AI in Medicine: Decision-Making and Relational Vulnerability of Healthcare Professionals

Abstract

The integration of artificial intelligence (AI) in medicine is transforming clinical decision-making and the doctor-patient relationship. While AI promises enhanced diagnostic accuracy and efficiency, it also raises concerns about physicians’ vulnerability, particularly regarding decision-making autonomy and the relational dimension of care. This study explores the perceived vulnerabilities associated with AI adoption in medical practice, focusing on decision-making and relational dynamics. Using a qualitative methodology, we conducted 20 semi-structured interviews with physicians from various specialties (radiology, genetics, internal medicine, oncology, etc.), analyzing their perceptions of AI through an interpretative phenomenological analysis (IPA). Our findings reveal two key forms of vulnerability: Decision-making vulnerability—Physicians express concerns about potential loss of clinical autonomy due to AI’s increasing influence over medical judgments. Many fear that AI could lead to over-standardization of medical practice, diminish their expertise, and foster excessive reliance on algorithmic recommendations. Relational vulnerability—This form of vulnerability concerns the process of interaction within the doctor–patient relationship. The delegation of diagnostic or communicative tasks to AI is perceived as potentially dehumanizing, reducing empathy and emotional engagement, and eroding trust between physician and patient. Therapeutic vulnerability—In contrast, therapeutic vulnerability relates to patient outcomes and experiences of care. Patients may feel disoriented or excluded from decision-making processes, especially when confronted with opaque algorithmic recommendations. Such experiences can lead to disengagement from care, passive adherence, and a weakening of the therapeutic alliance. These findings highlight the paradox of AI in healthcare: while it enhances precision and efficiency, it also generates resistance due to its perceived threats to physicians’ roles and medical humanism. The study underscores the need for a balanced AI integration strategy that preserves medical autonomy and safeguards the fundamental relational aspects of healthcare.

Ähnliche Arbeiten