Patient and Family Perspectives on Generative AI Tools in Rare Diseases: Exploratory Mixed Methods Online Survey

Editorial reading guide · Collaboration

How can we bring an AI-generated rare-disease question into a productive conversation with our child’s care team?

Shows the communication gap that can develop when families use GenAI to interpret an uncertain condition. Across the mixed sample, 63 respondents had used GenAI and 90 had not discussed AI-derived information with clinicians. Reported diagnostic or decision influence was self-reported, making this a conversation about sharing questions, uncertainty, and source checking rather than a validation of diagnosis by chatbot.

Study context

115 respondents, comprising 74 parents/guardians and 41 adult rare-disease patients, recruited largely through US rare-disease communities

Age / stage: Parents/guardians of children with rare diseases plus adult patients; child ages not specified

Read with three questions

1. Which tasks did respondents bring to GenAI, and which could influence a medical decision?
2. What might explain the gap between using AI and telling a clinician about it?
3. How do respondent-reported help or harm differ from independently verified diagnostic or treatment outcomes?

Keep the limits in view

This convenience survey mixed adult patients and parents; aggregate results are not parent-only estimates. Respondents were highly educated and predominantly US-based, and interest in AI may have affected participation. Reported diagnosis contributions, decision effects, and harms were not clinically validated; open responses were brief and subgroup comparisons limited.