Do We Have Enough Information? Assessing AI Clinical Decision Support Systems for Implementation in Primary Care

Abstract

This environmental scan examined commercially available AI clinical decision support solutions (AI-CDSS) across three domains: knowledge base, AI methodology, and privacy. Over half of vendors disclosed some information on their knowledge base, yet few demonstrated rigorous appraisal or alignment with Quality Standards or other evidence-based guidelines. Transparency on AI methods was limited as most cited proprietary algorithms but rarely described training data. Privacy information was more commonly reported but often high-level, with limited detail on compliance, storage location, or restrictions on secondary use. These gaps reveal the obstacles facing decision makers: without standardized, transparent information, organizations and governments cannot reliably evaluate AI-CDSS or provide the support clinicians need for responsible and informed implementation.