Beyond algorithms: A qualitative study on AI as a relational space for organ transplant patients

Abstract

Background: Organ transplantation is not a finite medical intervention but a lifelong process requiring continuous clinical monitoring, strict medication adherence, and sustained vigilance toward bodily changes. For transplant recipients, daily life is shaped by routine medical tasks and the constant fear of graft rejection, necessitating ongoing interaction with healthcare systems. However, patients often face significant barriers, including long waiting times, limited appointment availability, and brief consultations. For those whose bodily sensations oscillate between normal variations and potential complications, the inability to consult a physician quickly generates anxiety and self-doubt. In this context, AI-based conversational tools have emerged as alternative sources of medical information and emotional reassurance. While literature often focuses on AI efficiency, less attention is paid to how patients interpret, personalize, and emotionally engage with these tools. Aim: This study examines how and why organ transplant patients use AI-based conversational tools as doctor-like figures, and how these practices reshape understandings of medical authority, trust, and care. Methods: This study employed a qualitative research design informed by an ethnographic sensibility. Data were collected through semi-structured, in-depth interviews with 24 organ transplant patients (kidney and liver) who had undergone surgery at least one year prior and used AI for health-related purposes. Findings: Thematic analysis yielded five interrelated themes: (1) Access without burden, capturing participants’ experiences of AI as immediately available and non-judgmental; (2) AI as a Proxy Clinician, reflecting how patients mobilized AI to interpret test results, seek reassurance, and prepare for clinical encounters without replacing medical authority; (3) Being taken seriously without being judged, highlighting the emotional relief derived from articulating concerns outside time-pressured clinical interactions; (4) Naming, familiarity, and trust, illustrating how anthropomorphization transformed AI into a relatable and dependable companion; and (5) Redistribution of care responsibility, demonstrating how AI supported patients in navigating intensified expectations of self-management in post-transplant life. Conclusion: The findings suggest that the use of “AI doctors” reflects structural constraints within healthcare systems and unmet relational needs in clinical care.