Search

We aimed to determine the quality of AI-generated (ChatGPT-4 and Claude 3) Script Concordance Test (SCT) items through an expert panel. We generated SCT items on abdominal radiology using a complex prompt in large language model (LLM) chatbots (ChatGPT-4 and Claude 3 (Sonnet) in April 2024) and evaluated the items’ quality through an expert panel of 16 radiologists. Expert panel, which was blind to the origin of the items provided without modifications, independently answered each item and assessed them using 12 quality indicators. Data analysis included descriptive statistics, bar charts to compare responses against accepted forms, and a heatmap to show performance in terms of the quality indicators. SCT items generated by chatbots assess clinical reasoning rather than only factual recall (ChatGPT: 92.50%, Claude: 85.00%). The heatmap indicated that the items were generally acceptable, with most responses favorable across quality indicators (ChatGPT: 71.77%, Claude: 64.23%). The comparison of the bar charts with acceptable and unacceptable forms revealed that 73.33% and 53.33% of the questions in the items can be considered acceptable, respectively, for ChatGPT and Claude. The use of LLMs to generate SCT items can be helpful for medical educators by reducing the required time and effort. Although the prompt provides a good starting point, it remains crucial to review and revise AI-generated SCT items before educational use. The prompt and the custom GPT, “Script Concordance Test Generator”, available at https://chatgpt.com/g/g-RlzW5xdc1-script-concordance-test-generator, can streamline SCT item development.

Medical education always evolves to incorporate more tools for specific needs in assessing clinical reasoning skills. Among these tools, Script Concordance Test (SCT) has a particular importance due to its focus on assessing decision-making in uncertain clinical situations. However, development of SCT items is effortful. Artificial intelligence tools, such as large language models, offer significant benefits. These models are already used for generating multiple-choice questions, and their use in generating SCTs offers great promise. However, this requires well-designed prompts to generate SCTs. This article proposes a generic prompt for the ChatGPT-4, Claude 3, Llama 3, and ChatGPT-4o large language model chatbots to generate SCTs, which can be tailored to various fields of medicine and different stages of medical education. It can help to streamline the development process of SCTs. Initial findings are promising, and there is a need for generating SCTs using large language models and conducting research to assess the quality of SCTs.

This study introduces ContExtended Questions (CEQ), which is a tool both to teach and assess clinical reasoning particularly in the preclinical years, and the web-based program to implement. CEQ consists of text-based case-based multiple-choice questions that provide patient data in a fixed and predetermined sequence. It enables the examinees to develop and reshape their illness scripts by using feedback after every question. Feedback operates to transform the examinee’s failure into a “productive failure”. The preliminary results of the randomized controlled experiment of teaching clinical reasoning to preclinical students through CEQ is quite satisfactory. In the medical education literature, this would be the first time that students, who have no or very limited clinical experience, developed their illness scripts just by taking formative multiple-choice tests. The approach would be named “test-only learning”. The complete results of the experiment and then more experiments in other contexts and domains are necessary to establish a more powerful assessment tool and software. Furthermore, by changing the content of the questions, it is possible to use CEQ in every period of medical education and health professions education.