Protecting and Engaging vulnerable populations in the development of predictive Models in primary healthcare for inclusive, diverse and equitable AI (PREMIA)

This project aims to carry out an environmental scan of AI-based predictive models in primary health care in the published literature and by conducting interviews with key stakeholders; to form focus groups including experts in clinical, ethical and experiential knowledge of vulnerable populations who will work to identify and assess biases in predictive models; to implement recommendations from the working group into a model, gather experience from developers and users, assess acceptability, feasibility and relevance of parameters and fidelity of implementation of recommendations.