Grant: Health System Impact Fellowships - PostDoc Central Canada 2 year

A strategic priority for the Centre for Addiction and Mental Health (CAMH), the host organization, is evidence based care. The heart of this initiative is the Krembil Centre for Neuroinfomatics (KCNI), which employs state-of-the-art artificial intelligence (AI) and machine-learning (ML) techniques to identify, manage and treat mental illnesses, while shaping policy at national and global levels. KCNI manages and curates patient data to create algorithms capable of modelling disease pathways, predicting adverse effects, and facilitating personalized treatment interventions. Concerns about AI applications systematically discriminating against specific populations have spearheaded an urgent conversation about notions of fairness. Namely, who will/should benefit the most from these techniques and why? Thus, the primary objective of this project is to examine how fairness is defined, managed and produced along an Integrated Care Pathway (ICP) for Major Depressive Disorder (MDD). We ask: what makes training datasets "fair" and how are principles of "fairness" incorporated into decision-making by human (patients, clinicians, data curators) and non-human (AI) care teams? We anticipate that the concept of fairness is not fully developed, thus a key health system impact objective is to develop guiding principles for incorporating fairness into hospital-wide efforts to engage with AI/ML techniques for health equity. To achieve these objectives the fellow will conduct an institutional ethnography to analyze how concepts of fairness are used to inform the ICP for MDD. Patients, clinicians, and data curators who use and interpret data for recovery will be interviewed and observed. This project will foster the responsible interpretation of knowledge derived from AI/ML and is essential to ensure that policy uptake is relevant and beneficial for all.