From Principles to Practice: A Design Framework for Patient-Centered Conversational AI in Digital Medicine

Abstract

Patient-facing conversational artificial intelligence (AI) is expanding in digital medicine, yet development remains fragmented due to the lack of standardized design guidelines for conversational interfaces. Current frameworks emphasize ethics, governance, and technical risk management, but provide little actionable guidance for patient-facing design, leaving systems prone to usability, trust, safety, and accessibility gaps. This paper proposes a design framework for patient-centered conversational AI in digital medicine that addresses these challenges. The framework comprises three stages: (1) conducting user research through patient personas, literacy assessment, and device context; (2) designing conversations with plain language, accessibility, trust-building, and privacy-conscious personalization; and (3) testing effectiveness through metrics such as task completion, comprehension, safety, and accessibility compliance. By bridging governance principles with patient-centered design practices, the framework provides a structured pathway for developing conversational AI systems that are both compliant and clinically meaningful, improving patients’ comprehension, trust, and capacity for self-management in digital medicine.

Ähnliche Arbeiten