Patient information needs for transparent and trustworthy cardiovascular artificial intelligence: A qualitative study

Area of Science:

• Health Informatics
• Artificial Intelligence in Medicine
• Patient Engagement

Background:

• Healthcare systems are increasingly integrating artificial intelligence (AI).
• Ensuring transparency and trustworthiness of AI in patient care is a growing concern.
• Patient perspectives on information needs for AI trust are under-explored.

Purpose of the Study:

• To identify patient information needs for trusting and adopting AI in healthcare.
• To explore patient views on AI in cardiovascular care specifically.
• To inform strategies for transparent AI implementation.

Main Methods:

• Conducted three asynchronous online focus groups with 42 patients in the US.
• Focused on AI use in cardiovascular care settings.
• Analyzed data using a rapid content analysis approach.

Main Results:

• Patients require specific information on AI tools, oversight, and the healthcare experience.
• Key areas include information delivery, disclosure, consent, and physician use of AI.
• These factors are crucial for calibrating patient trust in healthcare AI.

Conclusions:

• Identifying patient information needs is essential for building trust in healthcare AI.
• Patient-centered engagement is vital for developing AI documentation and communication strategies.
• Addressing patient information needs promotes transparent and trustworthy AI adoption in clinical settings.