It has taken me a little while to notice what a biggie this is:
My Health Record on a path to
decentralisation
Justin Hendry
Innovation Aus
20 June 2023
My Health Record could adopt a new model where health data is primarily stored
in its original clinical system of record from as early as next year,
potentially reducing the appeal of the system as a honeypot.
The Australian Digital Health Agency has begun working with industry,
healthcare providers and governments to design the plan, which relates to a new
National Health Information Exchange, but is yet to receive sign-off.
ADHA Chief digital officer Peter O’Halloran, who joined the agency from ACT Health
in February, told the Digital Transformation Live conference that My Health
Record was moving to become the national repository for core clinical
information without storing all the data.
The centralised record currently includes information about an individual’s
health from a range of sources, including from the Australian Immunisation
Register, and Medicare Benefits Schedule and Pharmaceutical Benefits Scheme.
“We’re… talking about My Health Record moving away from being the be all and
end all to being the core national repository for core clinical information
about healthcare consumers, but not all data being stored in there,” he said
earlier this month.
“The concept we’re moving towards is that in many cases, data will be stored in
original system of record. The My Health Record system will be a core system to
move forward that gives a summary of what’s happening for a patient but doesn’t
do everything.”
The comments come just weeks after $429.7 million was set aside in the federal
Budget for a two-year modernisation of the My Health Record, which the
Strengthening Medicare Taskforce in February found was a barrier to health
information sharing.
Health minister has acknowledged that the “clunky, pdf format system” that
underpins My Health Record is outdated, with its future design expected to be
capable of supporting a “real-time, fully-integrated digital health system”.
Mr O’Halloran said around $38 million of that funding will be used for a data
platform to transition My Health Record from a pdf-based clinical document
system to a structured data system. A further $13 million will got to
implementing a “share by default” setting for health providers.
Lots more here:
https://www.innovationaus.com/my-health-record-on-a-path-to-decentralisation/
Reading this one really does have a sense of coming full cycle with the apparent recognition that it may not be a wonderful idea to cram the totality of clinical information held in the myHR in one huge database and that some distributed alternatives may make more sense!
There are pros and cons for each approach but I have to say that when dealing with sensitive private data there is some value in using a distributed approach to minimise the adverse consequences of data leakage.
Of course, there can be some real technical complexity in making the distributed approach work from a diverse range of sources and a good deal of planning and discipline is needed to have it all have together!
I have to say I would not expect any transition to be quick or pain-free!
David.
18 comments:
They lack basic fiduciary duty, why would we trust them to run a bath let alone this sort of monumental shift of concepts? This really has become a vanity project, a luxury item we can not afford at present.
This is the original architecture - connect existing sources of health/medical data and add some summary data in a central repository.
If you look at the new architecture, there is an additional feature at the back-end which allows the government and researchers to access all the health data in the system.
The original architecture had a feature, NASH, that identified every individual health provider who accesses a patient's data. That was never implemented and does not seem to have made it into the new architecture.
There is a reason why the original architecture was not implemented - it was all too hard. It remains to be seen if ADHA can do what was not possible over ten years ago.
It also remains to be seen if creating a firehose of uncurated data has any useful purpose or benefit.
"There are pros and cons for each approach but I have to say that when dealing with sensitive private data there is some value in using a distributed approach to minimise the adverse consequences of data leakage."
I would challenge that view. Once a hacker got access to the system, they would be able to see everything on record about a patient. There is more security in isolated systems.
I believe you will find bigger challenges than just the technical implementation.
On the technical front, just what do you do with all that data with legislative retention wrap around it?
The conformant repository approach also treated the original data stores as though they were a government data stores, along with all of the encumbrances of government data governance. There would need to be some nimble policy work to make this work for both private sector and government purposes.
Some things are abundantly clear.
1. the Medicare Taskforce found the MHR was a barrier to health information sharing
2. the “clunky, pdf format system” underpinning the MHR is outdated
3. the centralised My Health Record currently includes Australian Immunisation Register, Medicare Benefits Schedule Data, Pharmaceutical Benefits Scheme Data with the last two being quite useless
4. the ADHA tried a centralised data repository approach, but it didn’t work so now they are going to try a decentralised approach with information being stored in the original clinical system.
The most problematic aspect of the above is that a cohort of experts have been trying unsuccessfully for many years to get the government and the ADHA to understand that what the ADHA had developed and the approach the ADHA was using was never going to work. At long last it seems they may have accepted these facts. That being so the big questions now become:
Q1. Do they understand what they have been doing wrong?
Q2. Do they understand what is the health-system problem that needs to be solved and how to go about it? or,
Q3. Are we about to witness another decade of more of the same in a different guise?
The following provides a clear indication of what the answers to Q1, Q2 and Q3 are likely to be.
We are being told that:
- a new National Health Information Exchange will be needed
- a plan for the new Exchange is being designed
- the My Health Record will no longer be the core national repository for storing core clinical information about healthcare consumers, because only some data will be stored in the MHR
- the MHR will be “a summary of what’s happening for a patient”
- $429.7 million has been budgeted for a two-year modernisation of MHR
- and “its future design is expected to be capable of supporting a “real-time, fully-integrated digital health system”.
All for the intent. 2 years to modernise MyHR?
Is modernising MyHR the same as changing to a new health information exchange?
Have national consultations started?
What is the changes needed in legislation, policy etc?
How does consent work?
What work is required for source systems?
Etc. etc,
2 years? Sounds bold Minister or sounds like someone has been dropping a few to many pptx, or FHIR is the answer, just sign a few grants and we will save the MyHR.
Supportive of the shift towards a real-time, fully integrated healthcare system. I was unaware that layers of agreements had been reached and that the technology was implementable to have this injected directly into clinical workflows to support clinical decision-making. Can we expect the market for personal health apps to blossom now my health data is accessible, allowing for my personal control over who I do and do not gift it to for whatever reason?
Most impressed, well done ADHA, CSiRO. and HL7-Aus.
The fundamental problem is managerialism. The "elite" managerial bureaucracy within the health department has been given billions to play with and they are surrounded by snouts feeding off a very rich trough who would never think of pointing out fundamental flaws in the strategy. The next meal is ensured through the failure of the last plan. As with many things central planning is the worst possible path to success when more $$$ is the reward for failure.
I have no confidence in any of the ADHA plans, they continually put the cart before the horse and cover up deep failures and poor choices by plastering over the gaping cracks with taxpayer dollars. There is no understanding of fundamental IT concepts and we have seen the management of many IT entities and organizations become dominated by people with knowledge of how to appease the overlords and get their trough filled tomorrow, rather than solve the technical problems. It's all about "change management" apparently? I disagree, its about compliance with low level standards and basic code quality issues that makes building anything of value a challenge.
We actually need a withdrawal of these obscene amounts of taxpayer $$ flowing. They go to people who appease the bureaucratic overlords rather than allowing what works to flourish. People who have to pay themselves will be far more interested in success and quality, some qualities that have been buried under the avalanche of government $$ that continue to flow as if there was an endless supply.
"We can't solve today's problems with the mentality that created them."
Albert Einstein
The first principles, the basics, the fundamentals, the obvious, the essentials, so relevant to developing a real-time, fully-integrated digital health system, continue to be overlooked ,repeatedly in favour of new esoteric, seductive, technologies so loved by bureaucrats, politicians and technology practitioners, of every persuasion.
@8:51 AM Bernard noted that "There is a reason why the original architecture was not implemented - it was all too hard. It remains to be seen if ADHA can do what was not possible over ten years ago."
And to reinforce this most potent of observations we should also add:
Where in the world is there a demonstrable working model of the architecture being proposed today which was also proposed 10 years ago and not implemented?
@2:01 PM "4. the ADHA tried a centralised data repository approach, but it didn’t work so now they are going to try a decentralised approach with information being stored in the original clinical system.
BUT
they tried originally attempted to use a decentralised approach 10 years ago and they failed to implement it because it was too hard!
Are they planning to use blockchain this time around which didn't exist 10 years ago?
The search for a silver bullet goes on.
"The first principles, the basics, the fundamentals, the obvious, the essentials, so relevant to developing a real-time, fully-integrated digital health system, continue to be overlooked ,repeatedly in favour of new esoteric, seductive, technologies so loved by bureaucrats, politicians and technology practitioners, of every persuasion."
Yes exactly! To achieve that you need a messaging system with workflow built in and transfer of data in a lossless manner. What amazes me is how well that was understood by the people who developed HL7 V2 and how efficient it is in terms of message size and storage requirements. A document repository of pdf versions of data with no workflow built in is a joke, its a multi-billion $ joke which has been played on the Australian public. No decision support possible, just try leafing through potentially thousands of pdf documents?
Realtime HL7V2 is very possible, I have been doing that for decades and the workflow that it allows for is very complete, way beyond what is imagined today, I am in awe of its designers, I think they stand head and shoulders above what has come after and has continually failed miserably. Possibly the 90s was Peak Health IT and we are now in the dark ages?
"What amazes me is how well that was understood by the people who developed HL7 V2 and how efficient it is in terms of message size and storage requirements. A document repository of pdf versions of data with no workflow built in is a joke, its a multi-billion $ joke which has been played on the Australian public."
Yes exactly!
The problem with HL7 V2 is that for years, since NeHTA days, 'all the so-called knowledgeable experts' (except for a few experienced practitioners like you) have discounted HL7 V2 in favour of the latest 'technologies'; ie. out with the old and in with the new. Their reasoning being that the old is out-of-date and the new is the way-forward into the new future.
Embracing the 'old' was too stodgy for all the 'so-called knowledgeable experts' - they knew better - or so they thought, and look at where it's got us! Oh, and the other thing I need to say is that practitioners with an extensive proven track record of success behind them (like you ad some others) were not listened to despite repeated efforts to be heard.
Whilst I am not confident these huge barriers can be overcome we should not stop trying to be heard.
Sept 22 3:14 PM - I challenge that thinking: would you have humankind taking clinical notes by spitting paint on a cave wall? Without new consideration, we would not advance. Perhaps it is not a question of one standard over another standard that is at question here (after all hundreds of standards come together to deliver a healthcare service) but rather the lack of recognition of the value of standards and the commitment to meeting conformance and compliance agreements that has let us down is the software world?
"I challenge that thinking: would you have humankind taking clinical notes by spitting paint on a cave wall?"
Well that art still exists, which I doubt will be the case for digital art in 30,000 years time.
The main issue is the focus on documents, rather than documents/orders etc as part of a workflow with messages to support that workflow. HL7V2 supports documents in messages with the workflow designed into the system from the ground up.
It also promotes atomic data which allows for decision support to, for example look, at Liver function tests over time. The learned government solution is a large collection of badly labelled opaque pdfs that require a human to look at. This is the electronic equivalent of a stack of paper results, which was my motivation to advance eHealth. Imaging opening a patients record with 2000 pdfs where you have to open each pdf to even know if they include liver function tests. With atomic data you could (and can) ask for a graph of all liver tests over time and have it within seconds.
When you look at the V2 specs you see complex workflow is built in and its clear the designers were switched on. The later standards focus on documents and often there is no workflow and a lack of atomic data suitable for decision support. So even the documents are dumbed down! We have seen a focus on trendy technologies in an endless stream of irrelevancy to the underlying problems. eg Must use xml/json/soap/xml encryption rather than the core functionality and understanding of the data quality and workflow.
Yes compliance, or lack of it is a huge problem, but when you keep distracting vendors with the latest trendy technology they lose focus on doing the core job well. It appears that the "elite planners" don't understand the core requirements so just dismiss the quality and compliance issues as "geek stuff" and focus on the latest trendy fix which is mostly irrelevant to the core problems.
It's quite clear to me that their solutions will fall over, no matter how many $$ they slip into the pockets of multinational vendors. Its a challenge to not be distracted by fools however.
Post a Comment