1 Recommendation

Project Spotlight: Interoperability in Australia March 28-April 1

By Joaquin Blaya, PhD Moderator | 23 Mar, 2016

Hi everyone,
We’re trying a new activity type which we’re calling Project Spotlight, where we highlight a specific project. In our first Project Spotlight, we’re going to focus on Australian Interoperability. This Spotlight will occur from Monday, March 28 to Friday, April 1.
Australia has a long history of interoperability and many organizations working on it, especially NEHTA, the Australian National eHealth Transition Agency (https://www.nehta.gov.au/). This project spotlight is meant to help us learn from the Australian experience about what is needed for national level health data interoperability.

We’re very lucky to have 3 excellent presenters from Australia
Tim Blake is the Managing Director at Sematic Consulting and a Strategic eHealth Advisor to the Australian Department of Health. Previously, he was an Advisor to NEHTA and Chief Information Officer for the Tasmanian Department of Health.

Andy Bond is the Manager of Standards and Informatics at NEHTA. He has also been the Manager of Interoperability and Standards and Chief Architect. He received his PhD from the Victoria University of Wellington.

Grahame Grieve is a Principal at Health Intersections Ltd and a Consultant for NEHTA. He is one of the co-founders of HL7 FHIR (Fast Healthcare Interoperable Resource) which is probably one of the most popular health IT standards currently.

I’m including documents that are panelists recommended along with the URLs where they are located for people to read before the spotlight next week.



Jean Galiana Replied at 5:20 PM, 23 Mar 2016

Thank you for this wonderful idea and import and topic.

Noela Prasad Replied at 5:47 PM, 23 Mar 2016

Thank you for the opportunity to 'meet' Australian experts and especially the three persons that can guide navigation of the challenges encountered around interoperability.
An ophthalmologist (erstwhile clinician) and public health professional, I'm the Medical Officer at The Fred Hollows Foundation, an INGO based in Australia.

A/Prof. Terry HANNAN Moderator Replied at 8:42 PM, 23 Mar 2016

Noela, a small world. I worked with an knew Fred Hollows in the Prince of Wales Hospital in Randwick. I saw him 6 weeks before his untimely death and even at that time "life was beautiful"-la bella vita for him. Terry

Elaine Baker Replied at 4:13 PM, 24 Mar 2016

Thank you for setting this up and for the great links - I am very
interested in participating. However I am not clear on how to attend or
participate - what is a spotlight - is it a webinar or what format does it
take ?

Joaquin Blaya, PhD Moderator Replied at 4:48 PM, 24 Mar 2016

Hi Elaine,
The Project Spotlight is a new idea we had and it will be the three members who we mentioned who have worked on the Australian Interoperability project giving us their views on this project and its implications on other potential projects. As far as to how it works, it will happen here on this email chain or on the web page of this discussion <http://email.ghdonline.org/c/eJwdUMtugzAQ_Bq4gYwdIDlwiNJGkaK-1EOlXiKDDXZqbMde...> (both will show the exact same thing). Starting this Monday, March 28 and throughout the entire week, all members can send questions or comments.

Patrick Crisp Replied at 12:41 PM, 25 Mar 2016

Thanks for the links and setting up this discussion.

David Grayson Replied at 2:42 PM, 25 Mar 2016

Kia ora and thank you for this initiative. I am interested if Australia is considering the blockchain as a way forward for secure interoperability ?

Attached resource:

Joaquin Blaya, PhD Moderator Replied at 9:43 AM, 28 Mar 2016

Hi everyone,
This officially opens this very exciting project spotlight.

I wanted to start off by asking our panelists, what is the current status of Australian interoperability? I'm thinking things like, what institutions are able to communicate what type of information between them, what have been the great wins of having this interoperability (have they been measured), and so on.



Lauren Wu Replied at 10:01 AM, 28 Mar 2016

Hello - thanks for hosting this great topic. I would like to know if Australia has any formal data sharing agreements in place and if so, what the data sharing agreements cover. I have been looking at data sharing agreements for health data sharing globally to compare and contrast how these legal tools are structured, the content they cover, the parties involved, etc. My general review especially of low- and middle-income countries is that many do not have data agreements in place, or I have not been able to identify these agreements. I would love to know what the Australian model has been.

Vilas Mandlekar Replied at 1:43 PM, 28 Mar 2016

Hello - Lauren's I am also interested in getting information on data
sharing agreements. What are the data formats used in sharing data? Are
there any standards involved for the different types of data - e.g.
surveys, time series, or for GIS data? What are the data ownership and data
disposal restrictions at the termination of an agreement?

The Australian model or any other model would be of great help to me in my


Andy Bond Replied at 6:33 PM, 28 Mar 2016

Hi all,

Australia started on a journey to formally recognise e-health interoperability about 10 years ago. Specifically to distinguish it from technical integration. That is to establish layers of policy, workflow, semantics, and technology to support agreements to work together and equally as important, to support the co -existence of things we don't agree upon. All this in the context that interoperability needs to exist over time and allowing for changes in the layers of interoperability.

We have a range of clinical communities electronically interacting with varying degrees of policy support, behaviour consistency, semantic agreement, and technical integration. Data sharing agreements happen in clinical, public health/reporting, and financial agreements. Do they seem least adhere to consistent models of healthcare? Not really. But we do, I think, have a start in recognising why interoperability is critical to a complex, evolving system working together over time with no single point of control.


Grahame Grieve Replied at 7:52 PM, 28 Mar 2016

"What is the current status of Australian interoperability?"

Mixed. There have been bottom-up and top-down initiatives, which have achieved the kind of outcomes that these usually achieve - a mixed picture where some things work well, and others... don't. And it's rather idiosyncratic what does and doesn't work.

"institutions are able to communicate what type of information between them, what have been the great wins of having this interoperability"

Within hospitals: HL7 v2 is ubiquitously adopted for patient management, and fairly widely for lab result reporting to clinical viewers/repositories. v2 is also used for medications management and to a lesser degree for scheduling. Sometimes for master file management. A few sites use electronic order management, but this is still being bedded down. DICOM is widely used for image sharing

There's a widely supported v2 based delivery system that allows labs and imaging providers to deliver electronic reports directly to any provider. (it works, but it's costly to support for administrative reasons - someone has to install delivery software, and manage the registry that enables delivery).

Between hospitals and their local community, there's a mix of arrangements - some re-use the delivery mechanism above for discharge summaries, but others are still using letters or faxes. Referrals are less likely to be down electronically, and even less likely between GPs and specialists.

In addition, to this, there's a national system, where individuals can opt-in to having a record that they control. Healthcare providers are able to upload health summaries, event summaries, referrals, and prescription and dispense records to an individuals personal record, and then the individual (or their designated agent) can share these record with other healthcare providers. This is a national HIE using XDS+CDA, with a layer of patient control added. This system is entirely new, culturally, so it hasn't had wide spread adoption yet

The great wins out of this? I don't think there really has been any great wins except reduced adminstrative burden for hospitals and GPs. But Tim would be in a better position to comment on 'wins' or otherwise. The system is on the cusp of doing useful things, but there's a lot of detailed reconciliation to go before we can leverage it well.

A/Prof. Terry HANNAN Moderator Replied at 8:13 PM, 28 Mar 2016

Grahame, I deeply appreciate this "readable and understandable" summary of the Australian status on interoperability. I speak for myself but I suspect many people including others are befuddled when experts talk about interoperability. You have cleared the air and this discussion with the other Australian experts will be very educative. Terry

Andy Bond Replied at 8:34 PM, 28 Mar 2016

This is one of the challenges when we talk about interoperability as there is more recognition of the outcomes of interoperability rather than the cause.

Tim Blake Replied at 11:14 PM, 28 Mar 2016

Thanks Grahame.

To build on Grahame's points, I wanted to add a bit about the My Health
Record system (formerly) the Personally Controlled Electronic Health Record
(PCEHR) system. The following comments are a summary of the status quo,
and not a reflection of my personal opinions...

This system is a National summary EHR, which has had mixed results to
date. The system has been in operation since July 2012. We have been
clear that the patient (consumer) owns the data, and is able to control who
sees their data. This position is extremely popular with consumers and
privacy groups, and largely unpopular with clinicians, due to the potential
for patients to "hide" information. In reality, however, only 0.1% of
patients have restricted access to clinical documents within their record.

Just over 10% of the population currently has a record. Approximately
8,000 provider organisations are connected with the ability to send and
view clinical documents. About 50% of public hospitals and health services
are connected, and about 15% of private hospitals and health services.
Around 10,000 clinically curated documents are sent to the system each
week, and this number is growing quickly.

Further, we are about to trial an "opt-out" approach in a couple of
geographic regions, looking at whether this approach works and delivers
value. If it does then we may move to a national opt-out approach within
the next couple of years, which would likely result in over 99% of
Australians choosing to have a record.

Whilst volumes of clinical documents are picking up, the meaningful use of
the system has frankly been poor. This is a longer conversation, however,
and probably for another time, given that this is a discussion about
interoperability. Happy to have this discussion separately.

It is also worth noting that Australia has National Health Identifiers for
patients, health providers and health organisations. This infrastructure,
supported by federal legislation, has been a key component in identifying
patients and underpinning the interoperability of the national EHR system.

The My Health Record system is currently (exclusively) using HL7 CDA
documents as its interoperability standard. We permit CDA levels 1, 2a/b
and 3a/b. We are, however, developing a plan to move towards FHIR (which
will co-exist with CDA for some time). I can discuss this further if
anyone is interested.

One observation that you may find interesting - although we have put a lot
of work into ensuring "structural" data quality through the use of
implementation guides, sample code, CDA validators, etc, it has become
increasingly clear that the "structural" data quality issue is quite
independent from the underlying clinical content quality. For example, you
can have high quality, structured, CDA level 3b, SNOMED coded data that
still represents poor clinical practice and (for example) a
contra-indicated list of medications for a patient.

Consequently, I now strongly believe that the next major step forward in
interoperability lies in medical education and provider incentivisation.
Until we can establish that high-quality clinical documentation is a
clinical task, not an administrative task, all efforts at meaningful
interoperability will fail. I know that most of the people in this
discussion will understand this point, so it's not aimed at you, but rather
those clinicians who think documentation is a task to be delegated to the
most junior team member.

Anyway, with that fuel on the fire I'll sign off for now, but look forward
to picking up on some of the threads I've laid out here.

Steven Wanyee Macharia Replied at 11:55 PM, 28 Mar 2016


I particularly like your comment that "high quality clinical documentation is a clinical task, not an administrative task...". I'd like to extend your comment that high quality clinical documentation needs to be promoted to comprise of mandatory quality of care, and enforced as such. The efficiency of electronic tools (and other factors) to support that is a secondary question. From our routine programmatic M&E and anec dotal observations in Kenya of adaption and use of point of care EMR systems in Kenya, this has been a consistent observation, clinicians relegating this very important task to their junior colleagues, which defeats the purpose of this investment, and additionally, the health system then loses out on a critical value addition of the utility of EMRs (especially CDSS).
What is Australia doing to improve high quality clinical documentation by clinicians?

Grahame Grieve Replied at 12:14 AM, 29 Mar 2016

Steven: In order to do something about it, we'd first have to have the clinicians acknowledge the truth of Tim's statement: "high-quality clinical documentation is a
clinical task, not an administrative task". As far as I can tell, that's not a message that's being understood anywhere. I've been consistently speaking about "clinical interoperability" as a clinical problem rather than an IT problem in the last 12 months, but most clinicians seem to assume it's an IT thing that will transform their clinical experience (while resisting any actual transformation that's not clinically driven)

Andy Bond Replied at 12:26 AM, 29 Mar 2016

Outcomes that are seen as secondary (clinical documentation) to achieving the primary goal (delivery of healthcare) are often difficult to convince others to achieve. When the outcome becomes a bi-product of achieving the primary goal, people tend to take their eye off the inconveniences associated with the secondary outcome. Lot of things to play with around this ... tools, culture, gamification, peer ownership, etc.

Tim Blake Replied at 12:32 AM, 29 Mar 2016


Good observation.

I'd back Grahame's comments and add that there are a couple of things going
on here.

1) We need greater recognition of the problem. Clinical documentation is a
key part of of clinical care. If we are going to have true
interoperability then clinicians need to get serious about data quality.
To this end I have been exploring various strategies for the gamification
of data quality, looking at non-financial incentives to drive this. If we
can make data quality more synonymous with the quality of health service
delivery in the mind of the consumer then I think this is possible. Happy
to discuss these ideas further.

2) It is incumbent upon Health IT professionals and Heath software vendors
to produce systems that make producing high-quality clinical documentation
easy. Systems that structure and code clinical data (as well as performing
clinical decision support to detect data and care that is inconsistent,
wrong or dangerous) without placing a heavy cognitive load on already busy
clinicians are key. And let's be honest, we don't really have that today.
A large number of the clinical systems that I see don't get this, and are
not very good in this regard.

Tim Blake Replied at 12:35 AM, 29 Mar 2016


Thanks for your question.

There are definitely some benefits to blockchain in healthcare, but I think
we're at the peak of the hype cycle on this topic right now. I see a lot
of talk about how blockchain will improve interoperability, but I don't see
how it helps at all with the fundamental issue of the quality of clinical
documentation, as we're discussing in another thread right now.

For me, given my focus on a national eHealth record in Australia, the
benefits to that project are arguably more to do with security and trust.

Pratap Kumar Replied at 12:57 AM, 29 Mar 2016

Dear all,

I totally agree with Tim in that we need better systems to support documentation.

Else there is a risk of making this an 'us and them' problem - it's important to recognise limitations in each cadre but patient data must work for everyone and problems not seen as a 'clinicians' vs 'administrators' vs 'researchers' vs 'IT developers' issues.

Our two bits in this regard in Kenya has been to develop a system that focuses on narrative documentation of clinical cases and we're developing a program to train clinicians on story telling skills!

But also important is that we're asking for such documentation only in cases needing clinical decision support and not document every case electronically.

A manuscript describing our approach just got accepted to the Archives of Disease in Childhood and I'll be more than happy to post the e-pub here when available.

Tim - I'd love to hear more about your 'gamification' efforts. We're also working on incentives to clinicians, for example, using CPD points when cases are used for CDS.


Grahame Grieve Replied at 1:02 AM, 29 Mar 2016

On the subject of blockchain... that's a persistence solution, not an interoperability solution (though it might be part of it). Right now, multiple countries are starting to feel out what it might mean to try sharing data via the patient. The advantages are obvious, but the disadvantages are considerable. People interested in blockchains might also be interested in HEART/UMA, a workgroup that is working on the interoperability side of this, though with a US perspective - and they don't have any engagement.

Tim Blake Replied at 1:28 AM, 29 Mar 2016


Totally agree, as long as this is 'us and them' or Health IT forcing a
solution upon clinicians then it won't work. That said, there is no
shortcut to clinicians taking responsibility for the quality of clinical
documentation. Some currently do, but many don't.

Regarding gamification, it's just thoughts and ideas at this stage, not
something that will imminently be tried...

These thoughts are based on an underlying assumption that applies to a lot
of the work that I've been doing, that "with the wave of coming digital
disruption, patients / consumers are / will be a more effective lever for
clinical behavioural change than Governments".

This assumption is not yet widely shared. However, experience in Australia
shows that whenever we have wanted clinical behavioural change, we have
always used (poorly managed) financial incentives to try and drive that
change. As a result there is no emotional engagement with the outcomes and
little change outside ticking the box. Consequently, the meaningful use of
our My Health Record (and many other systems) has not (yet) been good.

Fundamentally this is because we are pushing a poorly governed extrinsic
motivator. My hypothesis (from my experience as a Health CIO and Strategic
Advisor to Health systems for many years) is that we need to find a better,
intrinsic motivators that are 'closer to home'.

Putting these ideas together, we should work to educate patients /
consumers that the quality of their care is tightly associated with the
quality of documentation that their health services produce. Therefore, we
should reward the highest quality (and most improved) clinical
documentation contributions to the system (the My Health Record system in
this case). This reward could take several forms, but could be a 'quality
mark' that health service providers are then permitted to use on their
branding. This becomes an intrinsic motivator as health services compete
to show pride in their brand.

In an increasingly competitive, global digital health market, this
(non-financial) incentive could drive behavioural change, and ultimately
improve interoperability.


P.S. This is just one example of gamification at work in this domain...
I'm sure that others can think of even better ones.

Lauren Wu Replied at 10:30 AM, 29 Mar 2016

Hi Andy - thanks for your reply regarding data sharing agreements. Are there any agreements available in the public domain you could share as exemplars?

Joaquin Blaya, PhD Moderator Replied at 10:50 AM, 29 Mar 2016

This has been extremely interesting to read, thanks for all of the participants.

Tim, you mentioned a couple of things that stuck out in my mind.
The first was the national identifier for patients, providers and health organizations. Most of the time, only unique patient identifiers are mentioned. I’m assuming that having all three identified was essential in moving forward with interoperability in Australia, but do you see a case where a national system could work with only patient identifiers in place?

The second comment was the low clinical quality of the documentation. Are there any plans (or is it even feasible) to do quality metrics at the national level to feedback to providers and health organizations? i.e. do quality control at the national level. Or do you think this would be an even higher disincentive for clinicians and health organizations to submit their data?

Patrick Crisp Replied at 2:03 PM, 29 Mar 2016

To the panel:

In the My Health Record system, who is responsible for the quality of the record? Does the individual have to check that each encounter/document/test etc. has been uploaded to the record? Is there a legal requirement for providers to add encounters/documents?

I am just concerned that if you have responsibility shared between multiple providers in different organisations then you are not going to have an up-to-date, complete record that you can trust.


Grahame Grieve Replied at 2:40 PM, 29 Mar 2016

A national system can work with out unique identifiers for providers and institutions. It just won't work as well; there will be degeneracy all through the system; just like if there's no single patient identifier.

With regard to trusting record: we don't have a pathway forward to having a record of trust for either the patient or their healthcare providers. I think that the most useful thing about the system we've built is that it will spur that discussion: who trusts who, and why? My personal view is that the existing models of trust are broken in the healthcare system, and that real change will only be possible when everyone confronts this. But it's much easier to have other less significant discussions instead.


Tim Blake Replied at 7:04 PM, 29 Mar 2016


That's a good question, but one that currently doesn't have a complete
answer. The Federal Dept of Health (the operator of the My Health Record)
and their partner organisation NEHTA (National eHealth Transition
Authority) are responsible for structural data quality and compliance, as
well as clinical safety.

We're still feeling our way on on issues of the quality of the clinical
content. Some of this falls under guidelines published by our Australia
Commission for Safety and Quality in Health Care, but arguably more
responsibility needs to be taken by the originating health systems.

We're not yet in a place where all health encounters generate a clinical
document, as many health providers are only now rolling out the capability
to connect (3 years after the system started). Generally speaking, if a
consumer has a record (and remember only 10% currently do) this is taken as
standing consent to send documentation, if the provider is able. However,
consumers can elect in a particular instance not to send, if that is their
wish (or they can always change their mind and remove or hide content later
on - another huge area of debate).

There are no legal requirements for health providers to upload documents,
as this would be too heavy handed. This area is the most controversial and
politically charged part of the My Health Record. There are however,
incentives that increasingly require the contribution of content to the
system to be attained.

However, particularly in primary care where GPs are (currently) paid on a
fee for service basis, most are unwilling to do extra work for the same
money. I completely understand this view, but it points to the heart of
the problem - 1) that clinical documentation is not seen as a key part of
their role and 2) that fee for service is no longer a fit for purpose model
(for many reasons, especially when considering eHealth / digital health).

Your last point is entirely correct. CDA documents do not generate an
up-to-date, complete record, nor do they easily permit solving this
problem, due to the variety of difference conformance levels, and reliance
upon non-machine processable narrative, etc.. This is most problematic in
the area of trying to build a current medications list. Hence why we are
looking at transitioning to the use of FHIR.

Hope that helps.

Tim Blake Replied at 7:18 PM, 29 Mar 2016


Thanks for your questions. Not much to add to Grahame's earlier response
on the first one, other than to reinforce that identifiers for health
providers and organisations will ultimately be extremely important for a
variety of reasons, not least clinical safety, audit, security, etc.

The following views are my own, not those of Government...

I personally believe that we need stronger (automated) data quality metrics
that support our understanding of the data flowing in to our system. This
is relatively easy to do for the structural quality of the data (i.e. CDA
conformance level, etc), but hard to do for clinical content quality.

We do currently do periodic external clinical data / safety reviews of
de-identified data, but I believe that we should automate some of the work
that gets done in these audits into some kind of operational dashboard.

Any use of data quality metrics would be aimed at incentivising providers
(i.e. a carrot) and not punishing them (i.e. a stick). We know that
achieving high quality data is hard, and we're looking for mechanisms to
lever and encourage improvement, not discourage people who are at an early
stage in their maturity journey.

Hope that helps.

A/Prof. Terry HANNAN Moderator Replied at 7:59 PM, 29 Mar 2016

All, our discussions on GHDonline have been valuable and of these some stand out as exceptional as does this one on interoperability. Previously we have created summaries of the input and queries to these discussions. For this one I would like to make a request of Tim, Andy and Grahame to create a summary document of the thoughts expressed here that include;
1. Their ideas and perceptions
2. The ideas and queries from the different contributors
Over the last week I have been tempted to post a reply but the quality of the postings have been so good that any comments from me would be redundant and all my queries have been answered-well done.

Grahame Grieve Replied at 8:24 PM, 29 Mar 2016

I want to comment on one really important part of interoperability which hasn't been noticed as part of this discussion. I said:

"There's a widely supported v2 based delivery system that allows labs and imaging providers to deliver electronic reports directly to any provider."

It's not every country that has achieved this - in fact, few have. It's grown over time, and it's one that the healthcare practitioners really like - except that now they take it for granted. But it does cost - the sender of the message pays somewhere between $0.01c and $1.00 to send the message (depends on business arrangements and message size - and my $$ costs may be a little out of date). That doesn't sound like much, but when your basic margin is <$1.00, and you operate at scale, this is real money.

There's several different providers that mediate the delivery as a cloud service. They compete with each other on cost and service and coverage - each of them have different practitioners signed up (some/many signed up to multiple providers). Now they can exchange content between themselves, but my impression is that this doesn't happen much yet.

Technically, the sender drops an HL7 v2 ORU message into a local diretory, from where the cloud agent picks up the message, determines from the message header where it's to go, and drops it into a local directory there, from where their system picks it up and puts it in the 'to review queue'. The format of the ORU message is dictated by an Australian standard (there's several).

However the format is still ambiguous, and/or there's technical mistakes on the part of many implementers, so one feature of this arrangement is that the cloud messaging providers maintain large remediation tables that says what fiddles they have to make to a message for a combination of sender and receiver. There was some effort to create a process to iron this out a few years ago, but since the costs of that process would fall on the source/target vendors instead of accruing as income (not profit) to the cloud vendors at the cost of the senders, it never really made any progress. (I often wonder whether the senders understand how the economics of this arrangement are working out for them)

So, the system requires effort to setup and maintain, and to ensure reliable delivery - but everyone likes it a lot. Except for the small number of providers who are not hooked up to it, and reliably can't find imaging and lab reports on their patients (per personal complaints from my friends)

Grahame Grieve Replied at 6:25 AM, 30 Mar 2016

More about block chain: http://geekdoctor.blogspot.com.au/2016/03/evaluating-blockchain-for-health.html - I pretty much agree with these comments, particularly with regard to Arien's note about network and business model.

Charlotte Hespe Replied at 6:13 PM, 30 Mar 2016

I thought it would be timely to add my comments about the MyHealthRecord system as a GP user!
I have been actively involved in the process of getting GPs to engage with the system since the launch of the PCEHR.
I have no IT expertise but understand a lot more than the average user about how the system works and it's limitations.
When the system was first launched I was part of a Quality Improvement collaborative looking at getting General Practices to improve their data in the medical records to ensure that the uploaded health summaries were accurate and reflected what the patient and Dr saw as a complete and appropriate health summary (remember that the information that GPs can upload is ONLY a summary document - it contains demographic details, allergies, medication list, past and present medical history (divided into surgical and medical) and immunisation records).
It is important to note that the collaborative started before the PCEHR was actually functioning so we had to practice with other tools that were available to see what the health summaries looked like and patient satisfaction with the documents.
There were 60 practices from around Australia that participated and we worked over 12 months to look at how we could generate accurate health records.
The data accuracy was much harder than a lot of people had thought and I would like to preface this with the statement that this is NOT because clinicians don't value accurate data but it is due to competing agendas when working out in the real clinical world!!!
Clinicians have not been taught to prioritise data input and in fact have been taught that this is important but someone else can (and should) do it!!- as a GP I am always being told that the issue of incorrect data in the hospital system is due to "clerical" errors- never is it seen as the clinicians job. Medical students are taught to be "clinicians" and are still not taught about the value that data adds to their ability to be a diagnostician or how to manage the problems in front of them!

So, what it boils down to is that the current medical record systems are all only as good as the data that gets put into them and currently there is no system designed to assist clinicians in being able to input accurate data at the same time as provide essential clinical care.
Hence, we need to be careful not to criticise clinicians for not prioritising data but encourage them to look at ways that will assist accurate data input in a timely manner.
In our practice we spend a LOT of time teaching and encouraging all team members to check the medical records for data accuracy. We check our entire patient population data each month and target data gaps with the clinicians.
This has resulted in a much wider understanding and ownership by all the Drs regarding accurate data!!
We also can enrol our patients into the MyHealthRecord system during consultations or at the front desk.
As a Practice, we are currently targeting our >75's to all be enrolled and have a current health summary uploaded into the system.
When we have achieved this, we will work on ensuring all patients with chronic diseases are similarly enrolled and uploaded.

There is currently no financial incentive for my practice to do this work. It is due to our desire to be a high quality patient centred healthcare home that drives this work.
From May we will get an incentive from the Govt to upload 6 medical records per month which will at least assist the administrative costs.

So in summary, I agree with what TIm and Grahame have been saying but want to ensure that the clinician perspective is not lost in this discussion and a real understanding of what is needed to create quality and useful health summaries in our current system.

My dream is a truly integrated medical record that is able to be shared across health systems.:)


Charlotte Hespe

Joaquin Blaya, PhD Moderator Replied at 6:24 PM, 30 Mar 2016

So, if there was a country that wanted to go down the interoperability route, what could they learn from Australia, both in terms of Lessons and what they could take and replicate in documentation, planning strategies or software?

Grahame Grieve Replied at 6:50 PM, 30 Mar 2016

Charlotte, thanks. I think the really difficult thing is that good documentation is about the future. Not having it represents increased technical debt, and it seems that we're all so busy drowning in technical debt that we don't have time to invest in the future.


I observe that Australia has very poor governance with regard to healthcare provision. There's multiple centers of authority over different aspects, pursuing different priorities. with overlapping jurisdiction. That makes change hard - that has both good and bad aspects. I think that in such an environment, ground up initiatives are more likely to penetrate the system, but they need to paired with a regulatory framework that wants to leverage them, rather than supercede them.

At a technical level, I know that HL7 v2 and DICOM are good for well controlled closed networks, but that model has had it's day. CDA is good for documents, but they have only a limited place in a well connected healthcare network. That's why I started working on FHIR. It's success has not caught us by surprise: it has the capability to allow systems to connect and to serve the future that people hunger for. But the pace of it's success has - we deliberate set out to create an iterative process to build something with real quality, which takes time. But people are using it now.

But FHIR is only an enabler. The technology is all nearly there, but without clinical governance - the kind Charlotte talked about - all it will be is technology. The challenge of interoperability has moved from being a technical one to a clinical one: just what is required to allow us to have seamless processes when a distributed team (both in terms of geography and governance) provide care for a patient?

A/Prof. Terry HANNAN Moderator Replied at 6:57 PM, 30 Mar 2016

Grahame, if there is only one major point to come out of this discussion (and it is not the only major point) is that embodied in your statement. "The challenge of interoperability has moved from being a technical one to a clinical one".
It strongly echoes the main factor in Mamlin and Biodich's paper. That is the system must be clinically useful.

LOW COST: preferably free/open source
CLINICALLY USEFUL: feedback to providers and caregivers is critical. If the system is NOT CLINICALLY USEFUL it will not be used.
Mamlin BW, Biondich PG. AMPATH Medical Record System (AMRS): collaborating toward an EMR for developing countries. AMIA Annu Symp Proc. 2005:490-4. Epub 2006/06/17.

A/Prof. Terry HANNAN Moderator Replied at 11:10 PM, 30 Mar 2016

Graham, Andy and Tim, I read the 'blockchain' posting and how it may be of benefit to health. My limited knowledge seemed to perceive that 'technically' it would not add much to what we already know. What seems to be the stumbling block is the end-user, data capture component of our eHealth systems. Please comment because I suspect I am missing an important point(s) here. Terry

Andy Bond Replied at 11:54 PM, 30 Mar 2016

While blockchain has some interesting qualities, I don't see them being a fundamental enabler of what most people are trying to achieve from e-health interoperability. I've no doubt their ability to support a shared, trusted journal of content has value to e-health but I still come back to my premise that we still struggle to understand what a good interoperable system looks like (other than two random systems that can ubiquitously work together across all aspects of technical connectivity, shared semantic understanding, common work practices, and consistent policy frameworks). So for me the starting point for trying to achieve interoperability is having a mature conversation about expectations and how success will be measured. Do you bind it all into a single dimension and just call it connectivity or do you recognise the layers of agreement that need to work together to enable systems to work together. What degree of difference do you allow? Can you work with alternative technical solutions but meeting common business and semantic models? I would suggest that agreeing on working with differences is as import and working with things that are the same. This is the foundation for an interoperability framework. A description of how evolving systems work towards contracts of working together through changing collaborators, technologies, semantics, and work practices. Interoperability is best seen at scale across communities with multiple points of control, complex/emergent behaviour, and leverages building blocks like common standards, common terminologies, policy frameworks supporting innovation, etc.

Elaine Baker Replied at 12:57 AM, 31 Mar 2016

I would like to ask the panelists and other Australia experts: if you were starting a shared health record system in a new country, what successes of the Australia shared record would you build on, and what shortcomings of the Australia shared record system would you avoid? Both technical and change management issues.

Andy Bond Replied at 1:12 AM, 31 Mar 2016

Who will be the record custodian? A clinician, many clinicians, consumers, or a mix of clinician/consumer.

Will the record be based on clinical documents, structured clinical content, a mix of documents and structure, or maybe an audit of clinical information sharing?

Grahame Grieve Replied at 1:16 AM, 31 Mar 2016

Well, I'd start with the key question: how much agreement do you have amongst clinical providers around what kind of data sharing agrements (both in regard to the nature of the data, and the accountability processes around the sharing). If you have good agreement (though it's hard to tell, because early on, no one wants to be nay-sayer, so they just shut up), then you can set up data sharing agreements, and use FHIR. If you don't, then all you can build is a pile of documents that everyone will be suspicious of, using CDA/XDS, like we did.

From my point of view, we have some successes: a national patient identifier, a national certificate system. Those are good (could be better, but that's always the case). But we built the base in too much of a hurry. Hard decisions had to be made to meet political imperatives with regard to timeline. We couldn't bring the community along with us (my 1st rule of interoperability: it's all about the people). And now we've cemented mistakes in place, and it's too hard to change. We can't even acknowledge that what we built is a beta, so we can actually learn from our mistakes.

Take the time to build the community slowly. It works. Perhaps Tim or Andy can talk about the NT system, because I think that's a success by anyone's measure.

A/Prof. Terry HANNAN Moderator Replied at 1:58 AM, 31 Mar 2016

I think these answers are appropriate and highlight the limited successes in Australia. As a clinical informatician I believe the natioanl identifier is important but it is not essential for a national EMR as there have been systems around the world that have multiple identifiers to assist with patient identification accuracy where the population is migratory. What the natioanl identifier provides is the "key" to access the clinical information within the EMR. NZ and Chile and others have had natioanl identifiers for decades but they do not as yet have a good or complete natioanl EMR.
The successes that I have seen, Hopkins, Regenstrief, BID, HELP and others and OpenMRS and other open source systems are built using known and evolving clinical information management tools where the direct patient care problems are addressed and solved using standardised clinical data models and concepts that can inter operate with other systems with the same or different infrastructure models. The process must be incremental and involve the users at all stages with weekly or fortnightly reviews of implementation evolution.
From first hand experience I have seen and been involved in the successes of systems such as the John Hopkins OCIS system in Australia (1984-1992), OCIS in Kenya (2000-) and more recently the EMR in the Universidad Catolica EMR in Santiago Chile.
I hope this makes sense.

Andy Bond Replied at 2:24 AM, 31 Mar 2016

As Grahame mentioned, the EHR in the Northern Territory (NT) is an interesting example of a successful EHR project. It comes with a set of constraints that have helped make it workable. The NT has a relatively small population of 250,000 with a large and mobile indigenous community who have been the core of EHR participation. The NT Govt plays a strong role in the provision of healthcare and management of healthcare contracts, thus the issues of cross-vendor interoperability are of a different nature than one would find in a larger and more diverse market. Healthcare contracts often drive the chosen technical foundation. Initial content integration was not atomic but more driven by structured narrative. I'd characterise their success as that of a very aligned approach driven by consolidated governance and investment.

Niamh Darcy Replied at 9:53 AM, 31 Mar 2016

Totally agree on the national identifier - even with having a national ID there is still a need for multiple ways to actually make sure this is the right person - even when the population is mainly stationary but then for example pregnant women often go home to where they grew up to deliver. When we worked on ZEPRS in Zambia, we used multiple IDs, as even though people were supposed to have an ID they did not necessarily have it with them (in Lusaka district), so we had I think 3 IDs as well as a lot of demographic information to link these together to a national ID. But this "national or sub-national ID" can be the key to access clinical information in multiple EMRs/Point of Care systems.

Steven Wanyee Macharia Replied at 12:01 AM, 1 Apr 2016

We did some work in Kenta as a proof of concept on data exchange and unique patient a couple of years ago. You can find some info here, https://sites.google.com/site/oeckenya/. I'll be happy to discuss this more.
The results of this project, our experience and lessons learnt provided a very important springboard for phase II of that project now called the National Unique Persons/Patient Identifier (NUPI) project. I'll be happy to discuss this some more too.

Steven Wanyee Macharia Replied at 12:12 AM, 1 Apr 2016


I'd really love to hear more about your efforts on "To this end I have been exploring various strategies for the gamification
of data quality, looking at non-financial incentives to drive this. If we
can make data quality more synonymous with the quality of health service
delivery in the mind of the consumer then I think this is possible. Happy
to discuss these ideas further".

Pratap, I'm curious to hear more about your efforts too around improving clinical documentation practices. Technologically, for example in OpenMRS, there are some pretty cool functionalities that have been developed along open notes concepts (with underlying machine learning functionality) to support such tasks with reduced cognitive burden on clinicians, but it isn't clear yet how that is working.


Amal Bholah Moderator Emeritus Replied at 3:00 AM, 1 Apr 2016

This has been a really interesting discussion. The NEHTA model has been used a lot in my assignments in Global eHealth. Is NEHTA ready to partner with other emerging countries to transfer knowledge and expertise? If so, how to proceed? Any contact point or any policy developed for that?


A/Prof. Terry HANNAN Moderator Replied at 3:11 AM, 1 Apr 2016

Amal, I make this statement in the hope that Grahame, Tim and Adam will correct or affirm my impressions. I am not sure the NEHTA model is a good model. It is a model that has 'stumbled' along with some positives (our Spotlight members are probably the best part) but there are other sections projects or implementations that have not been positive or maybe even negative. I am not sure if this is because the model is wrong or the implantation plans and support have gone awry. It is complex and I am not aiming to 'bag' the NEHTA model but we are a long way behind where we should be in Australia. I think the inputs from our panel are pointing out the issues. Over to our Spotlight panel.

A/Prof. Terry HANNAN Moderator Replied at 3:12 AM, 1 Apr 2016

Oops I meant to type ANDY...........apologies........its been a difficult week

Andy Bond Replied at 6:39 PM, 1 Apr 2016


It has certainly been an interesting journey for e-health in Australia that goes well beyond the more recent incantation in NEHTA. Its hard for me to put a finger on the exact NEHTA model and there has definitely been some stumbling along the way. I doubt anyone would look to repeat the same path if they were to start again but there is also a lot of learning along the way that I hope will enlighten us for better investments going forward. The hardest step for many is admitting to something that isn't working well and using it as a lesson for doing something better. There are so many factors that brought about the NEHTA approach, probably the majority driven by the environment in which the problem was being solved rather than any technical e-health problem. Government investment approaches, the conflicting interests of stakeholder groups, significant changes in technology, the rise of the consumer, etc. Someone once said to me that NEHTA's value was as a national mirror that if nothing else managed to reflect on a national scale the challenges that we face and the implications of solution approaches. Some things I hope can be done better going forward include a recognition that this is not a problem of complicated scale but one of emergent complexity where the many micro-communities are inextricably linked through technology, payment models, care pathways, politics, and so much more. Many of the challenges don't have an off the shelf solution but one day hopefully they might. We need to acknowledge the co-existence of things that work today, things that we think we know how to invest in for tomorrow, and things that need some clever minds to try novel solutions, potentially fail, but give us insight into ways that we can move those problem into quantifiable investments. We need to acknowledge that this is a complex space and we're not building something like a bridge that will last for 50 years. Our responsibility is to foster an environment in which evolution and innovation are core design principles so that we can harness all of the national assets, both public and private, that contribute to moving e-health forward rather than trying to invent prospective solutions to problems that we may not really understand well. An interoperable solution without a sound business model, in all senses of the word, is not a sound model for moving forward.

Joaquin Blaya, PhD Moderator Replied at 7:05 PM, 1 Apr 2016

We wanted to officially end this Project Spotlight, which has been extremely interesting. We will be writing a Discussion Brief for this and posting it on GHDonline.

We wanted to thank the members who took part in the discussion, and especially our 3 guests, Andy, Grahame, and Tim, they have done an amazing job of keeping up with a lot of questions and comments. We appreciate your time.

Hope everyone has a good weekend,

Tim Blake Replied at 4:12 AM, 2 Apr 2016

Thanks Joaquin, it has been a pleasure!

Happy to continue individual discussions if anyone wants to follow up on
anything that was raised.

Noor Afif Mahmudah Replied at 3:47 AM, 8 Apr 2016

good idea

Joaquin Blaya, PhD Moderator Replied at 1:05 PM, 19 Jul 2016

Hi Everyone,
We wanted to let you know that the Discussion Brief for this Project Spotlight is now available at http://www.ghdonline.org/tech/discussion/project-spotlight-interoperability-i...

It summarizes the points of the discussion and provides additional references.

Attached resource:

Rodrigo Cargua Rivadeneira Replied at 1:18 PM, 19 Jul 2016

Que buen Material Joaquin Gracias

Rodrigo Cargua
"No existe falta de tiempo,
existe falta de interes. Porque
cuando la gente realmente quiere, la
madrugada se vuelve dia. Martes
se vuelve sabado y un momento se vuelve

This Community is Archived.

This community is no longer active as of December 2018. Thanks to those who posted here and made this information available to others visiting the site.