A matter of life and data – Dr Clair Sullivan, Queensland Digital Health Research Network (QDHRN)

Clair Sullivan Metro Health

We need to stop looking at digital technology as a tool to make care quicker and more efficient and instead shift our thinking to see it as a quality improvement tool.

Dr Clair Sullivan has been the driving force of a data-led revolution in Australian healthcare.

As Head of the Queensland Digital Health Research Network at UQ’s Centre for Health Services Research, a research and educational collective tasked with fast-tracking the digitalisation of the state’s health services sector, and Director at Metro North hospital, Dr Sullivan has worked tirelessly to bring blue-sky innovations in digital health – from clinical informatics and analytics-based interventions – to the front-line of clinical practice.

We sat down with the University of Queensland Associate Professor and chair of the upcoming FST Government Queensland 2022 conference, to gather her thoughts on Australia’s Covid response and the Government’s National Digital Health Strategy, on reconnecting policymakers with the on-the-ground demands of healthcare academics and clinical staff, and what her ultimate wish list is for Queensland’s health sector.

FST Government: The hospital system, and indeed the healthcare sector more broadly, has come under considerable strain over the last two years.

Paint us a picture of Australia’s health sector today and perhaps what Covid has taught you about the growing importance of digital in healthcare?

Sullivan: Current methods of healthcare delivery were in fact strained prior to the pandemic, and the extra burden of Covid-19 created a pressure cooker situation resulting in the rapid and effective implementation of novel methods of care.

We saw this across the three horizons of digital health, the first being digital workflows, where we suddenly saw people who’d never used digital systems using them at the point of care. On the second horizon, data and analytics, we saw the compute power of data being used to drive decision-making in real-time.

This [focus on data] saw the sector starting to create real-time analytics about exposures, public health orders, and Covid cases to drive decision-making using digital tools.


And the third horizon is taking those digital workflows and that data and analytics to create new models of care. That’s where we’ve started to see virtual care with risk stratification being deployed at scale and at a rate we’ve never seen before.

Across all three horizons, Covid has really accelerated progress, and it’s allowed us to cope with what could have been an even worse situation in a relatively satisfactory manner.


FST Government: Digital transformation programs are fast maturing in the corporate world, with the public sector slowly making headway.

What gaps have you noted in the local healthcare sector’s digitalisation objectives, both in Queensland as well as Australia more broadly, and how can we plug these holes?

Sullivan: These themes are, of course, common whether it’s in Queensland, Australia or globally. Barrier one is the capital required to implement digital health solutions. Large scale transformation projects are expensive and that’s not palatable when there are many other pressures. It is, however, essential.

The second barrier is the operational costs to run them. Just deploying an EMR [Electronic Medical Records system] won’t push you along those three horizons that I’ve described; there has to be an acceptance that, although there’s capital expenditure, there’s also significant operational expenditure to maintain momentum in your transformation. Those are two barriers that I think aren’t yet well articulated or understood.

The last one is the funding model for care.

I still don’t know why funding telehealth is so controversial!


Re-thinking how we fund healthcare to include virtual care, to include continuous contact rather than discrete appointments needs to be discussed.


FST Government: Would you say then that Australians have become more comfortable utilising telehealth services, embracing what one might call a ‘hybridised’ care model?

Sullivan: You certainly don’t need to be a digital health expert to see that! You’re seeing elderly people for the very first time using technology to interact with their care providers. It’s wonderful! The evidence shows us that and, anecdotally, we see it as well. It’s been a great progression.


FST Government: The healthcare sector is no doubt grappling with an abundance of legacy infrastructure. How is Metro North, in particular, managing the legacy burden and what’s on your technology roadmap as you look into a post-Covid future?

Sullivan: It’s really a matter of ‘Keep going!‘. To be honest with you, I don’t think I’ve ever had a technology roadmap. I do, however, have a transformation outcome agenda; the priority in this is to continue with maintaining digital workflows, to continue with data and analytics, and to further that third horizon of implementing new models of care. While the technology itself changes, those are the principles and the strategy that we follow.

Of course, we still prioritise low code and no code development. We’re not looking to point solutions, but rather using enterprise architecture and infrastructure and undertaking low code or no code innovations is absolutely the way to go in a complex digital health care environment.

We really are shifting in digital health from an entrepreneurial mindset to an enterprise activity.


FST Government: You mentioned budget constraints earlier and issues securing funding. There still appears somewhat of a disconnect between the concerns of academics and front-line staff with the priorities of policymakers.

How can healthcare staff and the academic community better translate the benefits of investing in a robust digital health ecosystem?

Sullivan: In fact, I very recently published a paper in npj Digital Medicine which addresses this very problem.

Expecting to use digital health as a cost that can be recouped through financial ROI is a road to failure.


What you really need to think about is healthcare outcomes, the data and analytics, the precision medicine, and the AI, which come five to 10 years down the track. At the moment, those impacts have not all been costed nor included in traditional business cases.


FST Government: No doubt you’ve had a good look at the Federal Government’s National Digital Health Strategy (NDHS), which is set to come into force this year.

What are your thoughts on the strategy? Do you consider the priorities laid of the NDHS achievable within the government’s set timeframe?

Sullivan: It’s critical that we have a united strategy and I think Health has done a good job of bringing that together, so I certainly support it.

Of course, nothing’s ever fully achievable in health! It’s a never-ending piece of string. But it’s important that there’s a plan and a strategy.

The roadmap laid out is ambitious and our job is to get behind them to achieve it.


FST Government: Clinical informatics is an increasingly important discipline in modern healthcare. Walk us through what clinical informatics offers for Australia’s healthcare providers and some of the ground-level benefits for clinicians?

Sullivan: Five to 10 years ago, healthcare was digitally naïve. A developer or a technologist could walk in and implement a piece of technology to specifically address a point problem. Over the last five to 10 years, however, complex digital health ecosystems have evolved in our healthcare settings.

No longer can you just walk in and implement a point solution – it will clash with existing workflows, there’ll be interoperability issues, there’ll be standards issues, and issues with terminology and nomenclature.


Clinical informatics is really that layer between the technologists and clinicians or other practitioners, and our job is to understand what the digital health ecosystem is.

It is about connecting the two and driving the digital transformation agenda. The technologists won’t necessarily be driving the agenda – they’ll be driving the implementation of their product. Clinicians won’t necessarily be driving for those better outcomes – they’ll be wanting to do that job as quickly as possible. So, your job is to be that layer between the two, with a firm focus on strategic outcomes and pulling everyone along behind you.


FST Government: Clinical informatics of course depends on quality data. How can healthcare providers ensure they’re getting the most out of their data assets?

Sullivan: I don’t think healthcare providers are getting much out of their data at all! Most data in Australia is still used for simple point of care treatment.

We need to stop looking at digital technology as a tool to make care quicker and more efficient and instead shift our thinking to see it as a quality improvement tool.


Once you do that, you realise that data isn’t simply about providing a patient with quicker care on a Saturday morning, it’s about the compute power, resolving the questions of, ‘How many patients are there?’ ‘What problems do they have?’ ‘How can we get in there to intervene so that they don’t develop complications?’. That step really hasn’t been taken by many healthcare providers.

Compute power is where the real benefits of digital transformation lie, not in digitising the paper workflows.

In terms of use cases, hospitals right now are starting to think about this in terms of using the compute power of their data to improve decision making to predict and prevent complications.


FST Government: Looking to another data-driven technology, one that’s been hyped by healthcare technologists for at least the last few years – artificial intelligence. While touted for its potential to revolutionise health diagnostics, there remain legitimate concerns around algorithmic biases, misdiagnosis, and of course data privacy.

What role do you see for AI in healthcare and how can we ensure our deployments remain in all patients’ interests?

Sullivan: To me, it’s just another tool. Early in the piece, developers claimed that it would replace doctors; that generated a lot of emotion and angst. That time has now gone.

When an MRI, for instance, comes back with a report saying that somebody has a tumour, we don’t automatically assume that they do. We use that information, synthesise it with other sources, and then make a diagnosis – the correct decision for the patient.

I’m relegating AI to the same. It’s simply a source to help us, as practitioners, make a decision, like a blood test or an X-ray. But it shouldn’t make the decision for me.


FST Government: Have you noted any encouraging local use cases in AI?

Sullivan: There’s quite a few in very simple workflows to determine ‘What is it?’. Things like X-rays, retinopathy, where you can essentially identify things from pictures – a very simple, very limited workflow. There aren’t, however, too many use cases looking at the ‘Why?’ or ‘How is it so?’ of a diagnosis.

These are the workflows we go through when we see a patient deteriorating in front of our very eyes and we don’t know what’s wrong: ‘Is it sepsis?’ ‘Is it heart failure?’ ‘Is it that complex decision-making that requires multiple inputs, multiple outputs and multiple variables?

While we’re singing the praises of these simple workflows that are potentially better than, say, a radiologist in finding a spot on a picture, that’s a very small part of doctoring.


But the actual ‘hard doctoring’ is when you’ve got someone who’s desperately sick in front of you and you need to figure out what’s wrong with them. In that setting, we really don’t have many success stories from AI yet.

However, we are working on some AI algorithms. We’ve been fortunate enough to receive some funding for a project (see Dr Anton van der Vegt) using AI to help earlier detect deteriorating patients in hospitals in Queensland, which is well underway as we speak. There’ll be plenty of hurdles, but it’s also very exciting to be making a difference in an area where we know that patients are suffering unnecessarily.


FST Government: Finally, what is on your immediate wish list for Queensland health?

Sullivan: We need to ultimately become a learning healthcare system. My challenge is to ask, ‘Are we learning or are we just doing the same thing over and over again trying to get faster?’, and ‘Are we measuring what matters, looking at our outcomes and continuously improving them, or are we just delivering care and hoping we’ve seen 50 patients today?

At the end of the day, are we just delivering healthcare or are we a learning healthcare system? And I believe – and I think consumers would hope – that we’re monitoring our outcomes and continuously working to make them better. And we need digital health to do that.

We are privileged to welcome Clair Sullivan as chair of our FST Government Queensland 2022 conference on 31 March 2022.

Across the day, public sector leaders will explore the Sunshine State’s post-Covid response, front-line service innovations and back-office efficiency gains, cybersecurity best practice and posture, and how leading government agencies are negotiating shrinking ICT budgets, plus lots more.