Do patients prefer a human doctor, artificial intelligence, or a blend, and is this preference dependent on medical discipline? Empirical evidence and implications for medical practice

Today the doctor-patient relationship typically takes place in a face-to-face setting. However, with the advent of artificial intelligence (AI) systems, two further interaction scenarios are possible: an AI system supports the doctor's decision regarding diagnosis and/or treatment while interacting with the patient, or an AI system could even substitute the doctor and hence a patient interacts with a chatbot (i.e., a machine) alone. Against this background, we report on an online experiment in which we analyzed data from <i>N</i> = 1,183 people. The data was collected in German-speaking countries (Germany, Austria, Switzerland). The participants were asked to imagine they had been suffering from medical conditions of unknown origin for some time and that they were therefore visiting a health center to seek advice from a doctor. We developed descriptions of patient-doctor interactions (referred to as vignettes), thereby manipulating the patient's interaction partner: (i) human doctor, (ii) human doctor with an AI system, and (iii) an AI system only (i.e., chatbot). Furthermore, we manipulated medical discipline: (i) cardiology, (ii) orthopedics, (iii) dermatology, and (iv) psychiatry. Based on this 3 × 4 experimental within-subjects design, our results indicate that people prefer a human doctor, followed by a human doctor with an AI system, and an AI system alone came in last place. Specifically, based on these 12 hypothetical interaction situations, we found a significant main effect of a patient's interaction partner on trust, distrust, perceived privacy invasion, information disclosure, treatment adherence, and satisfaction. Moreover, perceptions of trust, distrust, and privacy invasion predicted information disclosure, treatment adherence, and satisfaction as a function of interaction partner and medical discipline. We found that the situation in psychiatry is different from the other three disciplines. Specifically, the six outcome variables differed strongly between psychiatry and the three other disciplines in the "human doctor with an AI system" condition, while this effect was not that strong in the other conditions (human doctor, chatbot). These findings have important implications for the use of AI in medical care and in the interaction between patients and their doctors.