• 16 Jan 2024
  • 14 min 50
  • 16 Jan 2024
  • 14 min 50

Jo Cheah speaks with Jodie Austin, a clinical informatics specialist, about using new technologies to deliver health care. The conversation focuses on using digital health to improve prescribing and medication management, through the use of digital dashboards and machine learning. Jodie also outlines the role of the clinician in the developing digital world. Read the full article by Jodie Austin and her co-authors in Australian Prescriber.


I think with digital health, it's never about taking the place of the clinician. I think the clinician is always going to have that best judgement call with the patient there in front of them. These are just tools to be able to help them make the best informed decision.

[Music] Welcome to the Australian Prescriber Podcast, Australian Prescriber. Independent, peer-reviewed and free.

Hi. Welcome to the Australian Prescriber Podcast. My name is Jo Cheah. I'm a hospital pharmacist in Melbourne and your host for this episode. It is my pleasure to welcome Jodie Austin, Clinical informatics director at the Queensland Digital Health Centre at the University of Queensland. Welcome, Jodie.

Thanks for having me, Jo.

Jodie has co-authored an article about digital health and prescribing in Australian Prescriber. To begin, Jodie, what is digital health and how does it impact on prescribing and medication management?

Digital health is really just about using information and technologies to deliver healthcare. So it's really a broad term and it encompasses a lot of different things. So for example, mobile health, health information technology, wearable devices, telehealth, personalised medicine, all these things fall under that broad banner of digital health. I think also we're seeing now that digital health is really multifaceted in the sense that it's not just about those technologies and devices but it's also about the data that's being generated through those technologies and the analytics that we're using to capture that information. So I think in terms of the impact it's having on medication management, we're seeing the ability for it to really promote precision medicine so we can use clinical decision-support tools where we're using the individual's specific characteristics, but we are matching it to large databases of information to guide best practice.

And I think health information technology is also impacting medication management, not just really at that prescribing level, but all the way through that medication management cycle from the prescribing to the administration to the monitoring and the supply etc. So it's really trying to capture safety features across that whole cycle of managing your medication.

It is good to know that, yeah, digital health is a very broad term but today we'll be discussing prescribing and the medication management pathway. So in your article you describe digital health transformation across 3 conceptual horizons. Can you explain these horizons? And if you don't mind as well, comment on the current status of digital health in Australia, noting that there are differences across states and territories.

So this conceptual horizons is looking at Horizon 1 being really where organisations are focused on that initial laying of the digital foundations. So implementing, for example, electronic medical records using the basic system functionality that comes with that. So for example, computerised physician order entry or clinical decision support etc. And that's really becoming part of routine workflows and then capturing digital data for every patient at every encounter during routine care. So that's really that Horizon 1 concept.

And then building on that, Horizon 2 is where those digital foundations have been laid, the digital workflows are embedded, and then we are starting to be able to capture that data and harness it using clinical analytics. And then we can really start to aggregate that information and look at high-level summaries in real time or near real time, and then drill down on that data for individual patient care.

And then building on that again is Horizon 3, which is ultimately what we're trying to head towards with this digital transformation journey where we become a learning healthcare system. And that's where we're using that data in real time and integrating it into new models of care and clinical workflows so it becomes the new way of practising health care. It's accepted and it's embedded, and we are being able to close that feedback loop in close to real time.

If you're looking at it in terms of how we are across Australia, I think for the best part, Australia has really been focused on that Horizon 1 up until this point. We've been busy across the states and territories in implementing electronic medical records and we are starting to see pockets of Horizon 2 and I guess potentially Horizon 3 happening as a generalisation. But really we're only just starting to establish that Horizon 1 and at the cusp of understanding how do we harness this big data that we're now generating through that Horizon 2.

If you're looking at it in terms of how we are across Australia, I think for the best part, Australia has really been focused on that Horizon 1 up until this point. We've been busy across the states and territories in implementing electronic medical records and we are starting to see pockets of Horizon 2 and I guess potentially Horizon 3 happening as a generalisation. But really we're only just starting to establish that Horizon 1 and at the cusp of understanding how do we harness this big data that we're now generating through that Horizon 2.

And I think there is a lot of discrepancies across the state. And if we look, for example to the US, who are potentially a little bit further along with their digital transformation journey. So they had the HITECH [Health Information Technology for Economic and Clinical Health] Act, which was introduced back in 2009, I think it was. And that was for 10 years allocating incentives for hospitals and healthcare providers to implement electronic medical records. And so they've seen this widespread adoption across their country, whereas in Australia it's been a bit more of a fragmented approach. So we're seeing that it's the state governments that are funding the rollout of integrated electronic medical records, for their public hospital system, for example. And so everybody's at different stages and we are seeing Western Australia where they've just announced recently that they're going to start their electronic medical record rollout. We've seen other states, Queensland, New South Wales, Victoria, Northern Territory, South Australia, which are a little bit further along on that journey, but everybody tends to be using different vendors, for example. So I think we've got a little way to go in terms of collectively being able to harness this data across the nation.

It sounds like we've got a long way to go to get to the maximum capability of digital health in Australia.

Yeah, definitely.

So what benefits does digital health offer for clinicians and patients?

I think, as I said, because it's so multifaceted, there's a lot of different benefits that are on offer. So if you're giving that with the telehealth, the virtual care, you're providing greater access to healthcare for people that are living in marginalised communities, rural remote areas of Australia, for example. So that's one area that we're benefiting.

The premise is we're trying to create a more efficient health system. So that's, for example, less time searching for patient data, reducing duplication in terms of pathology tests, X-rays, other forms of radiology imaging. And so by reducing those inefficiencies, hopefully we are reducing the costs associated with those inefficiencies as well. The other thing that digital health is really trying to promote is the improved quality of care. So for example, reductions in adverse drug events, reduction in medication errors.

Could you provide examples of digital dashboards as mentioned in your article and how these are used in clinical practice?

Yes. So digital dashboards is really what we're talking about as we're starting to move into that second horizon of digital transformation. So the data is being generated through routine care but we can start to try and capture that information and display it in high-level succinct summaries for our clinicians so that they can easily understand the current status of what's happening. And it would depend on specifically what that digital dashboard is intending to capture. So if we're looking at high-risk drug management, we can have dashboards that will be capturing patients that are being prescribed opioid therapy in an inpatient setting. And then we begin using that to flag those patients that are either poorly managed, so routinely high pain scores at rest, or they're experiencing potential adverse drug events as a result of that opioid therapy. So we begin tracking it and looking at things like reductions in respiratory rates, or their bowels not opening over a certain period of time.

And we can do the similar thing at a population health level. So where we can start to capture information out of the electronic medical records looking at, say, obesity rate, and then matching that with Australian Bureau of Statistics data so we can start to get the geographical locations of people who are overweight, a healthy weight and underweight across a state or a nation, for example. And then that can really help for policymakers in terms of targeting intervention strategies. So these are all the types of things that we can start to do, creating digital dashboards and harnessing that data that we're now generating through routine care.

You also mentioned machine learning algorithms within the article. So would you be able to provide some examples of machine learning algorithms and their clinical applications?

Machine learning is really, it's the idea of giving computers that ability to learn without explicitly being programmed. And so it really is an emerging space. So most of the applications are still really in the research and development stage. And so some of those uses, for example, are image analysis. So automating the systems that will help to examine X-rays or other diagnostic images. Also looking in terms of pathology, maybe analysing biopsy samples. And the other space we're seeing a lot of interest and a lot of research is in clinical decision support. So creating these algorithms that are trying to screen the data to look for people who might be at risk, for example, of developing sepsis or developing poor glycaemic outcomes in diabetic patients. So there's a lot of interest in this space and a lot of work happening. There's just not a lot that's being applied clinically just yet.

Very interesting. And you mentioned some examples there but I guess you would still need people to analyse what the machines provide as outputs. Would you agree with that, or are we moving towards a solely machine-generated decision?

Yeah, definitely not. I think with any of this digital health, all of the different types of strategies, and machine learning included, it's never about taking the place of the clinician. I think the clinician is always going to have that best judgement call with the patient there in front of them. These are just tools in their tool belt to be able to help them make the best informed decision at that point in time.

And what are the key barriers to maximising digital health implementation and how can they be overcome?

Key barriers at the moment are things, like we touched on previously with the different states, is the interoperability. So we've got a lot of silos still happening. People are becoming digital but the systems aren't talking to one another and that's a big wicked problem that's not easily fixed. But we have the Australian Digital Health Agency that's looking at creating frameworks for action to enhance standard specifications to address that siloing that's happening across the digital system.

Another issue that springs to mind is the digital readiness of the workforce. And I've certainly seen that in my time working clinically, the disruption that's caused when you're implementing these large-scale digital systems and it can be really taxing on the healthcare workforce. So I think there's a lot of emphasis now that's going into trying to make sure that our healthcare practitioners are sufficiently trained in how to use the different digital technologies that are coming up.

The younger generation of clinicians are coming through, certainly a lot of them are very technology savvy as it is. They've just been introduced into that really from birth. So I guess it's becoming easier as time goes on. And we're also seeing in the university sector specific postgraduate courses that are looking at digital health and clinical informatics for people that are interested in specialising in those areas. And also introducing it into the undergraduate courses, whether it be medicine, pharmacy, nursing etc. to really try and introduce them to these concepts from the very beginning. Another key thing is the need to improve on the infrastructure to support data reuse.

But a lot of the legislation that's being used to reuse that secondary use of health information was written prior to the digital era. And then we are still not quite sure what that social licence is yet with the consumers as to are they happy for the secondary use of digital data to be used to advance medical research.

Whether or not it's ethical or legal to access that data is very interesting, especially with all the data leaks and things we're seeing around the place.

Absolutely. So that's where we really need to understand what that social licence is out there with our patients. Really it's their data that's being generated. So we need to understand not just internationally, but we need to understand nationally and locally what that social licence is to be able to make sure that we are tapping into that data, but we are doing it in an ethically appropriate way.

How do you think digital health might transform clinical practice and support new and innovative models of care, particularly in relation to prescribing and medication management?

We've touched on it a little with some of those other questions. So I think transforming clinical practice, what we're trying to ultimately achieve is that learning healthcare system. So it's a digital platform but the digital platform is generating that data. And I think clinicians may be starting to look at it as evidence-generating medicine. So by practising your routine care, you are generating the evidence that can then be tapped into to analyse that data and then put it back into the health system to close that feedback loop.

It's not just individual patient care anymore, but this data is being used collectively and feeding back into the system. So that's where we are really trying to transform the clinical practice. And I think we're also starting to see, in terms of new innovative models of care, and there was an example that we've used in that editorial where even though machine learning, for example, we're not seeing that used or applied here in Australia routinely just yet. There is an example where they've used that and implemented it across a set of 5 hospitals in the [United] States. And they were using that predictive modelling to try and pick up patients that were potentially at risk of sepsis and introducing antibiotic orders ahead of time. And so they did see a reduction in sepsis mortality. So it's that idea of maybe with prescribing, it's going to shift more to a potential predict-and-prevent model as opposed to where we're currently at with that break–fix approach.

So Jodie, that brings us to the end of the episode. Thank you so much for joining us today.

Thank you for having me.


Jodie's full article is available on the Australian Prescriber website. The views of the hosts and the guests on the podcast are their own and may not represent Australian Prescriber or therapeutic guidelines. I'm Jo Cheah and thanks again for joining us on the Australian Prescriber Podcast.