Paper Type
Complete
Paper Number
PACIS2025-1835
Description
Effective physician-patient communication in online medical consultations (OMCs) is essential for diagnostic accuracy, patient satisfaction, and platform sustainability. However, current assessments often rely on patient feedback or numerical metrics, offering limited insight into the role of language. This study introduces a linguistically informed, interpretable modeling–based framework for identifying communication quality in OMCs, integrating lexical, syntactic, semantic, and pragmatic features. Results reveal a key trade-off: while medical professionalism enhances perceived physician attitude, it impairs readability. Besides, Lexical simplicity and numerical specificity improve both attitude and clarity, whereas excessive diversity and verbosity detract from communication effectiveness. Syntactic analysis highlights the benefits of declarative structures and the cognitive burden of syntactic complexity. Pragmatic cues, though less impactful, offer relational value. Our work advances computational linguistics in telehealth by offering actionable insights for physicians to refine language use and supporting platform managers in developing intelligent tools to promote effective, accessible communication.
Recommended Citation
Liu, Haoran; Gao, Gege; and Wang, Liuan, "Identifying Linguistics-Based Physician-Patient Communication Quality in Online Medical Consultations through Interpretable Machine Learning" (2025). PACIS 2025 Proceedings. 20.
https://aisel.aisnet.org/pacis2025/ishealthcare/ishealthcare/20
Identifying Linguistics-Based Physician-Patient Communication Quality in Online Medical Consultations through Interpretable Machine Learning
Effective physician-patient communication in online medical consultations (OMCs) is essential for diagnostic accuracy, patient satisfaction, and platform sustainability. However, current assessments often rely on patient feedback or numerical metrics, offering limited insight into the role of language. This study introduces a linguistically informed, interpretable modeling–based framework for identifying communication quality in OMCs, integrating lexical, syntactic, semantic, and pragmatic features. Results reveal a key trade-off: while medical professionalism enhances perceived physician attitude, it impairs readability. Besides, Lexical simplicity and numerical specificity improve both attitude and clarity, whereas excessive diversity and verbosity detract from communication effectiveness. Syntactic analysis highlights the benefits of declarative structures and the cognitive burden of syntactic complexity. Pragmatic cues, though less impactful, offer relational value. Our work advances computational linguistics in telehealth by offering actionable insights for physicians to refine language use and supporting platform managers in developing intelligent tools to promote effective, accessible communication.
Comments
Healthcare