Abstract

Artificial intelligence (AI) presents opportunities and challenges in post-discharge psychiatric care. Leveraging structured data and machine learning, the Centre for Addiction and Mental Health aims to predict adverse outcomes, including readmissions, among patients recently discharged from psychiatric units. By identifying high-risk individuals, AI can guide referrals to resource-intensive outpatient clinics, enhancing continuity of care and improving outcomes. A governance framework addressing ethics, transparency and fairness underpins the development and implementation process. The study emphasizes using interpretable AI models over black-box systems to foster trust and clinical utility, aligning AI advancements with ethical mental health practices.