Abstract:

On online health platforms, physicians provide guidance for patients using written language, and it is unlikely that all guidance will be fully accepted by patients with diverse diseases. Based on the language expectancy theory, this paper identifies the medical professional language and interpersonal language of physicians under the online physician-patient interaction context and explores their effects on patients’ satisfaction perception and reward behavior, as well as the moderating role of patients’ disease severity. Approximately 52000 physician-patient interactions are collected from a leading health platform in China and processed by the text-mining, machine learning, and other methods. The empirical results indicate that the professional language of physicians has a positive effect on both the satisfaction perception and reward behavior of patients, while interpersonal language has a negative impact. Patients’ disease severity positively moderates the effects of the medical professional language, but it negatively moderates the effects of the interpersonal language. These findings provide guidance for physicians on how to effectively use written language in interactions to meet the expectations of patients with varying disease severities.

Key words:

online health platforms, physician-patient interaction, physician language, patient expectation, machine learning

摘要：

医生在健康平台上使用文字语言为患者提供疾病指导信息，而医生语言信息不太可能被患有不同疾病的患者全部接受。基于语言期望理论，识别在线医患交互背景下医生医疗专业性和人际性语言，并探究其对患者满意感知和回报行为的影响以及患者疾病严重程度的调节作用。以国内一家领先的健康平台为研究对象，采集约52 000条在线医患交互数据，采用文本挖掘、机器学习等方法处理数据。实证结果显示：医生医疗专业性语言正向影响患者满意感知和回报行为，而人际性语言负向影响患者满意感知和回报行为；患者疾病严重程度正向调节医生医疗专业性语言的影响，而负向调节医生人际性语言的影响。研究结果为医生如何在交互中有效使用文字语言满足具有不同疾病严重程度的患者期望提供指导。

关键词:

在线健康平台, 医患交互, 医生语言, 患者期望, 机器学习