Teaching and learning about making a clinical diagnosis have been considered essential in medical education because it is a major determinant of the quality of patient care. Teaching medical decision making should target not only the diagnosis itself, but also the reasoning process, which involves the application of medical knowledge, knowledge structure and problem-solving strategies.However, the clinical thinking process involves unconscious cognitive capacities, which makes teaching and learning in this field challenging. 1 The traditional pedagogies of clinical reasoning are lectures, group discussions, bedside and ambulatory encounters, and case studies. Standardised patients, simulation and technology-applied tools have been lately introduced to enhance learning.The above educational strategies, however, may have disadvantages, such as time cost, inadequate experience of patient encounters and a lack of representation of thought processes in the reasoning process. With advances in technology, the complex system of professional expertise can be extracted and recorded for further analyses.
| WHAT WA S TRIED?A virtual patient (VP) with artificial intelligence and natural language processing was developed to trigger the problem-solving activities in a simulated clinical encounter. The student played the role of a physician, who faced a VP with medical problems. The VP 'recognises and understands' student's verbal questions and reacts by voice and text in real time. Each decision made by the student instantly modifies the response of the VP. Texts or multimedia appeared in response to the requests for physical examination, laboratory tests or image studies.Based on the information obtained from the VP, students established their clinical reasoning and concluded with a list of diagnoses. The students were then asked to categorise the information they just gathered based on their preliminary diagnoses. To enhance learning, the system displayed the schemata derived from the think-aloud interviews to experts. Finally, the system generated players' performance scores and provided feedback based on the standards predefined by experts. The field testing of validation had been delivered to 112 medical students in Kaohsiung Medical University and E-Da Hospital in Taiwan.
| WHAT LE SSON S WERE LE ARNED?The virtual patient-based system gives students the opportunities of early clinical exposure to patients. It allows repetitive practices with immediate feedback from the system and tutors in a safe environment. This method facilitates students self-directed learning, at any time and in any place, and helps them formulate learning objectives and improve the quality of their knowledge structure. The questionnaire data from the 112 medical students showed high appreciation of enhancing their competence in clinical problem solving. Up to 80% of users reported an increase in their confidence and proficiency in dealing with real-life patients, and a more organised pattern of thoughts. The simulation technology successfully maximi...