CARECOS study: Medical students’ empathy as assessed with the CARE measure by examiners versus standardized patients during a formative Objective and Structured Clinical Examination (OSCE) station

To assess the Consultation And Relational Empathy (CARE) measure as a tool for examiners to assess medical students’ empathy during Objective and Structured Clinical Examinations (OSCEs), as the best tool for assessing empathy during OSCEs remains unknown. We first assessed the psychometric properties of the CARE measure, completed simultaneously by examiners and standardized patients (SP, either teachers - SP_teacher - or civil society members - SP_{civil society}), for each student, at the end of an OSCE station. We then assessed the qualitative/quantitative agreement between examiners and SP. We included 129 students, distributed in eight groups, four groups for each SP type. The CARE measure showed satisfactory psychometric properties in the context of the study but moderate, and even poor inter-rater reliability for some items. Considering paired observations, examiners scored lower than SPs (<i>p</i> &lt; 0.001) regardless of the SP type. However, the difference in score was greater when the SP was a SP_teacher rather than a SP_{civil society} (<i>p</i> &lt; 0.01). Despite acceptable psychometric properties, inter-rater reliability of the CARE measure between examiners and SP was unsatisfactory. The choice of examiner as well as the type of SP seems critical to ensure a fair measure of empathy during OSCEs.

鉴于目前学界仍未明确客观结构化临床考试（Objective and Structured Clinical Examinations, OSCEs）场景下最优的共情评估工具，本研究旨在评估「会诊与关系共情量表（Consultation And Relational Empathy, CARE）」作为考官在OSCE中评估医学生共情能力工具的有效性。我们首先在单个OSCE站点考核结束后，针对每名学生，由考官与标准化患者（standardized patients, SP，分为教师型SP_teacher与社会成员型SP_{civil society}两类）同步完成CARE量表填写，以此评估该量表的心理测量学特性；随后我们评估了考官与标准化患者之间的定性与定量一致性水平。本研究共纳入129名医学生，分为8个组别，两类标准化患者各对应4组。研究结果显示，CARE量表在本研究场景下表现出良好的心理测量学特性，但部分条目存在中等甚至较差的评分者间信度。配对观察结果表明，无论标准化患者的类型如何，考官的评分均显著低于标准化患者（*p* < 0.001）。不过，当标准化患者为教师型SP时，考官与标准化患者的评分差值较社会成员型SP更为显著（*p* < 0.01）。尽管CARE量表具备可接受的心理测量学特性，但其在考官与标准化患者间的评分者间信度仍未达到理想标准。本研究提示，考官的选择以及标准化患者的类型，对于确保OSCE场景下共情能力评估的公平性至关重要。