Comments on: The crisis in e-health standards

By: Ruminations on ‘design’ in e-health « Woland's cat

Ruminations on ‘design’ in e-health « Woland's cat — Sun, 19 Dec 2010 20:55:37 +0000

[...] have often bemoaned the state of standards for the e-health sector. Earlier posts provide details. The main argument is that the key specifications the sector needs are for [...]

By: Should an Organization like HL7 be Engaged in Building Standards? | HL7 Blog

Mon, 29 Nov 2010 17:19:29 +0000

[...] A nice summary of problems with current approaches to standards development in the eHealth domain is presented here. [...]

By: Should an Organization like HL7 be Engaged in Building Standards?

Should an Organization like HL7 be Engaged in Building Standards? — Fri, 13 Aug 2010 17:28:03 +0000

[...] A nice summary of problems with current approaches to standards development in the eHealth domain is presented here. [...]

By: iNTERFACEWARE

iNTERFACEWARE — Sat, 19 Dec 2009 21:29:57 +0000

How does HL7 work?…

Whenever I travel for work, one of the most common questions I hear is “How does HL7 work?” HL7 is not always one of the sexiest subjects, but as interoperability and connectivity continue to be huge drivers in the health care space, the qu…

By: Gerard Freriks

Gerard Freriks — Wed, 04 Nov 2009 08:39:43 +0000

Almost 14 years of semantic interoperability activity in HL7, CEN and ISO as formal standardisation bodies has learned me several things.
-1- When we want to make semantic interoperability a commodity we need to realise that this is daunting. Many will have to play their natural role.
-2- Like language we need firm stable agreements with respect to the character set, the language and the language syntax. For the words we need regularly updated versions of the dictionaries. Finally we need local agreements of what to store, retrieve and exchange that can be changed any moment in time.
In other words we need various types of standards and organisations. Some producing stable components and other being very agile. This determines the type of organisation, it working mechanism, its publishing mechanism and business model.
In general, commodity means that we need an infrastructure where stable and agile components plus governance are standardised in some way.
-3- To often actors in this field try to play all standardisation roles and types of organisation at the same time. It is obvious that this will lead to problems.
Or in other words extend their role to an unnatural one.
-4- Without (national) legal regulations it is impossible to create a functioning stable infrastructure. Stable standards produced by CEN/ISO (like EN13606 EHRcom, EN13940 HISA) can play a role in legislation creating the free movement of goods, people, money and services in Europe.
Legislation should not be extended to the agile, non-stable parts, other (new) organisations will be needed to handle the agile parts (archetypes/templates, coding systems and ontologies)
-5- For things to work we need to have all agreements, standards, managed in a stable way and be real publicly owned entities.

By: wolandscat

wolandscat — Wed, 30 Sep 2009 23:54:03 +0000

With respect to the solution, I did not mean to imply that open source was the solution (I don’t think it is); I think that an open organisation with some of the attributes of the open source organisations is needed. But unlike open source orgs, it needs to have a focussed engineering team(s), not a ‘bazaar’. As it happens I don’t think that open source is nearly as important as open data standards and open interface (service) specifications. But reference software for standards needs to be available in source form for obvious reasons.

By: Jens Eliasson

Jens Eliasson — Tue, 29 Sep 2009 22:46:58 +0000

This is the best description I have read so far of the current state of affairs and why close to nothing has been delivered an unimaginable pile of money spent later. Also time slips much to fast. Defining the problems and criticism on what has been not done right is very exact and well put IHMO. We, a creative “biblical David”, in this environment have argued quite a lot along the same lines, but never managed to capture a more complete picture as you have done here. Thank you.

But, when it comes to finding the remedies I can not agree to all. In the article I think some about the role of open source and .org is an after construction of history. Actual inventions or creation of new designs have yet to be developed in such an environment. Open source is so far a means of marketing to other developers and to the academic world. Such orgs. may be very good at governing and extending standards, but that comes well after the engineering development process has completed its first couple of loops. It is not a prerequisite.

By: wolandscat

wolandscat — Wed, 23 Sep 2009 08:21:11 +0000

Gunnar, I take your points on board. I should clarify what I meant by calling the solution a ‘.org’ – I actually mean a new kind of organisation that houses a self-consistent set of specifications, uses software as a means of validating the specifications, and uses some of the other well-known ‘.org’ features, such as community mailing lists, wiki, version control, issue tracking and so on, to provide maintenance. The official standards organisations (as you point out they are literally also ‘.orgs’) fail to do this, and the open source software developers certainly do. The former suffer from no/limited internal coherence in their specifications, no maintenance path, and weak accountability to stakeholders, while the latter don’t even tackle the problem of interoperability.

If we look around today, there are at least 3 other other organisations (all .orgs that have broken the current mould in certain ways: 1) IHE – the existence of IHE indicates the problem of consistency within and across SDOs; 2) IHTSDO – the structure of this SDO has an international board of government stakeholders, solving the accountability in at least some measure, and 3) openEHR.org, which has tried to build a set of specifications which are internally consistent, and also validated in software (it is weak on accountability though). The last of these of course is the one I have been most involved in (critics of the main post here will accuse me of writing a critique that suggests that openEHR is the solution, but that would be misunderstand the causal order of things: openEHR.org is the result of years spent discovering the problems I have documented in the main post.)

In the end, the following observations seem to be incontrovertibly true:
a) formal SDOs don’t have a commitment to a self-consistent technical framework, never mind being consistent with the outputs of other SDOs
b) the committee process of formal SDOs does not lead to the creation of new technical artefacts of quality
c) formal SDOs have no maintenance plan other than a 4-year wait

To those such as Gunnar who have struggled with the impossibilities of the process and the recalcitrance of many human beings on the way, I salute you. But I don’t think it is the way of the future in e-health.

By: Gunnar Klein

Gunnar Klein — Wed, 23 Sep 2009 05:02:36 +0000

Response to the “Crisis in e-health standards” by Tom Beale, 2009-09-17

Despite being one of those silly persons that due to a personality disorder has continued for 17 years to struggle within the de jure standards world for eHealth that is so useless, I cannot escape agreeing with Tom on much of the description of the issues. Yes, interoperability in this field is hard and, as a total solution, far away. The specifications world is indeed messy. And I can feel embarrassed by some of the remarks on the weaknesses of the present results.

However, I think the “crisis” is misguiding in a few ways.

a) Technical development of complicated systems to support production of large parts of society (health is some 5-7 % of GDP) is not in general successfully carried out by well working open source groups (“.orgs”). The growth of both knowledge on “how to” and real implementations is far from a straight line. The evolution of a “solutions” as specifications and projects is as a rule messy and most of developments are blind alleys and wasted efforts. Is file sharing using MPEG4 building on Betamax? The same is of course true for development of clinical practice. I am pleased to learn that Tom is not planning to engage in paediatric oncology guideline development even if that could benefit from some input from good engineering, (eg. see paper by Rong Chen et al, Stud Health Technol Inform. 2009;150:653-7.) Medical Science and clinical practice is also not developed by but heavily influenced by both “de jure committees” and open groups but not according to the grand plan for progress. That was a failed strategy tried in certain countries but generally not used any more, certainly not in China the largest communist country. The examples so far of national eHealth programmes have not been very impressive as to cost-effectiveness. Agreed?

b) Large scale deployment of complicated technology is based on a combination of meeting a need with a number of different integrated engineering results many of them based on years of specifications in the public domains, from de jure standards organisations and other sources plus of course skilful marketing and other commercial efforts.

c) Health care and eHealth is very complex and we do not need THE solution from standards bodies nor “.orgs” (note that iso is actually .org as a domain). We need many and the standards organisations active in this field have helped people to create many useful solutions that are in use. But of course we could ask for better quality in specifications work including maintenance and let us discuss how we can achieve that. Concretely, not in a general abstract way. I doubt that creating yet another organisation is the solution.

Gunnar O Klein now a practising physician
Chair of CEN/TC 251 1997-2006, founder of ISO/TC 215 and still chairman of the ehealth standardization co-ordinaton group which has with little success tried to join forces between WHO, ITU, ISO, CEN, IEEE, HL7, GS1, DICOM, CDISC and IHTSDO.

By: Erik Sundvall

Erik Sundvall — Tue, 22 Sep 2009 08:43:35 +0000

Interesting summary! You wouldn't happen to recommend or be involved in any suitable .orgs or .orgs would you?