Cynical soul that I am, I must say that the arrival of all this largesse on the second last business day before Christmas did make me wonder if the holiday period was being used as a cover to slip material out which might need to avoid scrutiny.
A more charitable interpretation is that NEHTA was concerned that I may have been bored over Christmas and wanted to keep me busy!
What was released? There were three documents.
The first document is entitled “Technical Architecture for Implementing Services Concepts and Patterns Version 1.0 – 21 December 2006 For Comment.”
The purpose of the document is said to be:
“This document describes the technical concepts and patterns for service implementation within the national E-Health Infrastructure. These technical concepts and patterns form the basis for a Technical Architecture of the national E-Health Infrastructure.
The national E-Health Infrastructure is the infrastructure that is being developed by the National E-Health Transition Authority (NEHTA) to support electronic health in Australia.”
The document is also intended to provide guidance as to how to develop systems which will make up the national E-Health Infrastructure.
Bluntly, this 27 page document is a joke. All it offers is a Services Orientated Architecture tutorial and a recommendation that information flowing between the various services be modelled as follows:
“Information can be modelled and represented in many different ways. This technical pattern recommends that information can be represented using the XML infoset and encoded using the syntax of XML. It also recommends the use of XML Schema to describe those XML documents.
This technical pattern ensures that the data works natively with Web services. Other forms of data can be used with Web services. However, they would have to be represented as binary data which is encoded inside an XML wrapper.”
It then concludes with a set of totally obvious and self evident architectural principles – as an Appendix – but offers no framework as to how these are to be implemented and no use cases so the implications can be understood.
This is another “get no one anywhere” effort from NEHTA.
The second document is entitled “NATIONAL PATHOLOGY TERMINOLOGY Draft National Reference Lists – Pathology Request and Results” and comes with two spreadsheets of the SNOMED CT codes, the fully specified term and the NEHTA preferred term and some linkages.
The most amazing thing about this release is that – while being at best half done -NEHTA is claiming it is a major step forward and at the same time saying this is not ready for use!
“NEHTA has delivered Release 1 of a national pathology terminology, comprised of draft National Pathology Request and Result Reference Lists for comment. It is important to note that these Reference Lists are not yet considered suitable for implementation.
These Lists are the first step towards the development of a national, standard pathology terminology, for use in all general practice, pathology and clinical information systems.”
Worse it just dismisses all the work done with AusPath by airily saying LOINC is not supported – despite being at the heart of ELINCS – the pathology messaging approach now being adopted and supported by HL7 in the USA! What is it NEHTA knows that HL7 does not I wonder? Some detailed report of the analysis that led to a decision of this importance is surely warranted?
There is also some evidence of haste in the completeness of content and spelling in some areas.
All in all this could have waited for release until it had been properly quality controlled, spelling normalised and Australianised and tested in a trial implementation or two prior to public release. Given the resources available to work on this area I would expect more refined and developed products.
The third document is entitled National Discharge Summary Data Content Specifications Version 1.0 – 21/12/2006. Staggeringly this document has been under-development for well over three years – first as part of the HealthConnect program and now (since 2004) as a NEHTA initiative. The time taken, for what should have been a relatively straightforward task, is a true reflection of the lack of focus and direction that exists in the E-Health sector under NEHTA’s leadership.
NEHTA described the document thus:
“This document describes a specification for standardising the content of a discharge summary. The specification is a template that divides the discharge summary into sections based upon topic-specific data groups such as medications, problems/ diagnoses, diagnostic investigations, etc. The template is part of the care record summary suite of specifications that NEHTA is developing for the Australian health informatics community. One of NEHTA’s goals is to standardise the suite of priority care record summaries and their data content to achieve semantic interoperability amongst healthcare provider systems.”
The document runs to 374 pages (or so) and defines a Discharge Summary Template of genuinely biblical complexity. While the rest of the world works to simplify clinical communication of key information (as seen in the recent design changes to the UK shared record and in developments like the Care Record Summary (HL7) and the Continuity of Care Record (ASTM) in the USA) NEHTA has been investing time and effort in rampant, essentially unimplementable information modelling overkill.
Frankly I have no idea how anyone could persuade busy interns to promptly and correctly fill in forms of this planned complexity – no matter what the incentive. This will simply not be used in my view – defeating its noble purpose.
There is a critical need for the simplest of basic information communication between hospitals and practitioners – about the amount of information that can sensibly fit on an A4 page – not this over engineered monster.
The adage of “walk before you run” is ringing in my ears as I type.
Frankly I admire the dedication and work ethic of those who have put this together – despite its obvious lack of practicality. Pity about the out of touch leadership who sponsored such a brave, but ultimately what I believe will be a fruitless, exercise. Let’s get this excellent team to review where the US and the UK have headed for sound and practical reasons and quickly come up with a basic sharable specification that has some hope of being implemented. For heaven’s sake let’s master walking before we try and run!
It can be done! This is shown by New Zealand where they now have approx 2.5 million summaries going out to GPs annually, using an HL7 2.2 standard. I understand this is now being upgraded to V2.4 after six years of getting momentum and use with a very basic standard. Seems like the Kiwis have a jump on us in more than Rugby
A unifying flaw in all these documents is the lack of any reference implementations to confirm any of this is useful or valuable to even the minutest extent. Standards organisations have, I believe, a responsibility to prove what they propose works and can be successfully implemented before expecting it to be adopted. All this is a long way from passing that test.
All in all this looks to me like the response to a command from on high to “get as much as possible out before Christmas to show the last two and a half years have not been wasted”. What do you think?
David.
Hi David,
ReplyDeleteI have been following your blog with great interest, I have had experience as a Project Officer in a 12 month project which aimed to capture clinical data (from the Medical Director product) and convert this to the OpenEHR standard for upload to a central server. This information, in theory, would be accessible to authorised players such as Hospital Emergency Departments, After Hours GPs, Home Nursing groups and the patient's GP.
It all sounds great, and there are many examples of good indictions that this type of intitiative will work, but I must admit it didn't take long for the combination of special interests, commercial IP, proprietary ("lock in") systems, patch protection, security issues, and good old-fashioned FUD (Fear, Uncertainty and Doubt)to kill off any large-scale implementation effort.
Perhaps we need to look at establishing real technical 'ground rules' by looking at something like the OSI communication model - which defines layers of abstraction, each isolated from the rest - so that communications can be allowed to develop at a more targetted level rather than the big bang approach, which I doubt will ever succeed.
This OSI model underpins TCP/IP which of course makes the Internet possible, so I think there is some merit in looking at e-health communications in this fashion.
Just a thought, but keep up the blog, it's ultimately great for the tax payer that NEHTA know someone is looking over their shoulder!