The Malinauskas Government is trialling the use of Artificial Intelligence (AI) to support the public health system’s frontline clinicians, embarking on a series of projects that will investigate whether the technology can reduce the time spent on administrative tasks and improve patient outcomes.

If successful, these trials could lead to the roll-out of AI Scribes that enable patient consultations to be converted into clinical documents, administrative work which CSIRO data shows can consume nearly 20% of doctors’ time.

The time-saving benefits of these digital transcription tools have already been embraced by many allied health professionals and up to 40% of Australian GPs, who already use the technology to streamline their clinical work.

Two AI Scribe trials are getting underway at the Central Adelaide Local Health Network (CALHN) and Women’s and Children’s Health Network (WCHN), both requiring patient or family consent:

• CALHN is trialling an AI note-generating system that has been purpose-built for SA Health. It is being used in emergency and urgent care settings to save clinicians’ time and provide patients with easily understood records of their care upon discharge. The patient notes produced by this system are subject to the same data security protection as existing SA Health records.
• WCHN will trial a commercially available AI Scribe, developed by practicing Australian clinicians, in Child Development Unit outpatient clinics and could later adapt the system for use in other departments. Led by the WCHN’s Artificial Intelligence Governance Committee, this pilot trial will look to reduce waitlist pressures by enabling more appointments to be booked.

The WCHN trial is due to begin later this month, while the CALHN trial began last month with the Queen Elizabeth Hospital ED producing transcripts using the program.

Following positive feedback from clinicians, the CALHN trial is now being expanded to include other emergency clinicians and clinicians at the nearby Sefton Park Urgent Care Hub. It is expected that 500 patients will take part in the current research phase over a six-week period, with results to then be analysed to establish whether the tool can benefit emergency and urgent health care.

Traditional notes are still being written at the QEH and Sefton Park Urgent Care Hub while the Scribes are used, to evaluate AI transcripts for accuracy and ensure patient care is not compromised.

These projects are candidates to benefit from funding through the Digital Investment Fund, managed by the Department of Treasury and Finance, which received a $28 million boost in this year’s Budget to support the adoption of AI across the South Australian Government. This funding supports a dedicated AI Proof of Value Fund, which invests in AI trials in priority areas – including healthcare – to evaluate how AI tools can reduce costs and improve operational efficiency.

Both projects have been produced in alignment with national guidance on AI and comply with a new proposed SA Health policy for the safe, effective and ethical integration of AI That policy, which will help guide trial and deployment of AI technology across the health system, is also being released for consultation this week.

The AI Scribe trialled in CALHN has been developed in partnership between the LHN, Australian Institute for Machine Learning, and Commission on Excellence and Innovation in Health, with funding from Health Translation SA and the Health Services Charitable Gifts Board.

Quotes

Attributable to Chris Picton

We are already seeing clinicians the world over utilising Artificial Intelligence to help them work smarter, rather than harder.

If these trials prove successful, AI Scribes could be a real game-changer when it comes to reducing the administrative workload for doctors and nurses, so they can spend less time on the computer and more time with their patients.

We are taking every step to ensure that this technology is safe, secure, accurate and effective, which is why we are trialling these AI Scribes under strict controls.

The very same technology is already being used in the private health sector, and we’re hopeful that, if these trials are successful, AI Scribes will be a powerful tool for frontline clinicians in our public health system.

Attributable to Michael Brown

AI has the potential to improve healthcare delivery by making services faster, more effective, and more efficient.

AI Scribes can record and transcribe consultations, turning these into clinical notes. This reduces the administrative burden for frontline healthcare workers.

Less admin means clinicians can focus on treating patients. More time with patients means more care delivered, faster.

Attributable to Senior Clinical Lecturer at the Australian Institute of Machine Learning, Dr Emily Kirkpatrick

The Australian Institute of Machine Learning, Central Adelaide Local Health Network, and Commission on Excellence and Innovation in Health have collaborated to develop a public hospital ED scribe, trained on SA Health data and purpose-built for busy urgent and acute care settings.

Designed for SA Health, our sovereign AI tool aims to help improve interactions between clinicians and consumers in ED and urgent care settings. This would enable our doctors, nurses, and allied health teams to spend more time with patients and with less administrative burden.

Our goal is to enhance the consumer experience in EDs, with our tool focused on generating a consult note, discharge summary, and providing consumers with a summary of their care in consumer-friendly language.

Attributable to Director of Artificial Intelligence at the WCHN, Dr Melissa McCradden

WCHN is proud to be initiating a trial of a commercial AI Scribe tool to test the safety, quality, and fairness of AI generated draft documentation.

Our goal is to use AI to help us provide better care and service to our consumers.

This trial allows us to estimate the potential improvements to care delivery while maintaining the safe, high-quality and person-centred documentation standards that we are accountable for as an institution.

This trial will allow us to test whether AI can support our clinical staff as they deliver leading care to women, babies, children, young people, and their families.