26 May 2019

Bettina McMahon in conversation with Wild Health on interoperability

Insights Interoperability


WH: Where does the agency see secure messaging in the context of the overall pitch of what interoperability means for the Australian digital health ecosystem?

BM: When Tim [Kelsey] started in his role in August 2016, and I started in my role, he began by listening. He spoke to hundreds of people in the country, and one thing that came through consistently, especially from GPs but also specialists, hospitals, others in allied health, was that “this all sounds fantastic, the digital health vision you’re pulling together – but it’s really frustrating for us that we have to rely on our letters and fax machines for secure messaging. It’s just not working as well as we need it to – and if you could just fix that one thing that would make our lives a whole lot easier.”

And because it kept coming up, as a theme consistently across different sectors, we set up a program that year – in September 2016 – to try and unpick what were the barriers to secure messaging working at a national scale and in an interoperable way.

The first two years was largely technical and now we’re looking to transition that so it’s not just technical.

The background is that we sort of jumped ahead of the national digital health strategy which we didn’t complete until 6 months after we started the secure messaging program. We and started working on secure messaging [early] because it was such a priority for healthcare providers.

In the meantime we developed a digital health strategy, and  interoperability was a big theme.

We’ve got multiple systems available for providers to choose from in Australia. In other regions a government might say ‘this is the one that everyone has to use’. Which brings things down to interoperability and standards being adopted and used consistently and making that work.

When we first started this work one of the things that we did was listen to the ideas that the industry has, because they largely knew a lot of the answers. But they needed some national co-ordination and a bit of authority applied to a system.

There were some things that we did in the program that went in a different direction to where we thought they would’ve gone because of the input from industry largely, and their commitment and willingness to participate. Things like helping us to develop the different FHIR APIs to allow for different directory searches. It ended up going down a slightly different course, because the industry, to their credit, gave us the benefit of the doubt that we really wanted to make this work. We’re in a lot better place now than where we were a few years ago in terms of industry relationships.

WH: There still seems to be a lot of commercial underlay to the progress, or not,  of secure messaging. For instance patient management systems are a bit all over the place in how they accept and process messages, and the messaging vendors are very frustrated. Medical Objects, which has about 25% of the national market has walked away for now from the ADHA programs. Where do you think you are now given where you were when you started two years ago? What  are the good parts of where you are, and what are the obstacles now? And what do you have to do next? 

 BM: We’ve got messages moving across different messaging vendors.  That didn’t happen before. Through the development of standards we can now have messages moving across different platforms, which wasn’t the case two years ago. Another important step is we have agreed to formats that each of the clinical information systems need to be able to display. And it’s not one format only. We haven’t picked one and said “right it’s just got to bed PDF” . The clinical information systems will soon be able to build forms so that a message can be received in a multitude of formats.

WH: Which formats?


 WH: So not HL7v2?

BM: No, the message formats can’t include HL7, but the actual payload can: there’s two parts to it – there’s the message itself, so that’s like the envelope, which is the message format, and then within that there’s a payload. HL7 we do recognise. Basically all the pathology requests are sent with HL7. I think the problem is where we got to a few years ago that everybody was saying there just had to be one [format] , at the cost of all the others. That’s not the approach we’ve taken. We’ve just said we need to recognise there’s a few standards in use at the moment and to move forward we actually have to have standards that can accommodate all of that.

Some of the vendors were begging us to make a call and say this one format is it – and everything else just has to stop. There was an expectation that we’d pick one. That we didn’t go that way is why we kept people engaged I think. If we went the other way we may have  ended up with everyone splintering again.

The other thing that we’ve achieved is the idea of a federated provider directory. One of the other barriers to secure messaging is getting the pinpoint location.

A lot of the messaging vendors maintained their own directory- but if someone was using a different messaging vendor then their point of location wouldn’t be in the one directory or the other. Health Direct set up a directory service that we tried to test in Queensland but it needs a fair bit of redevelopment before it will be fit for purpose.

And then industry suggested we have a federated directory service so we developed APIs and all of the standards so vendor directories could talk to each other. We have  implemented this in the proof of concept trials and we’re in discussions with AHPRA about how we can connect that to their digital registries to provide an extra dimension of integrity. But that too is a longer term piece of work.

WH: Why is AHPRA involved – are they going to be a central source of truth? 

BM: We want to stick with the federated architecture. We think that is the right approach but there is a challenge where – by using a federated directory we might get duplicates, so you might get a health care provider that exists in multiple directories.  We think there is a place in the architecture for AHPRA to help with this sort of verification.

WH: So what happens – you get a conflict and you get two email addresses, and you call APRA and you say which is it?

BM: At the moment the way it works in practice  is if you’re, for example, using Best Practice , you’ll usually search your local directory first for performance reasons. Typically a practice will be referring to a local directory.  If you can’t find the end point location in that directory, then you can go to the global search which uses FHIR APIs and will then return you another result. It works in a similar way to an internet search where you get different results for users basically have to choose which seems most relevant.

WH: Is that something AHPRA collects, or is it something they could collect? No one collects that centrally do they?

BM: I’m not sure the messaging vendors collect what clinical information system their using. Some people are saying ‘vendors have built up this repository of providers in the directory’ and that’s one of their IP selling points. We had a lot of discussions about this and we all came together in June last year, and said  for a federated directory to work we need to agree that there is some base level data that shouldn’t be considered an asset for one system or another, but a national asset for everyone.

And then there will be value add that the vendor directory can add on top of that , which of course they can commercialise. So there’s a core data set which will be the name and the end point location, just some really minimal data that everyone agrees to share, and then if some vendors want to collect other data then that could be on top of that.  We agreed at that meeting that there would be a basic set of data that we would share, because it was agreed by everybody that by sharing that we will actually increase the use of secure messaging, and that’s going to be a good thing for all the industry.

WH: Some of them think their data is unique though and it’s a big advantage for them over other vendors, so will they give up enough core data in the end?

BM: It’s a good point you raise, because I think in the past – more so than now- there was a view amongst some that holding market share meant not interoperating. But I believe that in this space secure messaging people have moved on, partly because they’ve seen that we have to move on.  That the demand from healthcare providers is such that they’re not going to tolerate it anymore.

Most of the people working in this – they really believe in what they’re doing. They have a vision as to how they can assist healthcare to become more efficient and safer, so we’ve really been able to draw that out and show that there’s common ground. There’s still room for different companies to innovate and make money out of this but it shouldn’t be at the cost of interoperability.

Everyone in the space has agreed that making this work is the right thing to do and it also supports business innovation and profitability . I think people have moved beyond that idea that might’ve been around a few years ago – that it makes commercial sense to lock in customers. I haven’t really heard people saying that much in the last year or so.

WH: The saying goes, technology generally is an enabler, not a commercial advantage (at least not for long). From a commercial viewpoint, do you feel the secure messaging vendors recognise that where technology like FHIR and open APIs is heading and so they are actively trying to evolve elements of commercially sustainable advantage in new ways now?

BM: I think the realisation of this is what’s brought a lot of industry to the table in the end. They’re seeing that there are other areas they can add value into the future and that will be the major determinant of if they’re successful . For example, looking at some of the logic they can add into referrals. Some of the messaging is part of it, but it’s not just about the messaging, it’s also about transfer repair, hand over of patients between different parts of the health sectors, and other information that improves this process.

There’s others looking at consumers and the expectations of consumers- how they communicate with their healthcare providers. There’s a lot of talk about telehealth and broadening that out. Some of the messaging vendors are looking at how they can involve consumers more in meaningful communication. Health hasn’t been as disrupted as an industry as others by technology but they’re waiting for that to happen.  I think these guys are really expecting this to start and to be at the front of that looking at new business models like looking at what the expectations of our consumers are and how providers might meet them better.

For all this to happen though there has to be a base level of infrastructure. On top of that is the value that people will pay for.

WH: Why did Medical Objects walk away from the secure messaging program recently? As such a big component of the overall national system how much might that slow things down?

I don’t think they walked away. I think they worked hard on one of the proof of concepts, and then withdrew from that for various reasons. A couple of vendors walked away from the proof of concept, some of them because they weren’t compensated enough for the amount of effort they were putting it. But I think they are still really interested in making this work.

I do feel that we’ve reached a point of completion on some things. Where we’ve set out to do some stuff, we’ve done that stuff, and so that’s a milestone. But it’s like we’re now on the next horizon, and we’re looking at the next thing.

I think this next thing is national scaling. We need to look at it and go ‘how do we move beyond the technical demonstration [pilots] that the vendors have developed and which ones will work’.

This of course means there’s a lot more work to do. But we’ve started on that with the call to CIS vendors to start building their systems out to the standards developed.

We’ve had over 70 expressions of interest put forward, so we’re assessing them now for eligibility. We want to look at value for money, and how far we can get. But this initiative showed a real interest across the industry, to basically implement the standards we developed in the proof of concepts stage.

WH: What does the criteria of  ‘value for money’ mean?

 BM: For interoperability to work at a national scale we need most systems using the standards to get to the tipping point, a network effect. That’s why we put the industry offer out. We’re looking to do as much as possible. We have a budget, but we’ve had such a great response we might look at trying to stretch that a bit more.

WH: Do you think such a great response means that people are understanding what you are trying to do, or is it maybe the offer of money in a relatively tight market?

BM: The amount we are giving out won’t really cover the development cost. We think it provide a reasonable capital injection for those companies who want to get on board, who need a bit of assistance. It certainly won’t be a profit-making exercise for those vendors who are involved.

Ultimately health care provider clinical information systems need to stay focussed on usability and functionality. We are looking at what we can do to promote the need for interoperability to promote the value of interoperability to healthcare providers. If we do this providers will take into account interoperability when considering the purchasing of a messaging service, or a clinical information system. They’ll start to think about how well it will connect up to the rest of the healthcare system, and how easily they’ll be able to send secure messages. Our job as an agency is to make sure that when they’re making those decisions they put a lot of value on interoperability functionality.

WH: What about pathology labs in all of this? They do most of the messaging as things stand today. Are they going to change to help all this?

BM: We work closely with the pathology sector. I think less that they’re reluctant and more that they get a lot of demand from their customers for different modifications to messages. They’ve had to make a lot of modifications and a lot of them are not standard. They have been responding to their customers which is a pretty normal thing to do. So if people see them as reluctant to join in, we should consider that they’re thinking about the effort that’s required to transition their systems, especially in the absence of standardisation.  Also, it’s obviously a very competitive industry and they’ve each made major investments in their messaging. Remember also that these guys already have a consumer portal for consumers to look at their own results. They’re already moving ahead in those ways, and they’re really interested in looking at a standardised way to do this. There’s a lot of investment over time, and you’re not going to get a lot of organisations dumping that investment to go with something new without a pragmatic road map. That’s what we’re looking to develop.

WH: It’s a great vision but a cynic might say that they’re happy to take their time, because they’ve got their customers locked in downstream. Some vendors will tell you also that they don’t have much expertise left to pull this off now and not much interest either given other priorities. If it’s 40% of current messaging are they likely to meet any new standardised regime within the time period of all the others?

BM: I think the pathology orders and tests are a little different. They are absolutely secure messages – but they are a little different to the e-referrals, discharge summaries and the handover of care that is taking up a lot of the other messaging.  In their defence, they are big organisations, but they don’t operate on huge profit margins and they’ve got to make decisions on what they invest in.  A lot of them have technology programs, and rolling two year plans and like most businesses they have to make decisions on what they focus on in these strategies.  I look at it kind of practically going in, expecting that we’re facing standards that not everyone will comply all together. Australia just doesn’t work that way.

They are connecting to the My Health Record (MHR) in all sorts of ways now. So they’re playing ball in the national infrastructure here and they’re looking to connect and looking to share . But they’ve got to plan it out and make their own decisions.

WH: Is it because it’s more complex than connecting to the My Health Record?

BM: The key difference to sending a report to the My Health Record, and sending or ordering pathology results is the customisation. We’ve got specifications for the test and off it goes type of thing, whereas there is a lot of customisation that those ordering physicians require. This adds a lot of complexity from point to point.

WH: Mostly GP’s get their messaging for free… how do you get through that? Won’t someone have to pay in the end to get this all connected up eventually?

BM: There are costs but I guess this is the case we need to sell.  When we’re looking at a change management strategy, which we’re doing at the moment, we know there’s actual costs in relation to the fax machine and posting. But there’s more than that. It’s the quality cost. The time and the labour.  It’s scanning and things going to the wrong place and the need to have to follow up when things go wrong. What are the consequences of that?

WH: Who do you target when you’re trying to give that message, because everyone says ‘yes we want it’ – but when you go to GPs they don’t really have the time or the money to care about secure messaging that much. When it’s explained to them they get it, but your average GP is too busy trying to stay afloat to get on board properly don’ t you think?

BM: One of the dimensions of the problem is the change management. Even if you get all the technology working so it’s all interoperable someone has to make a decision to change it in the workplace.  In a small business or in a large organisation, what is the trigger for that?

We’re working hard on the strategy for this at the moment. I think where we’ll end up is it won’t be one thing. It will be a number of different things together, that has the vendors promoting that if the practice upgrades their software they can get these additional features, which include interoperable secure messages.

It could be from the local hospital saying that from a certain date they’ll be sending discharge summaries by secure messaging rather than by fax machines. It could be the Commission of Quality Healthcare coming out and saying, ‘look we’ve had a number of prior enquiries now and we’ve now got secure messaging being used more and more , so those practices not considering an upgrade, or not using it may start to face [compliance] issues. So, I think it will be from a number of angles that we’re looking to create some momentum that will get doctor practices to go ‘look its time now for us to upgrade, we’re confident the technology is working’.Ideally we would get to a point where it won’t be a big headache, and there enough other people using it to make users comfortable upgrading.

WH: There’s not an easy answer to that though really is there?

BM: No, I’m afraid not.

WH: Would you identify the key gaps as GP to specialist, specialist to anyone, and hospitals to GP? 

BM: The key cases we’ve been working on is  hospital to GP, GP referall, which can be inter-hospital or inter-specialist, and e-referral, or specialist and GP. We are also looking at how to include allied health because of the shift to chronic care management needs. These have been the key use cases that we’ve been trying to nail in terms of standards development work.

WH: What about specialists per say? They just don’t appear to use their patient management systems for messaging and interoperability much at all?

BM: It’s a really interesting part of the health sector . And we’re developing a work plan for next year. It’s likely there will be a focus on certain speciality groups. They usually have do a booking system – patient management system that is used for appointments, medical files and those types of things. But there tends to be less use of a clinical information system with specialists compared to GPs.

But if you look at specialists they’re on the bleeding edge of technology with different devices and diagnostic tools, so sometimes you’ll think they’re a bit backwards with technology but that’s not true. Where the technology works for them, they’re often the first on board. We haven’t seen that happen with clinical information systems. So there’s some work for us to do to go in terms of ‘what is it that we can do to support greater use of CISs’ because this will be critical for a connected health system.

Sometimes people say ‘oh look it’s just cultural’ but I just don’t think that’s right. It’s not the experience I’ve seen them have in their surgeries, or rooms, or hospitals. They’re enthusiastic users of technology.

WH: So what is the broader context of what you are trying to achieve with ‘interoperability’?

BM: If you look at the digital health strategy, which describes an Australia in 2022, it is digitally enabled. It’s looking at how services are more accessible to people who don’t have the same transport options, or aren’t as mobile. It’s making things more convenient to people, whether it’s around medicine or seeing different healthcare providers, and all those things are actually underpinned by an interconnected system.

So interoperability to me is really the pathway to get to that future in Australia. We will tick off some things like secure messaging, and e-referral, and discharge summaries, and it will take away some pain points and frustrations for healthcare providers today, which is important. But if we’re really looking at that vision – what sort of healthcare system do we want and how do we make that as seamless as possible for a consumer – to me I think interoperability is what’s  going to get us there. Past our fragmentation between a commonwealth, state, local, private, public and diverse provider needs environment.  It’s really the main game for the long term.

In our interoperability paper we crucially ask, “what are the practical steps we need to take? And let’s just start taking them.”

That’s what we’re trying to get through this in this interoperability discussion and consultation period. Importantly, what are some of the things we can start doing now. It’s looking at some of those small things that have the potential to have some leverage or some significant impact for people.

WH: What do you make of all the negativity around the progress of digital health in Australia and the Agency? Are we just a negative bunch compared say to New Zealanders who seem to be much more collaborative, even commercially?

BM: I don’t really think this that much much because I’m so down in the weeds all of the time. Occasionally you look up and look around, and it’s there. An interesting observation is that I will get international people come and see us, and they say to us that all around the world everyone is looking at Australia and how we’re doing things. What we’re learning, what’s working. We’re actually seen as global leaders in digital health. I’m not referring to the Agency, although we are some part of it.

Through states and territories to major private health institutions – and almost half of Australians has private health insurance –  we’re seen as world leaders in trying different things, and making some progress. Everyone I’m meeting internationally talks about how we’ve got healthcare identifiers for individuals and how it’s such an important underpinning infrastructure to actually build connectivity. In the US, they don’t have them, and they probably never will. They have legislation that that prevents them from actually doing any work on national health care identifiers. It’s easy to forget that, because it’s easier to focus on what’s not done, rather than looking back on what we have achieved.

In a way I think that’s a good thing though. You don’t want to rest, because there’s still so much work to do. It’s why we are consistently saying with secure messaging, there’s just no silver bullet – it’s just going to be long hard slog, we’re not going to solve it all, it’s an issue. We’re not there yet, but we’re going to focus on what we can do, and try to build momentum and try to make each year that we’re a bit better off than the year before. Hopefully that way we’ll build progress.

What’s great is that there are a lot of people who are really passionate about it. That can come through as being critical.  People in health just care so much about what they’re doing compared to other industries. So they just get frustrated with things that don’t happen the way they think they should. It’s sort of natural.

Occasionally I reflect on this and think why is it so different for healthcare and I think that it’s that level of passion, that people invest so much in it that they naturally get frustrated when they don’t see as much progress as they think we should be making.

They’re all good people.

Bettina McMahon will be on two panels to answer  moderator and audience questions on June 25 at the Messel Nano Science Hub Theatre, Sydney University for Wild Health No 4 Summit. Full program with speakers HERE. TICKETS HERE. Use the Wild Health Reader Code of WHP1 for a 15% discount by entering into promotional code when you buy. Look top right for where you enter your code.