Photo: andresr/Getty Images
A new AI tool developed at the University of South Australia can predict when a patient can safely withdraw from taking long-term antidepressants.
WHAT IT'S ABOUT
Using AI, UniSA researchers analysed passively collected and underused dispensing data from the Pharmaceutical Benefits Scheme (PBS) to identify the most successful cases of deprescription among 100,000 patients over a 10-year period.
The team trained and tested two machine learning algorithms – one assessed final prescription records and the other monitored patients' dose reductions and outcomes from their first prescription.
"By applying AI to the PBS database, we have identified patterns linked to successful withdrawal, forecasting which patients are most likely to succeed when taking them off antidepressants," explained Dr Lasantha Ranwala, UniSA doctorate student and AI researcher.
Based on findings presented at MedInfo 2025, an international conference on digital health and informatics, the models demonstrated 81% and 90% accuracy, respectively.
WHY IT MATTERS
Antidepressant use has continued to soar worldwide, including in Australia, where the medication was dispensed to 14% of its population in 2023-2024 alone. Among Australians aged 10-24, who already accounted for 35% of long-term users, prescriptions saw the sharpest increase in the same period.
Prolonged antidepressant use has been linked to such side effects as weight gain, sexual dysfunction and cardiac issues. Studies have also noted that about half of patients experience withdrawal effects after ceasing medication.
An AI that can predict safe antidepressant withdrawal can help general practitioners deprescribe obsolete medication "with greater confidence," according to UniSA.
"Healthcare providers are often reluctant to cease antidepressant prescriptions due to their concerns about withdrawal effects, making it difficult for doctors to know who can safely discontinue treatment," claimed Dr Ranwala.
A successful deprescription is defined as the absence of any antidepressant medication for at least one year following previous long-term use (more than 12 months). A deprescribing attempt is considered a failure if the medication strength rises again within six months of reduction.
The research team is now working to make the AI tool more accurate and easier to use, with plans to test with clinics and explore how AI could be used to help patients improve their medicine intake.
THE LARGER TREND
Studies worldwide have increasingly demonstrated AI's potential to support clinicians in developing optimal, personalised medication protocols. Another new study in New South Wales, for example, is using AI and genomic data to predict patient responses to antidepressants. The project, which recently received nearly $2 million in federal grant funding, aims to reduce trial-and-error prescribing.
A Monash University-led study has developed AI that can identify the optimal anti-seizure medication for newly diagnosed epilepsy patients.
