The complexities of patient-centred conversational artificial intelligence

Source

• Source name: arXiv
• Source kind: paper
• Authors / organization: João Matos, Olivia Buege, Donny Cheung, Gary S. Collins, Paula Dhiman, Nan Li, Bingyu Mao, Benjamin W. Nelson, Michail Ouroutzoglou, Paul Varghese, Jonathan Amar
• Venue: arXiv: cs.AI
• Published: 2026-07-09
• Collected: 2026-07-10
• URL: http://arxiv.org/abs/2607.08625v1
• PDF: https://arxiv.org/pdf/2607.08625v1
• DOI: UNKNOWN

Plain-Language Brief

Researchers evaluated the performance of large language models in patient-chatbot conversations and found significant variations in communication patterns and emotional expression among real patients. They developed a patient simulator to improve the realism of AI interactions and showed that communication style can impact triage outcomes. The study emphasizes the importance of accommodating diversity in communication for effective patient-centered conversational AI.

What the Document Says

• Consumer health chatbots using large language models (LLMs) are commonly used for symptom assessment.
• Chatbot development and evaluation often rely on simulated, cooperative patients.
• 2,053 real patient-chatbot conversations were analyzed to identify communication patterns and emotional expression variations among users.
• A patient simulator was developed to separately model clinical content, emotional state, conversational strategy, and communication style.
• The simulated conversations were evaluated by 15 human graders in a Turing-inspired realism assessment.
• Human graders achieved an accuracy of 55% in distinguishing between real and simulated conversations.
• Five distinct patient personae were used to evaluate the performance of four LLMs in urgency assessment.
• The study found that communication style can significantly impact triage outcomes.
• The authors suggest that patient-centered conversational AI should accommodate communication diversity to prevent underperformance and amplification of health disparities.

Key Concepts Mentioned

• consumer health chatbots
• large language models (LLMs)
• symptom assessment
• chatbot development
• patient simulator
• communication patterns
• emotional expression
• Turing-inspired evaluation
• realism
• urgency assessment
• patient personae
• triage outcomes
• patient-centered conversational AI

Claims Made by the Document

• Real patient communication patterns and emotional expressions vary widely.
• The developed patient simulator can realistically mimic human conversations.
• Communication style can significantly impact triage outcomes in patient-chatbot interactions.
• Patient-centered conversational AI must accommodate communication diversity to perform effectively.

Evidence Used

• Analysis of 2,053 real patient-chatbot conversations.
• Turing-inspired evaluation with 15 human graders.
• Five distinct patient personae used in clinician-graded cases.
• Evaluation of four LLMs in urgency assessment.

Methods / Approach

UNKNOWN

Limitations Stated by the Authors

UNKNOWN

Limitations Noted by the Agent

Agent interpretation — not from the source.

UNKNOWN

Open Questions

UNKNOWN

Notes for Future Q&A

May include interpretation. Labeled where it does.

UNKNOWN

Raw Abstract / Excerpt

Consumer-facing health chatbots powered by large language models (LLMs) are increasingly used for symptom assessment. However, chatbot development and evaluation often rely on cooperative, articulate, simulated patients. We analysed 2,053 real patient-chatbot conversations and found that communication patterns and expression of emotions vary widely across users. We developed a patient simulator that separately models clinical content, emotional state, conversational strategy, and communication style. In a Turing-inspired evaluation of realism with 15 human graders, simulated conversations were nearly indistinguishable from real ones, with human graders achieving an accuracy of 55%. We used five distinct patient personae, across 1,164 clinician-graded cases, to evaluate the performance of four LLMs in urgency assessment. We found that communication style can significantly alter triage outcomes. Patient-centred conversational artificial intelligence must accommodate communication diversity: systems designed for idealised, rather than realistic, interactions risk underperforming and amplifying health disparities when deployed in the real world.