JMIR Publications

ABSTRACT

Background:

The introduction of artificial intelligence (AI) into healthcare has sparked discussions and raised numerous questions about its potential impact. Health stakeholders, particularly patients, will be at the forefront of interacting with and being impacted by AI technologies. Given the ethical importance of patient-centered healthcare, patients must navigate how they engage with AI. However, integrating AI into clinical practice brings opportunities and challenges, particularly in shared decision-making and ensuring patients remain active participants in their care. The extent to which AI-supported interventions empower or undermine patients’ participation depends mainly on how these technologies are envisioned and integrated into practice. As healthcare aims to become increasingly digitized, fostering patients’ active participation is more crucial than ever to ensure that AI enhances, rather than detracts from, the patient experience.

Objective:

This study explores how patients and professionals envision patients' interaction with AI and their preferences concerning participation in decision-making about AI-supported care.

Methods:

Qualitative semi-structured interviews were conducted with 21 patients and 21 participants from different disciplines who were exposed to medical AI in their profession. The data was analyzed using reflective thematic analysis in an interactive and iterative process.

Results:

We created three themes to describe how patients and professionals envision patients interacting with AI. The first theme explored the vision of AI as an unavoidable and potentially harmful force of change for healthcare delivery. The second theme focused on how patients perceive limitations in their capabilities that would prevent them from having meaningful participation in AI-supported care. The third theme explores the adaptive responses patients see as feasible, such as relying on those considered experts or utilizing value judgments as the basis for their decision-making.

Conclusions:

External and internal preconceptions shape how patients and professionals perceive the patient's role and participation in AI-supported care. Patients have internally accepted that AI's complexity and unavoidability may limit their ability to engage actively, feeling they have little influence over AI’s development. While patients expressed their desire to rely on doctors or considered accepting or rejecting AI, these adaptive strategies risk placing them in a disempowering role where they passively accept AI “as it is”. These strategies may reduce patients' involvement to that of passive care receivers. Current adaptive responses might be insufficient to position them at the center of their care without adequate education about their rights and possibilities.