The Department of Psychiatry at Sinai Health delivers a wide range of inpatient and outpatient services — from General Outpatient and Medical Psychiatry to Perinatal, Geriatric, and Emergency Psychiatry. Across these programs, clinical and administrative data are collected in many different ways — often through Excel files or manual processes. While this information is valuable, it’s currently fragmented, inconsistent, and difficult to use for real-time decision-making.

This project aims to map the department’s current data landscape to understand what is being collected, how it is being collected, and how it could be improved. By creating a clear inventory of existing data and identifying opportunities for streamlining and standardization, we will lay the foundation for AI-enabled data systems that can support better, faster decisions in clinical operations.

Objectives

• Create a comprehensive inventory of all data and metrics currently collected across Psychiatry programs (using a structured Google Sheets tool).
  

• Assess data quality, relevance, and redundancy — identifying what can be discontinued and what needs improvement.
  

• Determine data needs for real-time clinical and operational decision support.

• Develop a roadmap for AI integration, outlining how machine learning or automation tools could help with data organization, quality assessment, or predictive insights.

In Ontario, more than half of people who visit an emergency department (ED) for a mental health crisis do not connect with follow-up care within two weeks, a gap that leads to avoidable harm and repeat visits. This project tests a practical, low-cost digital solution to close that gap.

Dr. Ben Rosen (Sinai Health) and Dr. Venkat Bhat (St. Michael’s Hospital) are leading the evaluation of MIRA (Mental Health Intelligent Information Resource Assistant), a privacy-preserving digital navigator co-developed with the Mood Disorders Society of Canada. MIRA translates patients’ plain-language needs into current, local, eligibility-aware mental health and addictions options, generating a one-page discharge plan that can be printed or emailed at the bedside.

The study embeds MIRA directly into the emergency psychiatry discharge workflow to replace static, outdated resource lists with personalized plans. Using existing follow-up calls by Psychiatry Emergency Services, the team will measure whether MIRA improves the rate of connection to community care within 14 days and reduces 30-day ED revisits.

If successful, this model could be rapidly scaled across Ontario hospitals, strengthening continuity of care, improving patient experience, and offering a data-driven path toward more coordinated mental health systems.

Partners: Sinai Health, St. Michael’s Hospital, University of Toronto, Alberta Machine Intelligence Institute, University of Wisconsin–Madison, and the Mood Disorders Society of Canada.