Hello All,
some comments on a few recent posts..
DHN> If NEHTA mandated a small development team to provide tools to translate each independent
DHN> export XML schema into a standard format, then all the companies would find themselves
DHN> in an embarassing situation if they objected to it, and also if they did not provide tools
DHN> to import from it.
Probably more important is the need to establish potential use of the data
Stipulate an open standard with certain components including medication
list
We quickly start running into all the old walls. We have no reliable way to identify individuals.
we can't agree on coding systems for diseases, and don't have any real candidates for coding medications, NETHA are
still thinking about forming a committee to discuss a work plan for key deliverables.....
DHN> In 1993, many people objected to using PIT format as an interim
DHN> measure. "Wait for agreed standards" they said.
DHN> Still waiting.
DHN> Hugh
DHN>
DHN> XML export is good. XML import is better. This presumes there is some
DHN> sort of agreed schema. In fact, they are all different. None of EHR
DHN> developers publish them and if you use something approximating HCN's
DHN> they'll take you to court.
DHN>
DHN> I feel depressed.
DHN>
DHN> David
I think it is a big mistake to try and come up with a "New" format to
transfer data. You could come up with an XML schema that would be
simple and do what *you* want, and then someone else would want a
little bit more detail, some atomic data etc and try and extend it.
After about 20 years of this and numerous committee meetings you would
end up reinventing HL7. HL7 V2.x has been around in some form for that
long and may not be perfect but can and is being used to transfer
atomic data. Its supported all over the world and by the vast majority
of labs in Australia. HL7 V3 is still out in the wilderness but some
say the sightings look encouraging - but I am less sure.
If you don't think you need HL7 then PIT is at least of some use, but
is not a good choice moving forward.
If you want an XML format then HL7 V2.3.1 has an XML schema already
and various free tools can convert between classic and XML format.
If you say HL7 is too complex then you have tunnel vision and can only
see your specific needs. Simple Text reports are actually quite simple
in HL7. It has to be able to handle the complexity of the data
requirements of all health care providers and reports, not just the
simple case.
Otherwise we end up supporting HL7 from Labs, PIT, GP1XML, GP2XML,
SPECIALISTXML, "Standard RTF" "Standard Ascii" - surely one standard
that can do it all is better?
The lack of a National Medications terminology is an absolute disgrace
considering that's really what government should be doing, not sure
what happened to the $250M that was spend on Healthconnect.
Snomed CT is the national terminology for the UK and US and has been
nominated as the choice for Australia, you can dispute the choice (I
actually think its a good choice) but if we have made a choice we need
to get on with it and make it available.
The lack of a reliable patient identifier is a problem, but only a
huge problem when you try and aggregate large amounts of data, within
any one practice it can usually be handled reasonably. You are "not
allowed" to use the medicare number by law ?????
Provider numbers are an excellent way to identify private providers
and are actually specified in the Australian HL7 standard. Based on
the track record of Government IT performance, waiting for a National
identifier is not wise. Now days most allied health providers also
have a provider number. Of course the lists of providers and provider
numbers are a "Top Secret" and not available unless you are a path
lab. Being required for getting medicare payment usually encourages
people to actually obtain this number.
The walls are more imagined than real. We have national HL7 Standards
and even have a free testing facility - see http://www.ahml.com.au. We
have PKI/PGP/GNUPG etc Rich semantic communication is hampered by the
lack of standard terminologies but basic communication is not. So far
in 2005 Buderim GE Centre has delivered 13894 HL7 clinical reports to
local GPs and specialists, and on the Sunshine Coast alone, nearly
100,000 clinical reports have delivered between Specialists and GPs.
I is doable and is being done, I think waiting for NEHTA to act is not
going to be very fruitful.
The idea of an Australia wide VPN is a bit strange, do you trust all
other health care providers in the country? Once it goes outside your
organisation the usefulness of a VPN drops of very sharply. Public Key
based Authentication seems the only workable option to me.
All the talk of "Web Services" is a distraction. Any messaging
environment should have a "Service Orientated Architecture" by
default. What would be more useful is interface definitions and
national standards for EMail, HTTP Web Services, XML Web Services etc.
Saying "Web Services" sounds like a repetition of whats been said at
the Microsoft sponsored dinners rather than a solution.
