New project: Predictive models to optimize mental health services in Quebec

Congratulations to  Dr. Massimiliano Orri‘s team and all co-applicants/collaborators for obtaining significant CIHR funding to develop predictive models in mental health. The project—Leveraging predictive models to optimize mental health service delivery and improve clinical practice—builds directly on the collaborations, infrastructure and expertise made possible by the Alliance santé mentale Québec, which enabled its emergence and provincial deployment.

A unique collaboration in Quebec

Mental disorders represent a significant burden for individuals and the healthcare system, leading to frequent readmissions, prolonged stays in psychiatric wards and repeated visits to the emergency room. To address these issues, the project uses machine learning to predict the risk of hospital readmission and the length of stay in psychiatric wards.

The data comes from four hospitals, allowing for inter-site harmonisation that is unparalleled in Quebec: Douglas (CIUSSS-ODIM), IUSMM (CIUSSS-ESTIM), CERVO (CIUSSS-CN), and IR-CUSM. This collaboration makes it possible to develop and test robust models that are representative and directly applicable to the clinical realities of the Quebec network.

Tools that will move from research to clinical practice 

Beyond predictive models, the project includes a significant implementation component. Interviews with clinicians, managers, decision-makers and patients will identify the winning conditions for the practical adoption of these tools in everyday practice.

Led by an interdisciplinary team combining clinical expertise, data science, implementation science and lived experience, this initiative will contribute to better service planning, resource optimisation, and improved quality and efficiency of mental health care in Quebec.

A committed team