a 'standards inventory' (fronted by my own modest piece - see Addendum One below). He pointed out that the specifications for the stack of standards available would almost reach the ceiling, and that there had to be an alternative approach. It was a case of 'here we were'; the problem was to make a global system, we should 'just do it' . The 'tabula' was nearly 'rasa'.

Peter Tresseder told us more of the ISO process. There was his earlier emphasis on the emerging global community, and he communicated the sense of controlled urgency. He explained the mechanics of the ISO process, in particular the procession New Work Item, Working Draft, Enquiry Draft, Final Draft, ISO standard. We are now just at the beginning, and the timetable for the evolution of the product is set. As we get more deeply into this process, these timings and the spacing of the meetings will dominate our work programme.

Peter Wagerman gave an elegant Powerpoint demonstration which has already been distributed to delegates by Peter Williams. Among the points he made was the assertion that Electronic Patient Record Systems are at the core of all of health information. He compared electronic and paper records, and wrote that the longitudinal electronic record should be platform and institution independent. It should be able:

The elements of it were interoperability, shareability, security. He talked of a vision of a new type of society where information will change many aspects of life. He conceived of the global, national, regional, and hospital information structures as being a simple nested hierarchy. He emphasised the need for a unique national and international patient provider and payer identification system. He emphasised decision support and documentation rules. It appeared that his timetable for the evolution of this process was similar to that of the ISO committee, which perhaps emphasises the prominent role he sees for it. One off-the-cuff remark should be noted: during the 'introductory phase' of the meeting he was asked what position he had been encouraged to take up, and he said that he'd been told he could agree to anything as long as it was 'RIM'. The Reference Information Model is the current US position (in so far as there is one) on an overall information model for healthcare.

Peter White presented the Australian data model. This model has been carefully constructed of units that are defined according to ISO standard 11179. He described how its content was achieved first through wide-ranging consultation among players (or "parties" as the model defines them) in the Australian health sector. He further described how he took a smaller group and "locked them in a basement for two years" (it was not clear if this was literal or metaphorical. They then came up with the framework for recording health data that is found on their Web site, and which has also been distributed as a Powerpoint version by Peter Williams. He displayed the prototype version, which had the relationships between the entities depicted with lines and arrows. For the current version they simply removed the arrows. The model relies heavily on the concepts of 'party' (any person in the healthcare process), and 'event' (the ultimate indivisible atom of healthcare reality).

I think this could be a set of 'useful pots for putting things in' (after Eeyore in 'Winnie the Pooh'). It has apparently been adopted by the US Healthcare Funding Agency, which controls a large part of the US health dollar. The idea is that there is nothing that can happen in the health sector that is not represented in it, and that all the little pots and clumps of pots in the model have been arrived at through hard work and consultation. As Robert Mayes of the US agency said in private conversation at Orlando, he did not find it necessary to re-invent the wheel.

A set of global pots could be convenient. However, I could not see how the model could be clinically useful. Clinically useful data could be mapped into it, and out again perhaps, but for the actual practice of medicine it would seem to be an empty exercise. Perhaps there is something I have not understood here.

Peter White did say that the space between his model and the interface with the patient was called 'business rules'.. All the wisdom and craft of medical practice appeared to come under this rubric in his scheme. It looks as if this framework is to be universally adopted, but it is not clear to me how it relates to the practice of clinical medicine, and I recall in conversation with Peter Schloeffel last year that he also felt that this was something for statisticians and administrators rather than doctors. It is, however, widely acclaimed, and Australian, and therefore a reality among us.

Ken Toyoda spoke of the status quo in Japan. He said there only one National Insurance System in Japan. For the medium to large hospitals there were only five vendors of IT, and usually only one vendor per hospital. He repeated his helpful statement that the healthcare information system is the key to improving the quality of healthcare delivery. He emphasised the importance of modelling of the clinical process, and of interoperability, and standardisation of terminology. Later in discussion the Japanese group, who struggled with the profuse vernacular of some of the contributors, made a particular point of opposing an American suggestion (Wagerman) that documentation of, for example, surgical procedures should be standardised. In the end this phrase was kept in but it was made clear that it was only in the context of the initial resolutions of the ISO committee TC215. It was the Japanese delegation who in the initial terms of reference for the entire ISO/TC215 committee had emphasised that the 'standard' should specifically not dictate how medicine should be practiced. It is an important disagreement which will be inherited, in some sense, at the next level of discussion because it was not resolved.

top of page

The Scope Statement

About half of the time of the meeting was taken up with evolving the Scope Statement, and the 'Terms of Reference'. There was the Orlando document as a starting point, which was set up as a 'straw man', and then modified. Each word was savoured, considered, rejected or accepted. To me, it seemed like a negotiation for how many angels might be felt appropriate to dance on the head of a needle. However, it was made clear to me that his would become a 'legal document', set in stone, deciding what could or could not be part of the 'scope' of ISO electronic medical records standards. I will just pick up on a few words.

The scope of WG1 is to develop standards for the trusted management of information concerning health and the healthcare process. In the context of these standards healthcare is very broadly defined and includes maintenance of wellness and support for all modalities of diagnosis and therapy.

Comment: 'all modalities of diagnosis and therapy' This wording was chosen in response to the Japanese delegates raising the issue of non-Western medicine, particularly Chinese medicine. Words like 'Traditional' and 'Western' itself were felt to be culturally chauvinist. The term all 'modalities' was thought to be inelegant, but kept in for lack of a better word. I tried to point out that some alternative therapies may depend on secrecy, but the point was drummed out by the rain ...

WG1 will address health record standards that are independent of setting and technology.

Comment: '... independent of setting and technology. 'Setting' was chosen very deliberately to be an inclusive term about physical context.

The standards will enable the availability of the appropriate information at the place and time of decision

Comment: Here there was heated discussion on the concept of 'Real Time'. Ralph tried to insist that this phrase be inserted here, and when asked why, responded that if the record were not available in real time, "the patient has died". I concurred as emphatically as I could with this assertion (I emphasised the importance of the concept of 'Real Time' in my submission to the meeting (see Addendum 1). We met opposition from both Europeans and Japanese delegates. 'Real Time', they said, was a technical term. Ralph muttered something under his breath about it being a 'conversion experience', but in the end all we got in was 'at the place and time of decision'.

They will also facilitate health consumer participation and support clinical, public health, research and administrative functions. WG1 recognises that health records are integral to managing both patient care and wellness.

WG1 will coordinate health information modelling with respect to standards that will be related to a general conceptual model of health information.

Comment: WG1 will coordinate health information modelling with respect to standards that will be related to a 'general conceptual model' of health information. This for me is the 'big wow', the possibility of there being a 'general conceptual model', and the possibility of having input into it.

Such standards work will ensure the compatibility of specific models which arise from the activities of the other working groups. The content of those specific models will be the responsibility of the working group in whose domain the activity occurs.

Comment: This claims the 'modelling coordination role' while being careful to 'limit the territory' so that the other working groups will not feel encroached on. However, the overview of the subject still is fluid enough to admit of different interpretations - eg, Does the 'work item' of 'Ownership' belong in 'Modelling' or 'Security' (see under work items below).

Terms of Reference

WG1 will:

Comment: framework of standards ... shared across all boundaries including systems, jurisdictions, disciplines, and professions. This is a very comprehensive wording designed to leave out no part or corner of the world.

The standards developed will:

Comment: modelling notation. This refers to a formal established set of building blocks for constructing models such as the UML and Rational Rose, which we definitely need to take a look at.
Comment: 'Unique identifier'. It was the serious proposition here that each citizen of the world should have a world number.
Comment: Facilitate the faithful preservation of meaning and clinical context. Here I found myself on the side of the Japanese. It seems to me that if all clinical context is preserved in the standard, one might as well send 'screen shots'. The idea of decomposition into elements or items means freeing them up from context to some extent. I stated that when doctors disagree, they 're-contextualise' items or elements into new configurations. In the end, the word 'facilitate' was put in here, whereas otherwise the passage might just have read '... preserve meaning and clinical context ...' which I and some others felt might beg the whole question.
Comment: ... not to be limited to application within computerised systems ... The point was raised (by me and others) that perhaps there should be some sort of 'third world ticket' in here, some recognition that it is important to be third world friendly, not to exclude anyone by virtue of limited resources. The result was to include paper records within the terms of reference.
top of page

Proposed Work Items

For the discussion of this, Peter Wagerman simply went to the front of the room, plugged in his laptop (into the overhead projector) , and took over. It appeared that there were proposed work items 'on the table', some from the Europeans, some from the US, and I am not quite sure how they got there, but the discussions proceeded at a clipping pace. Items were omitted and deleted by a process of consensus or alignment that appeared to depend on discussions to which I had not been party, or perhaps involving knowledge areas to which I do not have access. Everyone seemed to agree, after the Australian model, that 'Healthcare Party' and 'Event' were realities. Other definitions, such as encounter, episode, structure were mentioned. The purpose of the work item list is to have items raised which are to be resolved, and many items that were raised seemed altogether too hard. Against each item is written the name of the delegate who was to 'look after' that work item for presentation at the next level up of the process, the formal adoption of 'work items'.

One such 'too hard' item was Ownership/Access, but this was something on which I had thought and discussed a little, and I determined to drown out the rain and the opposition and Say Something to it. I said my piece about how records might perhaps be defined as 'not ownable', like the weather, or land (for nomadic peoples). I should perhaps have said 'like people' , except that people of course CAN be owned (just as the New Zealand health sector is trying at present to 'own' doctors'). The serious point got through, that instead of ownership, one could have rights and obligations around records. I am sure that possibility will have been identified before, in the many discussions that will have occurred in this area throughout the world. There was discussion too about whether the topic was validly considered part of 'security' and thus might belong to Working Group 4. It is obviously closely linked in. I would argue that it is part of 'architecture', ie part of the definition of 'what it is' that is to be made secure.

In the dividing up of the Work Items for formal presentation, we were allocated 'ownership /access', which means we have to be the sorting house for assorted views on this matter and then find out how to fill in the appropriate ISO Template, all by 28 February. We will have to have a meeting about that and find out how to do it, and agree on what to say.

top of page

Some Interesting Conversations

As well as the formal proceedings, I would like to report on a number of conversations with the participants that give clues about 'where we are at'.

Ralph Korman and The Model of Models

In the first coffee break I talked to Ralph over coffee about modelling, and how I felt we should be looking for a model of models, not a model. This was the idea contained in my brief note to the work group meeting (Addendum 2). It is the same simple idea I have encouraged all along, based on the linguistic notion of syntax which can generate an infinite diversity of utterances, each of which addresses a situation and is/can be considered a model. The model of models for Natural Language is the combination of rules of syntax and the lexicon.

Admittedly, we do not yet have a "Natural Grammar", despite the best efforts of linguists. When setting up a new language system (and the ISO standard might become one), we can set up just the sort of self-consistent language game that linguistic philosophers have long desired (eg Ludwig Wittgenstein, Tractatus Logico Philosophicus). Ralph said that he agreed that we were looking for a model of models but he would rather consider that we were looking for a meta-model, a containing model. I felt we were understanding each other, but the implications of the two points of view are in fact quite different. A 'containing model' might be like the Australian data model, the definitive set of 'Eeyore pots'. It cannot generate anything other than itself, but a true model of models creates contexts and models to suit situations, and is itself 'empty'.

We talked about consensus, and he said he preferred to call it 'alignment'. I said that this must be a cultural difference, which caused some amusement. However, the word 'alignment' rather than 'consensus' has entered the ISO debate, and was used by Peter Tresseder. It is no more the same thing as consensus than 'love' is the same thing as 'sympathy'. It appears that the two words may be used interchangeably in the work ahead.

We did briefly discuss Lawrence Weed's 'SOAP' unit concepts, and he said that the problem with it was that it proved impractical to track 'problems' individually in clinical practice. I agreed, but said that the 'mind set' of 'subjective, objective, assessment plan' as an approach to problems was certainly widely shared among doctors, and he agreed with that. I said that ultimately the SOAP scheme could be reduced to a concept of 'input/output' clusters that were a quantal 'chunking' of the clinical stream. This was irreducible and universal. He agreed with the notion.

Later during the debate on the 'scope' statement, Peter Kay said with resounding intonation: "Well if you are trying to take over the world it could take a little time." Ralph instantly replied "Well, you lot should know, you tried to take over the world", to which I intercede "No, only one third of it", and there followed some remarks on the circumstances of the US achievement of independence. The meeting was full of such banter.

Later, at the dinner, we were talking about the impossibility of ever providing a Health Resource adequate to demand, and I mentioned Prioritisation as a way of rationing the utilisation of a resource. I said it worked for Cataract, but both the NSW Director General of Health and Ralph Korman said that it would never be accepted in their countries. Ralph asked whether I wouldn't give a historico-cultural analysis of settlement patterns to explain the cultural difference. I said I thought that New Zealanders had always done what they were told because the people who settled New Zealand originally were largely like that, and nothing had happened to make them change their ways. I cited a common experience in eye clinics, of patients who say "I know I am not meant to ask" or "I know its none of my business Doctor, but am I going blind." This cultural pattern is also shared by the UK in some cultural niches.

Finally, among memorable 'bijoux' from this source, the conversation had meandered to 'what to do about it all', and Ralph was saying that the same conclusion was being reached from all sides, and I asked what that might be. He replied that there was universal agreement emerging that there must be some sort of "best practice" guidance of clinical pathways, and there must be monitoring of outcomes. He also suggested that the feedback from the data of ordinary practitioners would transform medicine, and he mentioned the concept of 'bulk' single-purpose surgery units which he said his data showed in some cases were cheaper, quicker, and had lower complication rates than some high-profile, high-status centres. I have to concur with all three of these propositions: some form of "best practice" guidance, outcome measurement, and use of the clinical experience of ordinary physicians to provide feedback on healthcare delivery practices are becoming agreed assumptions for the New Medicine of the information age.

Emergency Medicine

The concept of a 'quick pay-off' (so that everyone could see the process was worthwhile) was suggested by Peter Kay, and Emergency Medicine was to be the vehicle. The idea was to agree on a single international format and content for emergency medicine information, and the US standard document was produced and discussed. The Europeans seemed quite in favour of it. However, the idea appeared to be quashed by Ralph Korman. Vendors and doctors would never use it, he said, and therefore there was no point in it.

Peter Wagerman and the Demise of Corba

Near the beginning of the meeting, the question of CORBA was raised with Peter Wagerman. The Common Object Relation Broker Architecture is an American Pipe Dream, a bit like GEHR is for the Europeans. It apparently works quite well in many aspects of computing, and there is a subset called 'corbamed' that tries to achieve a dictionary of shared objects with shared structure, which can be intercommunicated between participating systems. It has never quite taken off. It was interesting to hear Peter say that he felt that the concept was superseded by the capabilities of XML. XML appears to be becoming the wonder 'solve-all' medium, with relevance to messaging, object definition (apparently), markup, multimedia access, etc.

GEHR and the New Zealand Model

Again, toward the beginning of the meeting, someone else raised the concept of GEHR. "Oh, that has been oversold", said Peter Kay, instantly corroborated by Tom Marley. I had approached the conference thinking I could say "We're with GEHR", and the conversation of which this was part had been about 'context' (see below). I said that I understood that the GEHR concept of transaction took its own context with it. Marley replied that the problem he had felt with GEHR was that the unit 'transaction' was also the minimum attestable unit. 'Attestable' here I take to mean the minimum unit that will stand up by itself, that has been 'vouched for', so to speak (by whom, and how, are separate questions). Marley said there was a Danish group wanting to make the 'item itself' the minimum attestable unit. He said 'transaction' was a useful concept, but not the only one nor the only way that notions of attestability and context were being handled. On looking into the CEN251 latest modelling contribution, it appears that with it the notion of 'context' has its apotheosis. This looks to me like a very interesting document that we should study. I was clearly given the impression that the Europeans had moved on past GEHR, leaving it stranded, so to speak. So when the question, from Peter Williams, was "Would you like to tell us about standards modelling development in New Zealand", I muttered on about the model we had produced, and said that after developing it we had become more familiar with the European initiative, particularly as manifested in GEHR, and that HL7 had achieved wide diffusion in New Zealand, and indicated that I was here to learn more, because we did not want to take off into an idiosyncratic direction.

Later, after Wagerman had spoken and there was a brief lull, I felt I should say more about our humble efforts, since I appeared to have missed the 'main chance'. I drew the attention of the group to our conceptual data model diagram of the 'Medical Process' which had been reproduced in the hand out after my little piece on the 'Model of Models'. I spoke to it briefly, and Ralph Korman was on to it at once. It wouldn't work for inpatients he said, it is an outpatient model. I did not have the chance to address it further; the group had moved on. However, much of the proceedings was like that. Matters were picked up and dropped almost straight away. The meeting was about 'scope', was about producing an 'enabling document'. I do not consider our model or its philosophy dismissed, rather the contrary, I think we are 'in there with a chance' for the next stage in the argument.

Later, in the break immediately before the 'work item' part of the agenda, Peter Tresseder asked me a similar question to the one Peter Williams had asked as convener: were we going to implement the model on our Web site? I again replied that we were waiting to see. He asked if Dick Whiddett was still involved, so I told him that Dick was still Chair of Working Group 3 and I had wanted him to represent us at the Sydney meeting, but that it was not to be basically for financial reasons. I told him that funding applications were now in for the support of standards work, and that there were rumours that the HFA might pick up the tab. He asked about our involvement with GEHR, and I spoke of Peter Schoeffel's visit and the enthusiasm it had generated, but that when it came to the point the GEHR kernel did not appear to exist. I mentioned that Ashwin Patel had said we would design our own, but had not yet done it. I said that as far as I knew the UK 'Support Action' one did not exist either. He said that that was what worried him. He said that this was what the work items were about. I said that that appeared to be 'our moment' for this. Then he moved that we commence the work item discussion.

top of page

Context

This concept has been given a much greater emphasis in the new CEN model. To quote from them:

'With respect to the electronic communication of patient information, the standards being developed are not so much concerned with enriching the context as primarily with safeguarding or preserving the context so that data is not lost, added, and/or wrongfully interpreted, thereby causing unintentional harm to a patient. Furthermore, in the EHCR Architecture, context is represented not just by surrounding 'text' but by architectural components that singularly or collectively surround, contain, link, constrain, and/or qualify the content so as to make the communication safe and unambiguous. Context then should not be considered as merely residing in just one specific place but rather as an architectural principle that diffuses all of the architecture.' (From the Introduction to 'Health Informatics' Version 1.0 1998-12-01, Part 1 'Extended Architecture'.)

This passage introduces the concept of 'Text' which is so crucial to my understanding of these matters. The new European contribution might certainly repay some serious study.

top of page

Where to in New Zealand Now?

The problem with implementation for New Zealand as it is for the rest of the world is that it is not yet self-evident what is to be implemented.. There is the GEHR process which has been so ably presented in New Zealand , but which in the final analysis has few if any genuine applications and whose so-called Kernel lacks substantiality. The other problem with it is that the latest version of CEN251 from which it derived has explicitly moved beyond it and is in conflict with it in certain respects. I daresay later version sof GEHR will occur, as will versions from the UK Action Support organisation, and there will be compatibility between them. Perhaps all contributors will now work together on the emergent 'meta' model, and a contribution from David Lloyd of CHIME in London implied this. But the ISO process is in evolution, and will include compatibility to CEN251.

In all probability they will not end up different documents. Why, now, therefore, go for a fragment like GEHR which is in a sense 'cast up along the way'. I had the opportunity to talk to David Lloyd in London, and he explained about the different 'flavours' of the European effort, and drew my attention to the latest CEN production which to my surprise he presented as one of the three alternatives. If anyone were to develop a Kernel, he said, it would work for the other two since so much was shared between the three flavours. They are after from the same 'stable'. The ISO process is 'timed' to last 36 months. It has the Americans 'on board'. Why not, then develop the 'Kernel' for that.

The constant use of the term 'Pre Standard' across so many generations of 'deliverables' is bewildering .There is an implication that due process faithfully followed will deliver an outcome. Perhaps it is my background as a surgeon which makes me impatient with this mode of thought, but for surgeons at least nothing ever happens until and unless you "happen it". The endless presentation of pre-standards appears like a performance anxiety, or perhaps a recognition that that which is being proposed could never become real. As I understand it, this was the American position at least until recently. The HL7 protocols actually work, and there was an argument that standards development should be bottom up, or in other words, proceeding from things that work reliably to 'more things that work' rather than trying to think the whole thing up in one 'bolus' like the Europeans and the GEHR enthusiasts appear to be doing (although Peter Kay explicitly denied a 'big bang' approach).

However, the US too clearly have recognised that there is an 'emergent level' called 'architecture' within Electronic Medical Records that needs to be shared if there is to be the degree of interoperablitiy and potential global sharing of healthcare information that the ISO process has been set up to facilitate. It is hard to escape some version at least of the 'component approach'.

We could use the ISO standards development process in New Zealand as the implementation process. We could implements the ISO standard as it develops. Some good guesswork might even position us to lead the process by implementation trialling. What we do is both open for us to decide, and matters enormously on all levels. I suggest that New Zealand fully support the ISO process and use it to help work out the National Plan.

We should send at least two delegates to the ISO/TC215 WG1 meetings, and seek admission to the messaging and content and security work groups as well (WGs 2, 3, and 4). It is also now time to contact the industry for their input, because the bare structure of the solution is now in fact beginning to 'stand out' from the background. They should know about it and have their chance to contribute.

There appears to be developing international consensus (or 'alignment' as the Americans would prefer to term it) about the need for some managed care, and for some attempts at 'best practice' guidelines and their monitoring. But those are among the least of the reasons why we should work hard to implement these standards.

Doctor-friendly electronic medical records have finally become available, and there are now designs which doctors are queuing up to buy. While I was in Sydney, I was shown an interactive teaching tool for Ophthalmology which was as accessible as a modern computer game (contact John Gribb for details), and the simple link of this concept to doctor and patient centred record systems such as we have already designed in New Zealand will open so many doors and corridors for doctors and their patients as to make the adoption of such systems the norm. 'Now' is indeed the hour to design the communication environment for this burgeoning reality.

top of page

Addendum 1

ISO/TC 215 N25.4 - Agenda items 7 and 8
ISO/TC215/WG1/N4
INVENTORY OF STANDARDS

Experts are kindly invited to bring examples from their jurisdiction or region, of existing standards and/or ongoing standards projects that are deemed to be of potential interest as an input to the ISO standardisation process. Please send in advance of the meeting, by e-mail if possible, copies of any documents you want us to consider, or bring a paper copy. A summary describing the scope of the standard would be very useful, particularly if document is not available in English or if it can not be freely distributed for copyright reasons. It may also be appropriate to reference a web site address.

It would be appreciated, if the information is available electronically, for it to be advised to me by 7 January 1999, to facilitate reproduction and distribution.

This will be an important component of our initial meeting. The scope of work is potentially very large. To be most effective, the Working Group will need to achieve a balance in its work programme between building on existing regional developments, where the initial focus may be, and undertaking activities to remedy 'gaps' through entirely new work.

A large number of relevant standards were advised at the August 1998 ISO TC215 meeting (particularly from Europe, the United States and Japan). There are clearly also other relevant activities which need be taken into account.

top of page

Addendum 2

ISO TC215 Work Group 1:
Position Statement submission for Sydney Meeting 21-22 January 1998

Michael Mair FRACO, Head of Delegation, New Zealand

A Generic Architecture must be very simple. In so far as 'making sense' of something is the same as 'modelling' it, a generic 'model' is a 'model of models', or 'universal model'. All the different 'ways' of practising medicine and the different sub-branches of medical practice, can be considered 'models in action'. A Generic Architecture might better be considered a 'kit' for making models out of.

If there were any 'Natural' or 'Universal' characteristics of human record-making activities that could be built in to a global standard, they would certainly come in useful at this point. I suggest there are at least two.

The meta-model, or model of models, should not be into the business of making stories or sense of all this harvested data. That is the task of those at the real time work face, and of people who subsequently rework the data by the application of other search procedures and templates to it.

There may be just enough in the simple architectural principle of the succession of clusters of harvested data items on the input side to give organisational coherence to an 'open but secure' labyrinth for the healthcare network. This might be regarded as the 'core process' of the electronic health care record. It could have the 'coordinating' role that the modelling WG 1 product is meant to have for the other aspects of the standard. I have called this the 'text generator' or 'story maker' model elsewhere ['Comprehensive Ophthalmology Management' in Proceedings "Towards the New Millennium" Conference of the New Zealand Health Informatics Foundation, Christchurch, NZ August 1998].

In the Standards New Zealand Working Group 3 which considered these matters, we felt that a hierarchy of levels including the 'encounter' were also validly considered 'Natural Units' and thus a good basis for a Universal Model. These levels are:

Some of the details would now be changed, but we continue to argue for the overall scheme, which has much in common with the European position (in all its flavours), and with some of the features of the HL7 architectural initiative.

We certainly hope that progress on gaining consensus internationally on these important matters is achieved first at the Sydney meeting, and then in Berlin.


top of page