Oliven A, Nave R, Gilad D, Barch A
Faculty of Medicine, Technion, Haifa, Israel.
Stud Health Technol Inform. 2011;169:233-7.
Structured Clinical Examinations (OSCE) are resource intensive, not practical as teaching tools, and their reliability depends on evaluators. Computer-based case simulations ("virtual patients", VP) have been advocated as useful and reliable tools for teaching clinical skills and evaluating competence. We have developed an internet-based VP system designed both for practice and assessment of medical students. The system uses interactive dialogue with natural language processing, and is designed for history taking, evaluation of physical examination, including recognition of visual findings and heart and lung sounds, and ordering lab-and imaging tests. The system includes a practice modality that provides feedback, and a computerized OSCE. The reliability of our system was assessed over the last three years by comparing the clinical competence of medical students in similar VP and human OSCE. A total of 262 students were evaluated with both exam modalities. The correlation between the two exams scores was highly significant (p<0.001). Alpha Cronbach for the computerized exam was 0.82-0.89 in the 3 years, and was substantially higher than that of the conventional OSCE each year. We conclude that a computerized VP OSCE is a reliable examination tool, with the advantage of providing also a training modality.
结构化临床考试（OSCE）资源消耗大，作为教学工具并不实用，其可靠性取决于评估者。基于计算机的病例模拟（“虚拟患者”，VP）已被倡导为用于教授临床技能和评估能力的有用且可靠的工具。我们开发了一个基于互联网的VP系统，旨在供医学生进行实践和评估。该系统利用自然语言处理进行交互式对话，设计用于病史采集、体格检查评估，包括识别视觉检查结果以及心肺声音，还可用于开具实验室和影像学检查医嘱。该系统包括提供反馈的实践模式和计算机化的OSCE。在过去三年中，通过比较医学生在类似的VP和人工OSCE中的临床能力，对我们系统的可靠性进行了评估。共有262名学生接受了这两种考试方式的评估。两次考试成绩之间的相关性非常显著（p<0.001）。在这三年中，计算机化考试的克朗巴哈系数为0.82 - 0.89，且每年都大大高于传统OSCE的系数。我们得出结论，计算机化的VP OSCE是一种可靠的考试工具，其优势还在于提供了一种培训模式。
