Utility of the Medical Knowledge AI Copilot, OpenEvidence: A Patient With 100 Cerebral Microhemorrhages

Cerebral amyloid angiopathy (CAA) is a highly prevalent condition that often poses diagnostic and therapeutic challenges, especially in people with comorbid mild cognitive impairment (MCI). OpenEvidence (OE; Miami, Florida, US), a leading free medical knowledge copilot, reports it has supported over 100 million Artificial Intelligence (AI)-powered clinical consultations from U.S. doctors and other frontline clinicians. The objective of this study is to use a single case with two common comorbid neurological conditions to retrospectively appraise the ability of OE to serve as a clinically relevant knowledge aid. We input the de-identified summary of a case of probable CAA and MCI in OE. Overall, the AI recommendations very closely matched the diagnosis and care that was actually provided. We identified some areas (e.g., prognostication, geriatric risk-benefit trade-offs, diagnostic terminology) where it could be improved to match the logic and knowledge of a subspecialist. OE used as a knowledge copilot can help inform point-of-care physician decisions by giving them more information than they may have otherwise considered.