Establishing International Consensus on Outcomes-Based Evaluation Framework for Artificial Intelligence in Primary Care
PRIMARY-AI: Patient-centered Research Initiative for Metrics And Responsible Yield in AI. Established in 2025, it is a partnership between over 100 academic, regulatory, policy, industry, and charitable organisations worldwide.
Learn More About PRIMARY-AI
People-centeredness deficit: Existing frameworks assess technical performance but fail to evaluate whether AI systems preserve or undermine therapeutic relationships, support shared decision-making, empower patients with health information literacy, or respect patient autonomy and dignity. The person-centered focus that distinguishes primary care from other settings remains unmeasured.
AI evaluation rarely addresses care coordination across multiple conditions (multimorbidity management), transitions between care settings, team-based care workflows, or interoperability with existing health information systems. Yet coordination failures represent major sources of medical errors and patient harm.
Validation studies typically assess point-in-time performance rather than longitudinal impact on continuity relationships, life-stage care trajectories, or chronic disease management over time. This episodic evaluation misses primary care's defining temporal dimension.
Interested in our work? For more information or to get involved in our projects, please contact us via the following: