Development and evaluation of AI chatbot tool for written communication training in self-care: Experiences of pharmacy students and faculty

Abstract

BACKGROUND: Effective communication is crucial in pharmacy practice, particularly in self-care counseling. As online pharmacies and chat-based consultations expand, training in digital written communication is increasingly important. Artificial intelligence (AI) systems based on large language models (LLMs) offer a structured and engaging environment to support skill development through conversational agents. This study explored the use of LLM-based chatbots to train pharmacy students in written synchronous communication for self-care consultations. METHODS: Three chatbot-simulated patients and an LLM-based feedback system were developed to reflect common self-care scenarios and provide communication-focused feedback. Fourteen pharmacy students and faculty interacted with the chatbots and shared their experiences through semi-structured interviews. Thematic analysis was used to identify patterns in the data. RESULTS: The analysis identified five main themes. Participants emphasized the authenticity of the simulated patient interactions, particularly their emotional realism. The AI-generated feedback was described as structured, detailed, and fair especially valued for its focus on communication skills. Faculty appreciated the consistency of the feedback and highlighted its added value to complement human assessment. Students discussed the cognitive and emotional demands of the experience, suggesting potential to tailor chatbot complexity to learners' needs. CONCLUSION: LLM-based chatbots represent a pedagogically grounded and scalable tool for developing pharmacy students' written communication skills in self-care consultations. This approach offers a foundation for building shared virtual patient infrastructures and integrating communication theory into digital education. It holds promise for broad implementation across pharmacy programs adapting to the demands of online and hybrid care.

Ähnliche Arbeiten