Contents:
- Introduction
- The Work Items
- Ownership and Access Work Item
- South Island WG1 meeting March 1999
- The Japanese Connection
- Summary
Introduction
The meeting started with the normal sense of surprise on seeing those familiar faces again in yet another environment. There were 25 people in the group. The weather was muggy and a little turgid; the resistant air opposed our meeting, with neither the cold clarity of Berlin nor the tumult of the waters in Sydney, nor the glossy sleekness of Orlando. But it was a serviceable environment, and it was good to be back in London.
The Work Items
There were only three work items from WG1 which had 'made it' to being actually proposed and voted by the member countries, and these were the General Domain Model, the Unique Country Identifier Standard, and Ownership and Access.
The discussion of the other work items was lengthy and somewhat inconclusive. The country identifier standards one is going ahead and I remembered that Yogesh Anand had volunteered to work on this with Trevor Hodges from Canada, who is the convenor for this work item. Many countries had pointed out that country identifier standards already exist. It was just a question of agreeing on how to do it. Trevor will have been in contact with Yogesh directly about this.
The General Domain Model discussion was at times wordy and obscure. It centres on the Australian National Data Model developed by Peter White and associates. As Peter put it:
"The approach proposed to resolve this was that, in the first instance, jurisdictions would be requested to map their existing models to the current Australian National Health Information Model Version 2.0 at consecutive levels of abstraction. As the mappings drill down, at the point where they are unable to be further mapped (which may be at any level of abstraction including the highest) they should offer an alternative approach, which accommodates their needs together with a brief explanation of the rationale. The use of the Australian model for this purpose was not intended to confer any particular status over any other model, but rather as a 'straw man' to commence the process."I have commented on this model before (Sydney meeting report). I do not see the value of an aggregation of 'Eyeore' pots or categories for their own sake, to copy data in and out of, even if they are internationally agreed. It is a statistician's model, but does not appear designed to support the practical business of making or exchanging medical records. There was extensive discussion of 'modelling', which led to some sharp exchanges. Peter Waegemann the US delegation head said he was sometime ashamed when asked what had gone on at ISO to have to report on such abstruse and empty exchanges (my words). Stephen Kay from the UK who has a role in 'modelling' with CEN defended the discussions on modelling in the abstract. It will be interesting to see what becomes of the 'General Domain Model' in Tokyo.
Of considerable interest was the new work item from Australia proposed by Peter Schoeffel on 'Functional Requirements for the Electronic Health Record'. Peter is an Australian GP and computer expert with close links to the GEHR which is being developed by Sam Heard and Tom Beale at Flinders University. There is also the UK CHIME institute headed by David Lloyd working on a similar model, although there are differences, and there is a meeting in London in November of all GEHR participants to get a common approach.
Peter presented his work item as yet another attempt to 'get it right'. I asked him how this stacked up with his commitment to GEHR, and he said 'no problem', and to my surprise, the US commercial representative Gary Dickinson jumped right on and said he was keen to work with Peter to develop these 'functional requirements'.
GEHR are going ahead aggressively and have just won (29 October) funding in Australia from the Australian GPGC to run a development trial lasting nine months involving a number of software companies to experiment with a common health record server behind different applications.
GEHR is offering a complete 'solution', if they ever actually unveil it that is. There is an exclusivity about it that is breathtaking. The proposition is that instead of continuing to debate the issues, let's just implement with the GEHR model and kernel. It is to be programmed in Eiffel to begin with, but maybe becoming available in a language like C++ when more fully evolved. There is a leap of faith involved here, unless one is already 'on board' and part of 'deep thought' communications, or one of the more frankly commercial companies around GEHR. It will be ironic if those companies end up marketing products that are not GEHR at all. This is a time of explosive development, and it will be interesting to see if the GEHR development cycle gets lost in the heady competition emerging between a very few 'best of breed' applications.
Ownership and Access Work Item
The Work Items had been subject to vote, and there was unanimous support for the inclusion of a work item on ownership and access from WG1 (14 countries) with 10 saying they would be prepared to help with it. There were two adverse comments on the item we had submitted, and which is available on the NZHIS site. The Germans said the format of it was wrong. The Japanese said they agreed with it but that it should be a technical specification not a report. The UK and US said it should be a report not a specification. The UK comment said that the present 'scope' contained only an outline of the proposal 'against which conformance criteria cannot be applied.'
My New Zealand Work Item contribution on the immunological paradigm (circulated in an earlier draft form to HISC members before the meeting) was received without much comment. Bernie Cohen, a senior UK computer technologist, said that the concept was similar to that of 'agents', which were Artificial Intelligence-like devices which can have a sort of autonomous existence, and we were five years too early since the concept was not developed yet. (see http://www.fipa.org/). He said that it would be unwise to use the immune system as a model since it did not work well, but Per Arne Lundgren from Sweden said that that was no objection. He suggested the immune paradigm be considered in the technical report we were to write along with other approaches.
I did say we (New Zealand) would be prepared to continue to convene global views on ownership and access. Our role was endorsed, and we were encouraged to present succinct two or three page 'Scope and Deliverables' statement for discussion at the Tokyo meeting.
The work item itself that David and I developed is brief and is linked to this report. I have had favourable feedback from Sam Heard of GEHR who has called it a 'safe place to start'. The crux of the work item is, I think, whether the 'technical report' is to aim for a 'specification', or to simply be 'guidelines' or a 'framework'. If it is just a framework, and if there are no conformance criteria, then it will be of interest only to framework collectors. To at least aim for a global implementation is the hallmark of sincerity, since if we are not after genuine global interoperability, one must ask why we are participating at all.
This will be a hard fought battle, with some commercial big players very anxious to avoid a prescriptive standard since they are, in fact, trying to capture a major share in the global market, and might prefer to define these things for themselves. So the will to develop a standard specific enough to be useful and not just a framework becomes the acid test of the integrity of the ISO/TC215 process itself. The standard might contain a framework within which diverse views on ownership and access might be expressed, and yet still interoperate. But if it does not deliver conformance sufficient to facilitate interoperability, then the whole argument will have proved to be cosmetic.
South Island WG1 meeting March 1999
I was asked casually at the social evening whether the next WG1 meeting (rather than the full TC meeting, which is in Tokyo) could be in Timaru, New Zealand. This meeting is to be in March 2000.
Later it appeared the group was serious about having the March 2000 meeting in New Zealand. It really seems there is a role for us in this argument, if only because of our perception as being unaligned and independent. Also the 'cross cultural' sensibility in our presentation has found favour, particularly with the Japanese delegation.
I am glad that Nick Manson (chief Information Officer at the MOH) has agreed to support the March 2000 meeting (9-10 March), but the venue is yet to be decided.
On reflection, Timaru would be a good site for it since the Timaru Health Sector is so much under siege at present, a microcosm for so many others nationally and internationally. However, with David Menkes' involvement, Dunedin is a possibility, and there is always Christchurch. Peter Waegemann has already confirmed his attendance. It will be a great opportunity for New Zealand workers in this area to work with many of these key international figures. I have had confirmation from Peter Waegemann, US delegation head, that he will attend, and so will Yasuyuki Horoshi from Japan.
The Japanese Connection
We are hopeful that close collaboration with the Japanese delegation might be achieved. They appear to have similar views on the virtues of a technical specification rather than guidelines (from Ken Toyoda), and we are planning research with Horoshi Takeda of Osaka. Dick Whiddett has had exchanges with him, and David Menkes also sees the enormous opportunity for cross-cultural studies of ownership/access attitudes that this affords. Yasuyuki Hirose from Okinawa has written a really exciting paper on the way these matters have been dealt with in Tokyo Dental Hospital, and explores the theoretical issues as well. We hope over the next two weeks and in Tokyo to confirm these contacts, and begin to work practically with them on our own implementation. I also had positive comment on the New Zealand data model from delegates, particularly Angelo Rossi Mori from Italy, who asked why we had not implemented it yet, and said that CEN were now looking at something similar.
Summary
We have a role with ownership and access, but can expect debate on the relationship between a report and a specification in Tokyo, with Peter Tresseder perhaps deciding on an interpretation from the ISO secretariat. We have a WG1 meeting planned for 'somewhere in the South Island' for 9-10 March 2000. We have considerable interest in our journey toward a National Implementation of EMR from many quarters. The GEHR development is expected to 'declare' soon.
The pursuit of the national and global electronic medical record, like the hunting of the legendary snark is now in full flood.
Mike Mair and David Menke