Conversational Capacity in Triadic Telemedicine: A Conceptual Framework and Evidence Synthesis

Area of Science:

• Medical Informatics
• Human-Computer Interaction
• Clinical Communication

Background:

• Telemedicine often involves a third party beyond the clinician and patient, including AI.
• Current models lack a framework to distinguish genuine facilitation from instrumental use of third parties.
• This ambiguity hinders research comparability and informed deployment decisions.

Purpose of the Study:

• To develop a technology-neutral conceptual framework and classification model for triadic telemedicine interactions.
• To establish criteria for identifying genuine third-party facilitation in clinical encounters.
• To guide responsible deployment of AI in telemedicine.

Main Methods:

• A narrative synthesis of literature on triadic clinical communication, shared decision-making, and AI-mediated interactions.
• Development of a conceptual framework based on 'conversational capacity' and its functions (interpret, translate, advocate, adapt).
• Creation of a classification model including a five-level Conversational Capacity Spectrum and a Clinical Situation Matrix.

Main Results:

• Conversational capacity, defined by interpret, translate, advocate, and adapt functions, is proposed as the criterion for triadic facilitation.
• A framework distinguishes genuine triadic architectures (Modes A/B) from augmented dyadic models (Mode C).
• AI currently struggles with advocacy, limiting its facilitation role; an equity paradox exists where vulnerable populations may receive less human support.

Conclusions:

• Triadic telemedicine should be defined by "conversational capacity," not just third-party presence.
• A gradable, measurable construct replacing the binary distinction is needed for research and AI deployment.
• This framework enables cumulative research and responsible AI integration in clinical consultations.