Comments on: Further on the US Healthcare IT Standards Debate

By: Simplified Interop: You Want it When?

Simplified Interop: You Want it When? — Thu, 03 Dec 2009 05:10:21 +0000

[...] there are a bunch of other use cases in David’s categories I and II. Beyond that, the lesson of the Internet crowd is that if we get something simple actually into operation we will find that enterprising [...]

By: Martijn Linssen

Martijn Linssen — Sun, 15 Nov 2009 13:09:08 +0000

We EDI guys (yup!) were really happy with our X12 and EDIFACT over FTP 20-10 years ago

The introduction of XML along with the formalisation of IT architecture as an actual profession have lowered the threshold on the one side, and obfuscated the debate on the other side. How come we allow architects to daze away in cloud metaphors?

The self-describing aspect of XML makes it totally unfit for conducting B2B or any other machine-to-machine interaction.
XML is fun for humans on one side of the line, but machines are not supposed to just suddenly change human agreements based on a single self-describing document that breaks the rules

SOAP and WS are immature and already so complex and compromised that they’ll just never work. Take a look at SMTP: that works perfectly for everyone, why not adopt that mechanism?

If we could just agree on the semantics and write that down in a legible, human, business-information form, then we can then let the techies loose on mess(ag)ing that up. The outcome is easy: the simplest (= cheapest) messages will win

More on my blog (http://martijnlinssen.blogspot.com/) about all that…

By: uberVU - social comments

uberVU - social comments — Sat, 14 Nov 2009 23:01:54 +0000

Social comments and analytics for this post…

This post was mentioned on Twitter by eleddy: My everyday battle: US healthcare [tech] sucks because “healthcare informatics” wanks are winning – http://bit.ly/7JvQS #ugh…

By: Gerard Freriks

Gerard Freriks — Sat, 14 Nov 2009 05:52:50 +0000

Wess I think you presented the present situation aptly.

There are two paradigms. Each with its strengths and problems:
- Message paradigm: Exchange between proprietary databases using an intermediate message information model (e.g. HL7 RIM based) The implementation of messages is time and money consuming. This paradigm is an example of the case of the healthcare informatics crowd.
-Two Level Model paradigm: Exchange (federation even) between standardised databases using a simple interface that Gets, Puts, Publishes, etc and self describing artefacts that get exchanged.
The self describing artefacts (objects) are called Archetypes and Templates. Archetypes and templates are implemented instantaniously without (re-)programming. This paradigm corresponds to the RESTfull case and the Internet folks.

At this moment there is an accepted CEN and ISO standard (EN13606) and an implementable specification that defines this Two-Level Model paradigm, because one model is implemented in the system and a second model is used to define archetypes and templates that define records and records fragments. The tools based on this second model allow healthcare providers to define what they need to store, retrieve, exchange and archive. And archetype/template is implemented in IT-systems on the fly.

Using this approach it was possible to have two proprietary hospitals systems produce and use the same patient summary. It took only two weeks to develop it. Agile development was possible, since each improvement was only a change in the archetype using the archetype editor.

By: Charles Parisot

Charles Parisot — Fri, 13 Nov 2009 18:02:18 +0000

I like very much your analysis, but I am wondering who may really be this \Healthcare IT crowd\.
If we compare apples-to-apples, the discussion should be between REST versus SOAP based Web Services. This is a rather internal IT debate, one between the MIcrosoft, Oracle, Intel, IBM, Linux crowd that have been pushing SOAP based WS in WC3 for several years, and the Internet Crowd of these same companies.
I am as much surprised as John Halamka, quoting his blog (the power of AND):
“The work I’ve done for the past 4 years aimed at unifying the industry on a web services approach, embracing web-centric standards such as SOAP, XML, and HTTPS. In 2006-2007, this was considered very forward looking. In 2009, RESTful data exchange of simple payloads with TLS and application level security is considered cutting edge.”
I too am puzzled that after having been encouraged to adopt SOAP and WS in healthcare, we may be faced with an attempt to turn healthcare in an IT battle field, where the real HealthCare IT crowd may simply be a sort of hostage.
Should this real healthcare IT crowd not exert a liitle wisdom here and step back, let that debate be sorted within the IT/internet Industry and see where the puck falls ? SOAP-based WS have a wide acceptance basis, and feel to me like the safer side to be on for the time being.

By: Charles Parisot

Charles Parisot — Fri, 13 Nov 2009 17:56:28 +0000

I like very much your analysis, but I am wondering who may really be this \Healthcare IT crowd\.
If we compare apples-to-apples, the discussion should be between REST versus SOAP based Web Services. This is a rather internal IT debate, one between the MIcrosoft, Oracle, Intel, IBM, Linux crowd that have been pushing SOAP based WS in WC3 for several years, and the Internet Crowd of these same companies.
I am as much surprised as John Halamka, quoting his blog (the power of AND):
\The work I’ve done for the past 4 years aimed at unifying the industry on a web services approach, embracing web-centric standards such as SOAP, XML, and HTTPS. In 2006-2007, this was considered very forward looking. In 2009, RESTful data exchange of simple payloads with TLS and application level security is considered cutting edge.\
I too am puzzled that after having been encouraged to adopt SOAP and WS in healthcare, we may be faced with an attempt to turn healthcare in an IT battle field, where the real HealthCare IT crowd may simply be a sort of hostage.
Should this real healthcare IT crowd not exert a liitle wisdom here and step back, let that debate be sorted within the IT/internet Industry and see where the puck falls ? SOAP-based WS have a wide acceptance basis, and feel to me like the safer side to be on for the time being.

By: Tweets that mention Further on the US Healthcare IT Standards Debate -- Topsy.com

Tue, 10 Nov 2009 23:10:15 +0000

[...] This post was mentioned on Twitter by Chris Winters, Elizabeth Leddy. Elizabeth Leddy said: My everyday battle: US healthcare [tech] sucks because "healthcare informatics" wanks are winning – http://bit.ly/7JvQS #ugh [...]

By: Brian Wells

Brian Wells — Mon, 09 Nov 2009 20:15:56 +0000

Wes – Great post! Having spent years coding software solutions that had to succeed in both camps I am well aware of the limitations and powers that reside in each. It would be a huge help for me to assess the applicability of the Internet camp approach, to learn of a non-healthcare business application domain (that has as complicated a data model as healthcare) that is successfully interoperating between disparate entities using Internet protocols without predefined content standards. I just cannot see how that could work in a dependable way.

Said another way, is there something more complicated than e-commerce type transactions going on over simply Internet protocols that we can all learn from and apply to healthcare?

I am an admitted healthcare centric cynic that is open to a new way if someone can show me one that works.

BPW

By: David McCallie

David McCallie — Mon, 09 Nov 2009 17:08:45 +0000

Good post Wes.

I think it would be useful to further clarify the notion of “architecturally neutral.” The HIT Policy Committee used that term to define a core constraint for HIE. What did they mean?

I suspect that what is most important are the following:

* the architecture is based on open protocols and open standards
* the architecture is not constrained by any proprietary hooks such as hidden patents, etc.
* the architecture can run on any modern operating system
* the architecture can be made accessible from any modern device
* the architecture can be developed using standard tools that are readily available to all

As a bonus:

* the architecture is simple enough that massive training is not required.
* the core components of the architecture are available as open source software

The Internet fulfilled all of these, yet it there was clearly a well-defined and precise “architecture” present. It would be a mistake to assume that the NHIN or any other national system of HIE could exist without some kind of “architecture” — preferably one that met the above criteria.