Co-Designing a "win-win" in Predictive AI: First Results from Interviews and Focus Groups with Persons with Parkinson's Disease.

This study explored the perspectives of people with Parkinson's disease (PwP) involved in the co-design of AI tools for PD care. The aim was to understand PwP perspectives on AI tools and identify factors influencing their engagement. A qualitative triangulation of 13 interviews and two focus groups were conducted with a panel of 14 PwPs from six European countries. Data analysis employed inductive content analysis which resulted in 3 categories. Patients emphasized the potential of AI tools to increase patient agency and improve healthcare through co-design. While AI-based medication response prediction was highly regarded, risk assessment tools were met with skepticism. Participants stressed the importance of taking into account disease complexity, individual factors, data protection, and transparency in AI model development. Participants strongly believed in the co-design process, and the value that it could bring to all parties involved in the development of the AI-based tools, ensuring the development of tools that are beneficial and trustworthy.