User preferences and trust in hypothetical analog, digitalized and AI-based medical consultation scenarios: An online discrete choice survey

Current developments in telemedicine and artificial intelligence (AI) are significantly impacting doctor-patient interactions. This study examined the interacting role of individual traits with different levels of digitalization in participants' user preferences and trust within hypothetical medical scenarios. Specifically, preferences and trust levels towards various digitalized and analog formats, such as face-to-face interactions, video calls, written exchanges with a doctor or chatbot, or conversations with AI avatars were compared using standard scenarios of varying health risks and potentially embarrassing content. In an online discrete choice experiment, 1009 participants rated hypothetical scenarios of varying medical concerns regarding their preferred conversation format and trust. User preference (n = 2018 observations) and trust (n = 9880 observations) were predicted using two multilevel models. Higher perceived efficiency of digital conversation formats predicted user preference for digitalized formats. However, users' preference for digitalized formats was generally lower compared to face-to-face interactions, especially when receiving bad news. The level of digitalization was negatively associated with trust, which was lower for consultations that involved receiving bad news or discussing potentially embarrassing content compared to good news. Trust ratings varied depending on the conversation topic. When comparing analog and digitalized medical consultation scenarios, digitalized medical consultations are not equally suited for every medical consultation. Participants preferred personal contact, particularly when bad news needed to be communicated. Additionally, trust in the doctor significantly varies depending on the topic of conversation.