For possibly the first time ever I think Tom and I agree on a messaging
issue! Good on ya Tom.
I think the messaging standard for pathology in this country has been
set and the practice management vendors have an obligation and
responsibility to implement that standard for the good of the
patients..Perhaps they have even a duty of care issue...
Acknowledgements of message delivery with HL7 are vital to saving
peoples lives and having a third party app return an ack that it reached
the MD2 directory is just not good enough. MD must report it
successfully imported the message.
Given there is a national patholology message standard in place and that
willing participation by the clincial package vendors in pathology
messaging is infered by the ability to import messages at all then,
perhaps there is a legal platform that in a case where a result "got to
the MD2 or other system import directory" (provable with third party
ack) and the clinical package silently never imported the file that
there is a failure on not meeting industry standards.....to my small
brain kind of like a toy that fails safety standards and causes harm to
a child. I know the standards are probably compulsory for the toy, but
just the same the outcome could be deadly....
food for thought:
punch biopsy of pigmented lesion --> recommend complete excision to
exclude melanoma gets missed
punch biopsy of clinically innocuous looking skin lesion --> amelanotic
melanoma gets missed
Can the software vendors afford to test the legal issue out? Its one of
those risk outcome things where the risk might be small but the outcome
potentially enormously bad for the vendor in the not completely unlikely
event a court made them partly or fully responsible for a missed test
result.
FFS just implement the damn acks as it should be "core" and "best practice"
Anyone able to show us how hard an ack is to generate? Im guessing its
about 5 lines long..has a little demographics, a unique message ID and
really wouldnt be that hard to produce for each imported .ORU
JD
Tom Bowden wrote:
Dear Colleagues,
At HealthLink we have been watching the discussion on messaging
responsibilities with great interest. It was timely that David More
provided an excellent link to a video on use of messaging etc in the
Dutch health system. The key point we think should be noted is that you
cannot get to this level of automation without all involved having
complete trust in the system, especially trust in the fact that it is
safe and 100% reliable
As readers are probably aware, HealthLink is a messaging and security
system provider active in Australia, New Zealand and Canada and
therefore we have a number of environments upon which to draw upon for
examples and comparisons.
In New Zealand, the issue of 'where the responsibility lies' was
addressed more than a decade a go when, after a terrible incident in
which a young girl's urinary tract infection went untreated for a week
as a result of an undelivered lab result. Fortunately the labs conceded
that it was indeed their responsibility to ensure that the information
was correctly delivered into the database of the computer system
operated by the intended recipient's practice (by monitoring receipt of
application level acknowledgements. It was clear then, just as it is now
that 100% reliability could only be achieved via fastidious monitoring
of application level acknowledgements and using systems that are truly
integrated with the intended recipient's clinical systems.
As a consequence of this clear delineation of responsibility, trust in
the system has continued to increase to a point at which the entire
Health system is now electronic, (a bit of parochial commentary follows)
NZ is even better than the Dutch system from that POV.
As Craig Barnett points out, HL7 is designed by its very nature to do
precisely this (provide guaranteed delivery into the clinical
application). When HL7 is correctly implemented by sending application
acknowledgements from the intended recipient's system back to the
sender,the HL7 standard can be fully relied upon to perform this
function. When we only rely on transport acknowledgements (also known as
accept acknowledgements) generated by the intermediary messaging system
then complete trust is NOT achieved.
If we wish to build a trustworthy, reliable and scaleable e-Health
system we must aim for the highest of standards. We must have
appropriate respect for patient safety and the necessity of reducing
clinicians' risk exposure. The requisite communications standard is
there and can be implemented.
It may or may not be known that HealthLink has proposed that all of the
parties involved in electronic messaging in the Australian health sector
be party to a code of practice that sets out the responsibilities
entailed in doing clinical messaging correctly. Following is a link to
a draft discussion document entitled 'Safety through Quality' that
backgrounds three key quality issues and presents the draft code of
practice we have proposed.
Please feel free to read it and publish your comments on this list, we'd
be keen to incorporate any useful input.
If we take a quality approach to electronic communications and security
across the health sector (reducing the unnecessary risk in patient
safety associated with poorly implemented or outdated messaging
systems), we may one day be in a position to match what is happening in
other countries across the world. Until then we can only watch the other
countries sail past us.
Kind regards,
Tom Bowden and Geoffrey Sayer (HealthLink Ltd)
http://www.healthlink.net/healthlink_documents/brochure/Electronic%20mes
saging%20safety%20Issues%20-%20HealthLink%20viewpoint.pdf
_______________________________________________
Gpcg_talk mailing list
[email protected]
http://ozdocit.org/cgi-bin/mailman/listinfo/gpcg_talk
_______________________________________________
Gpcg_talk mailing list
[email protected]
http://ozdocit.org/cgi-bin/mailman/listinfo/gpcg_talk
