AI models might advance human welfare in unprecedented ways, but progress will not occur without substantial risks that will have to be managed.

Clinical decision supports create ethically complex risks and need to align patients’ and caregivers’ professed values.

Nudges can improve patient outcomes, but decisions need to be made about where they fit in the flow of health service delivery.

Dr Mitesh Patel joins us on this episode of Ethics Talk to discuss nudges, how they can be used effectively in health care, and how to identify and avoid the potential ethical pitfalls of guiding behavior.

Patients can now easily view their health records, so clinicians must consider a reader’s interpretation of how they convey sensitive personal health information. What might this mean for ethics consultants?

Perhaps machine learning systems trained on patients’ electronic health records and social media footprints could be used as decision aids when patients lack capacity or face overwhelming decisions.

Clinical decision support systems leverage data generated in the course of standard clinical care to improve clinical practice. They need to ensure privacy and quality of patients’ data, but must also allow queries of electronic health records.

Questions about data privacy need to be addressed when research institutions negotiate with companies developing mobile health applications. Commercial terms of use and data sharing notifications should be reviewed before use in human subject research settings.

Social and behavioral data contained in electronic health records are essential for studying health disparities. Can researchers avoid bias when collecting, analyzing, and using such data?

Fair distribution demands new strategies for engaging patients in co-creation.