Leads: A/Professor Niranjan Bidargaddi, Professor Tarun Bastiampillai, Dr Jőrg Strobel

Non-adherence to medication is common among individuals with severe mental illnesses, and often leads to acute relapse requiring potentially avoidable hospitalisation. In practice medication non-adherence is difficult to detect due to lack of effective monitoring systems. The AI² decision support application (the first ever application of machine learning on My Health Record in Australia) is a real-time monitoring system that uses Medicare and prescription data to provide clinicians with objective information about medication adherence for their patients to support early intervention from health services.

This project further enhances the AI² decision support application to enable collaborative decision making about side effects, dosage and patient choice, and evaluate the benefits of implementation in practice. It will adopt a user-centred approach and develop new components for AI² to capture patient reported side effects and treatment response such as trigger alerts on worsening side effects or inappropriate dosing, extend the adoption of AI² in consultation with clinicians and consumers and evaluate the value/benefits of implementation ultimately closing the evidence-practice gap.

The outcomes:

• will result in comprehensive enhancements to AI² that support medication adherence and address side effect and medication dosage in an early and personalised manner, and thus help to close the current information gap in management of people with severe mental illness;
• produce new knowledge on how to effectively translate digital innovation implementations into practice at scale; and
• and evaluate the application in practice through a quantitative health benefits analysis, health economic evaluation and qualitative evaluation.