Interoperability: find out what it means to me
Australian users of clinical information systems have told us what information they need from interoperable clinical systems. Understanding these needs will enable us to shape and prioritise the development of Australia’s digital health system.
Interoperability between clinical systems is both the greatest challenge of digital health and its greatest promise. Given this central role, it has been a prominent topic in our consultations with clinical users of information systems across Australia. A single clear question has emerged from these consultations:
What does someone else know about this patient, that I don’t know and that is relevant to the current context?
On the face of it, this seems like a simple enough question. However being able to answer that question in real time, in a clinical context is extremely challenging. Let’s break it down to analyse it and understand it better.
“About this patient”
First and foremost, we need to identify the patient correctly across disparate clinical systems and resources. Within Australia, the National Health Identifiers Service plays a critical role in ensuring that we match the same patient’s information across multiple systems. This shared identity framework is the fundamental enabler of all interoperability.
“What does someone else know about this patient?”
To find out what else is known about a patient, we first need to know where to look. Timeliness is frequently a critical factor in a live clinical context, which rules out time-consuming searches through resources of unknown quality. Only resources that are known to provide high quality patient information are suitable, which in turn implies that the search will be restricted to a very small number of places.
There are some collaborative projects around Australia where networks of providers use a dedicated health information exchange to share data between providers, such as South Western Sydney’s Project iRAD.
In the main, however, the most useful repository for information about patients is their My Health Record.
“What is the information that I don’t know?”
In order to make a comparison between what is already known (the data in the local system) and inbound data from another source, there needs to be a common language. If my database says “Full Blood Count” and the incoming data says “Complete Blood Count”, how can the system tell that they are the same thing, and not new information?
When a GP clinical system and a laboratory exchange data, they do so using an agreed language. The two companies agree: “We’re going to call it a Full Blood Count”. That’s a relatively easy thing for two companies to do, especially when it meets the need both company’s customers. Unfortunately, within Australia many of these bi-lateral agreements have organically sprung into being with a large number of differences.
Within My Health Record, it’s impossible to know, in advance, what laboratory, or doctor, or imaging provider (etcetera) the information might be coming from. Therefore it’s equally impossible to know what “dialect” is in use.
So what’s the answer? Ultimately, we need to agree to implement a consistent language for clinical concepts at the interface. We already have large and well-defined national terminologies, but these have not yet been consistently implemented, and there are a broad range of reasons for this. One of these reasons is that there has never been a mandate for a standard terminology.
“What information is relevant to me now?”
How can we choose what information is relevant to the current context? For now, we need to rely on the expertise of people who provide services. What’s important in the GP context may not be what’s important in the Emergency Department context, or the elective surgery context.
In the future, machine learning and other advances may offer new ways for clinicians to keep abreast of the information that’s important about their patients.
So, how do we move forward?
What seemed, at first, to be a fairly simple question is actually quite complex. However, we also have some understanding of the human and financial costs of the current fragmentation of Australian digital health information, enough to know that these complexities are well worth tackling.1
It goes without saying that the resolution of these issues will arrive incrementally, one step at a time. In healthcare, these steps must not only take us closer to the best possible solution, each step must individually make the lived experience of participating in the healthcare system better.
After listening to our stakeholders, we know your concerns. We’ve distilled these down to a number of priorities for action. One common theme of your feedback has been the need to redefine, in partnership with industry, a development model that clarifies the process and roles for the development and ratification of standards and specifications for healthcare in Australia. Another prominent concern is to make the healthcare system safer for people with known allergies. And there are many others.
Returning to our question: how do we move forward? The answer, as always, is “Together”. We will continue to consult with our stakeholders as we co-design the journey toward delivering a truly interoperable digital health system for all Australians.
Australia’s National Digital Health Strategy – Safe, Seamless and Secure lists a number of forecast improvements in healthcare outcomes and systemic costs. To cite one example, it has been estimated that an effective digitally enabled medicines management program could reduce adverse drug events by 50%, which would prevent over 100,000 hospitalisations annually, saving taxpayers over $500 million each year (p.34). ↩