Shared Decision-Making and Clinical Decision Support Systems in Neurology: Autonomy, Trust, and Clinical Judgment

This paper examines the ethical questions raised by the development of AI tools for clinical decision support in neurology. I draw on my current work in the PRIMUS project, which develops a decision-support tool for therapeutic choice in multiple sclerosis care, and on exploratory ethnographic observations of neurological consultations in France. The paper starts from a question that has become central in the ethics of AI in healthcare: how should patient autonomy be understood and supported when an algorithm contributes to diagnosis, prognosis, or therapeutic decision-making? In many debates, this question is approached through the language of information, transparency, explainability, and consent. Patients should be told that an AI tool is being used; they should receive information about its function, limits, and risks; and they should be able to consent to its use. These requirements are important, but they leave aside part of what is at stake in ordinary clinical decision-making. In multiple sclerosis care, diagnosis and treatment decisions often take place in conditions of uncertainty. They are temporally extended, depend on clinical experience, and involve vulnerability, expectation, and trust. Patient autonomy is exercised within a relation in which the physician’s epistemic authority, the patient’s values and concerns, and the responsibilities attached to clinical decisions all play a constitutive role. I argue that shared decision-making offers a useful framework for examining these issues, provided it is not reduced to a procedural ideal of information exchange. Read through relational autonomy and care ethics, shared decision-making allows us to ask how AI tools may affect the conditions under which clinical decisions become understandable, acceptable, and trustworthy for patients. The paper explores how AI decision support might be integrated into neurological care without weakening the attention, responsibility, and trust on which clinical judgment depends.