Here are some thoughts from Grahame - republished with permission.
An alternative approach for resolving the Secure Messaging dilemma in Australia
Earlier this year, just as Covid-19 was getting going, I was talking to Nathan Pinskier about the many challenges fixing GPs at this time. Nathan asked me whether there was an alternative approach to solving the secure messaging problem – which is a consistent pain point for medical practitioners, especially this year.
I thought about it and wrote up a brief description of an alternative approach that I believed would offer a way forward. For me, it was most important to solve 2 inter-related problems:
- There needs to be a clear business rationale for all parties to get involved (vendors, GPs, specialists, hospitals, public health authorities)
- We need to solve the directory problem – how do you pay for maintaining it? (or who pays – it’s a key stumbling block)
A wider context is that the whole secure messaging concept as dying anyway – I already posted about that. These were the design inputs I considered when I wrote the document, and reviews from selected friends were enthusiastic.
However I did not want to derail the existing SMD project that the agency was running (or be seen to try to do that). If there was any chance that it would deliver, I thought that would be a better outcome even if it was a sub-optimal approach. But six months later, it seems clear that this program has run out of steam without actually making any difference to market outcomes, and there are participants out there looking for a different approach.
So here’s my proposal:
Note that this is just a rough outline – a lot of water needs to go under the bridge before it’s a solid proposal. I’m publishing it to stimulate thinking, and to suggest directions to various technical teams already thinking about this.
I want to thank Nathan Pinskier for starting me thinking about this (and also writing the first draft of the introduction), Brett Esler for technical input, and also Reuben Daniels and Andy Bond for review and comments.
Here is the link:
http://www.healthintersections.com.au/?p=3051
David.
28 comments:
Good to see some familiar names still working selfishly in the background. Might be hope yet.
HL7V2 is a messaging standard that can (and is) directly messaged over the internet using NASH and http. The issue is the ability of ALL the endpoints to handle and reliably render the content. FHIR is an alternative, but is not simple. Both depend on good tested implementations, its a case of same problem, different standard. I would welcome any move to insist on software quality, but would probably start with whats currently in use!
Well, in fact, this approach would stick to v2 ref messages, and pretty much is v2 ref messages over http + NASH.
Well HL7V2 over http + NASH/PKI directly messaged has been in place for over a decade on the Medical-Objects network so I can tell you it works. Its what and how reliably the endpoints handle the content that is the main problem. Universal adoption of NASH is required to make it seamless and once one endpoint uses a different PKI you have to fallback to your own PKI infrastructure to make things work.
Every GP surgery that uses medical-objects has had the web facing infrastructure in place for direct messaging via REST/PKI for over a decade.
Yep. In fact pretty much every vendor can say the same thing, but all the vendors are doing it their own way. That's not in anyone else's interest, and I don't think it's even in the vendor's interest either. Stop splitting the pie up, and collaborate to build a much bigger one.
Payload Compliance, both in and out is a 100% necessary requirement for commodity messaging to appear. If you don't understand that, then you are part of the problem.
I hesitate to make this observation but here goes.
It seems this 'conversation' is turning full circle 'once again'.
When I try to 'sift' the chaff from the grain each time Andrew, Grahame, and some others put their views forward (forcefully, respectively and convincingly) I consistently come to the 'conclusion' that Andrew's approach using well-established mature proven technology has been working successfully in many hundreds of sites for a long time.
I have not yet seen any compelling arguments as to its shortcomings and most importantly why that approach and technology should not be more widely adopted, nurtured and supported, to progress digital health.
I do acknowledge there are strong, often conflicting, views in play and that I do not fully comprehend the effect of various egos in the mix. Even so, it really does seem to me that the Medical-Objects approach has an awful lot going for it and that there is a reluctance to accept that!
@2:48 PM
What you have described is symptomatic of the whole Digital Health movement. Vendors are pushing their profits and make extravagant claims regarding the benefits of their products, usually only from the perspective of those products. From cloud solutions, block chain products, telehealth, interoperability to EHRs, EMRs, the variety is wondrous and full of potential and promise. Governments and a few medics are interested in saving money and believe the hype so they support it thinking that the magic must happen sometime soon.
The trouble is Digital Health has little to do with the processes of healthcare itself. Digital Health does nothing to help doctors make better diagnosis (a firehose of data does not help) and even less regarding treatment.
Real advances in medical care are happening elsewhere, they are just not called Digital Health
@3:25PM & 2:48PM I agree absolutely. So, before starting out to develop any solution the people problem needs to be broken down into all its component parts, then the business model(s) need to be established to ensure the people problem(s) can be overcome.
@3:45pm
That is a very negative view! I started playing with IT when I was a registrar in 80's because I needed to solve issues and have been using my own software ever since and I guess I am biased but the gain in efficiency as a result of eHealth has been enormous despite all the warts. I can do cumulative results, graphing, calculated values, view guidelines and send ERCP pictures to a surgeon in Brisbane within seconds and obtain pathology within 10s of minutes and regularly transfer bulk results to local surgeons for an opinion. I would like to do more advanced decision support, but the data quality does limit that somewhat. 99% of my reports are at the GP surgery before the patient is out of recovery. I can review results anywhere, even overseas, or in bed!
The potential is enormous, which is why I am still plugging away and always believe in "Eating my own dog food" (Sort of true in real life except the chicken necks ;-) )
4:31 PM You have just described how IT is used in a very small part of healthcare as a business. What you have not described is how IT assists better diagnoses and treatment.
"I would like to do more advanced decision support, but the data quality does limit that somewhat" says it all.
I repeat "The trouble is Digital Health has little to do with the processes of healthcare itself."
@6:28pm Given that I am doing healthcare and see huge advantages, and do a lot of R&D on trying to extend it I will have to disagree. Done well its an enormous support for clinicians who need ways of obtaining and analyzing patient results quickly, thats what it does for me and poor quality from lack of compliance with standards is the major block to making it more useful, especially in areas of medication lists and past history. For a clinician those advantages are a big part of the business. Population level results would be good, but getting point to point transfer working with quality data is first step.
Colleagues, this is a conversation with some rich appreciations of specific challenges and possible solution elements.
It is equally symptomatic of the absence of a larger frame and framework encompassing the human and digital technology aspects of modern day healthcare.
Most specifically it highlights the absence of a properly designed and operating collaboration mechanism able to:
a) operate at a whole-of-health sector level;
b) effectively connect the clinical and health service value propositions with the requisite establishment of clinical information flows and conformance to the norms of health care communications; and
c) attract growing participation by all relevant stakeholders to a new approach
IMHO such collaboration will require government participation BUT not control. In fact such a collaboration must be protected from capture in order to effectively partner the health sector.
I get a sense even the authors are a bit over secure messaging and see the future for the requirement in more modern business and technical approaches. It answered a specific question but I read in the report a broader set of thinking, hidden but it is there.
If you watch the John Pilger movie The Dirty War On The NHS he demonstrates how the NHS is being dismantled and privatised. There is a big push to adopt the American model of healthcare.
It started with Thatcher, was continued by Blair and is still on the agenda of Hancock, the current UK Health Minister.
What is not stated, but which is obvious, is if you are going to break the NHS up and create a multitude of separate health service providers there is one thing you have to have - interoperability.
A tightly integrated health service has little need for formalised data interchange - it is much easier to share raw data as needed.
So don't forget, interoperability is an enabler of a rampant healthcare free market. Are you sure you want to facilitate that outcome? I for one, don't.
Interesting idea. Come to think of it a national medical record would also act as "an enabler of a rampant healthcare free market"
The NHS and Australian healthcare are very different. Australia started as a distributed business model, that is why adopting a centralised solution was always doomed.
> A tightly integrated health service has little need for formalised data interchange - it is much easier to share raw data as needed
I've worked with many countries around the world. Fully integrated systems need the same amount of interoperability as fully decentralised systems, because the challenges of the healthcare systems are too complex for a top down solution. Russia, China, Vietnam, England, USA, Queensland, NSW, Victoria -- all of them have disparate information systems, irrespective of how apparently top down the command and control systems aspire to be.
Interoperability makes for better healthcare outcomes irrespective of the way the healthcare systems are managed.
And also, btw, what I've seen is that both government run and privately run systems are able to deliver awful outcomes. Quality of governance is always what matters (which doesn't make me feel great right now)
Its all about patients and moving patient data to wherever its needed! The "Centralized care" of government facilities are probably the least able to reliably move data to where its needed and have become hopelessly inefficient, at least where I live! This has nothing to do with the politics and if government are trying to make data mobile to privatize health care they are doing a pretty poor job of it, just like everything else they do it seems. So I wouldn't worry because with $2Billion + they have achieved nothing, perhaps even made things go backwards..
"The NHS and Australian healthcare are very different" that doesn't mean they aren't being pushed in the same direction.
"Its all about patients and moving patient data to wherever its needed! The "Centralized care" of government facilities are probably the least able to reliably move data to where its needed and have become hopelessly inefficient"
Integrated doesn't mean centralised. Integrated means that all the patient data is held close to the patient and accessible by all the providers concerned with the patient. It's a distributed system with the patient at the focus of their data and their care.
Yawn, Anon 10:31 you are repeating the same old mis conceptions and rabbit holes of the past.
.... and there is no reason why "... all the patient data needs to be accessible by all the providers concerned ...". This misconception goes to show how little some understand about 'the integrated health record".
@8:57 This misconception goes to show how little some understand about 'the integrated health record"
Who said anything about an "integrated health record"? Health records are an anachronism, a solution looking for a problem. What is needed is an integrated health service.
@12:04 PM Who said anything about an "integrated health record"?Health records are ... a solution looking for a problem. What is needed is an integrated health service.
An integrated health service would be best achieved if there were no patients, for there would then be no need for a health record and there would be no service providers to use the record.
There is nothing anachronistic about a patient having a health record.
What is anachronistic is the way the enthusiasts continue to approach the development of an integrated health system which will meet patients' clinical information record(ing) requirements needed by patients' service providers to investigate, diagnose, and manage, a patient's illness, health and well being.
Exactly what is done with a patients data is irrelevant and will vary depending on the situation and system, however its important that the data if reliably encoded and rendered for however views it or whatever system uses it. This is what compliance is about, high quality data correctly encoded in a message and handling that data reliably at the other end. Its actually a huge safety issue that is unreliable and limited currently. The privacy, security, addressing, messaging and politics of what happens to the data are separate (but important) issues. We currently have a myriad of fixes to allow to data to travel between systems and until this is fixed interoperable messaging is a pipe dream. Messaging via a usb and courier would not work currently, Automate transfer and you just amplify the problems and create mayhem. First get reliable message creation and handling of that message reliable and the information can flow, until then trying to turn on the tap will just get you soaked. Whatever format you use has to be reliable before you can expect free flow of information.
What Andrew is saying appears to be backed by Researcher Professor Meredith Makeham findings. The MyHR is simply acting as a huge and dangerous amplifier. No point spending money cleaning the lake if you allow the rivers to remain polluted.
Prof MM spent 3 years destroying her credibility on anything to do with Digital Health when she beca.e Kelsey's lapdog.
" an integrated health system which will meet patients' clinical information record(ing) requirements needed by patients' service providers to investigate, diagnose, and manage, a patient's illness, health and well being. "
Yes.
But if you read Andrew's comment, he is not describing a traditional health record. And as Sarah points out, MyHR isn't even a useful health record.
Post a Comment