Evaluating ChatGPT-3.5 and Claude-2 in Answering and Explaining Conceptual Medical Physiology Multiple-Choice Questions.

Background Generative artificial intelligence (AI) systems such as ChatGPT-3.5 and Claude-2 may assist in explaining complex medical science topics. A few studies have shown that AI can solve complicated physiology problems that require critical thinking and analysis. However, further studies are required to validate the effectiveness of AI in answering conceptual multiple-choice questions (MCQs) in human physiology. Objective This study aimed to evaluate and compare the proficiency of ChatGPT-3.5 and Claude-2 in answering and explaining a curated set of MCQs in medical physiology. Methods In this cross-sectional study, a set of 55 MCQs from 10 competencies of medical physiology was purposefully constructed that required comprehension, problem-solving, and analytical skills to solve them. The MCQs and a structured prompt for response generation were presented to ChatGPT-3.5 and Claude-2. The explanations provided by both AI systems were documented in an Excel spreadsheet. All three authors subjected these explanations to a rating process using a scale of 0 to 3. A rating of 0 was assigned to an incorrect, 1 to a partially correct, 2 to a correct explanation with some aspects missing, and 3 to a perfectly correct explanation. Both AI models were evaluated for their ability to choose the correct answer (option) and provide clear and comprehensive explanations of the MCQs. The Mann-Whitney U test was used to compare AI responses. The Fleiss multi-rater kappa (κ) was used to determine the score agreement among the three raters. The statistical significance level was decided at P ≤ 0.05. Results Claude-2 answered 40 MCQs correctly, which was significantly higher than the 26 correct responses from ChatGPT-3.5. The rating distribution for the explanations generated by Claude-2 was significantly higher than that of ChatGPT-3.5. The κ values were 0.804 and 0.818 for Claude-2 and ChatGPT-3.5, respectively. Conclusion In terms of answering and elucidating conceptual MCQs in medical physiology, Claude-2 surpassed ChatGPT-3.5. However, accessing Claude-2 from India requires the use of a virtual private network, which may raise security concerns.

Agarwal M ，Goswami A ，Sharma P 《Cureus》

Analysing the Applicability of ChatGPT, Bard, and Bing to Generate Reasoning-Based Multiple-Choice Questions in Medical Physiology.

Background Artificial intelligence (AI) is evolving in the medical education system. ChatGPT, Google Bard, and Microsoft Bing are AI-based models that can solve problems in medical education. However, the applicability of AI to create reasoning-based multiple-choice questions (MCQs) in the field of medical physiology is yet to be explored. Objective We aimed to assess and compare the applicability of ChatGPT, Bard, and Bing in generating reasoning-based MCQs for MBBS (Bachelor of Medicine, Bachelor of Surgery) undergraduate students on the subject of physiology. Methods The National Medical Commission of India has developed an 11-module physiology curriculum with various competencies. Two physiologists independently chose a competency from each module. The third physiologist prompted all three AIs to generate five MCQs for each chosen competency. The two physiologists who provided the competencies rated the MCQs generated by the AIs on a scale of 0-3 for validity, difficulty, and reasoning ability required to answer them. We analyzed the average of the two scores using the Kruskal-Wallis test to compare the distribution across the total and module-wise responses, followed by a post-hoc test for pairwise comparisons. We used Cohen's Kappa (Κ) to assess the agreement in scores between the two raters. We expressed the data as a median with an interquartile range. We determined their statistical significance by a p-value <0.05. Results ChatGPT and Bard generated 110 MCQs for the chosen competencies. However, Bing provided only 100 MCQs as it failed to generate them for two competencies. The validity of the MCQs was rated as 3 (3-3) for ChatGPT, 3 (1.5-3) for Bard, and 3 (1.5-3) for Bing, showing a significant difference (p<0.001) among the models. The difficulty of the MCQs was rated as 1 (0-1) for ChatGPT, 1 (1-2) for Bard, and 1 (1-2) for Bing, with a significant difference (p=0.006). The required reasoning ability to answer the MCQs was rated as 1 (1-2) for ChatGPT, 1 (1-2) for Bard, and 1 (1-2) for Bing, with no significant difference (p=0.235). K was ≥ 0.8 for all three parameters across all three AI models. Conclusion AI still needs to evolve to generate reasoning-based MCQs in medical physiology. ChatGPT, Bard, and Bing showed certain limitations. Bing generated significantly least valid MCQs, while ChatGPT generated significantly least difficult MCQs.

Agarwal M ，Sharma P ，Goswami A 《Cureus》

Evaluating ChatGPT's Ability to Solve Higher-Order Questions on the Competency-Based Medical Education Curriculum in Medical Biochemistry.

Background Healthcare-related artificial intelligence (AI) is developing. The capacity of the system to carry out sophisticated cognitive processes, such as problem-solving, decision-making, reasoning, and perceiving, is referred to as higher cognitive thinking in AI. This kind of thinking requires more than just processing facts; it also entails comprehending and working with abstract ideas, evaluating and applying data relevant to the context, and producing new insights based on prior learning and experience. ChatGPT is an artificial intelligence-based conversational software that can engage with people to answer questions and uses natural language processing models. The platform has created a worldwide buzz and keeps setting an ongoing trend in solving many complex problems in various dimensions. Nevertheless, ChatGPT's capacity to correctly respond to queries requiring higher-level thinking in medical biochemistry has not yet been investigated. So, this research aimed to evaluate ChatGPT's aptitude for responding to higher-order questions on medical biochemistry. Objective In this study, our objective was to determine whether ChatGPT can address higher-order problems related to medical biochemistry.​​​​​​ Methods​​​ This cross-sectional study was done online by conversing with the current version of ChatGPT (14 March 2023, which is presently free for registered users). It was presented with 200 medical biochemistry reasoning questions that require higher-order thinking. These questions were randomly picked from the institution's question bank and classified according to the Competency-Based Medical Education (CBME) curriculum's competency modules. The responses were collected and archived for subsequent research. Two expert biochemistry academicians examined the replies on a zero to five scale. The score's accuracy was determined by a one-sample Wilcoxon signed rank test using hypothetical values. Result The AI software answered 200 questions requiring higher-order thinking with a median score of 4.0 (Q1=3.50, Q3=4.50). Using a single sample Wilcoxon signed rank test, the result was less than the hypothetical maximum of five (p=0.001) and comparable to four (p=0.16). There was no difference in the replies to questions from different CBME modules in medical biochemistry (Kruskal-Wallis p=0.39). The inter-rater reliability of the scores scored by two biochemistry faculty members was outstanding (ICC=0.926 (95% CI: 0.814-0.971); F=19; p=0.001)​​​​​​ Conclusion The results of this research indicate that ChatGPT has the potential to be a successful tool for answering questions requiring higher-order thinking in medical biochemistry, with a median score of four out of five. However, continuous training and development with data of recent advances are essential to improve performance and make it functional for the ever-growing field of academic medical usage.

Ghosh A ，Bir A 《Cureus》

Performance of ChatGPT on Nursing Licensure Examinations in the United States and China: Cross-Sectional Study.

The creation of large language models (LLMs) such as ChatGPT is an important step in the development of artificial intelligence, which shows great potential in medical education due to its powerful language understanding and generative capabilities. The purpose of this study was to quantitatively evaluate and comprehensively analyze ChatGPT's performance in handling questions for the National Nursing Licensure Examination (NNLE) in China and the United States, including the National Council Licensure Examination for Registered Nurses (NCLEX-RN) and the NNLE. This study aims to examine how well LLMs respond to the NCLEX-RN and the NNLE multiple-choice questions (MCQs) in various language inputs. To evaluate whether LLMs can be used as multilingual learning assistance for nursing, and to assess whether they possess a repository of professional knowledge applicable to clinical nursing practice. First, we compiled 150 NCLEX-RN Practical MCQs, 240 NNLE Theoretical MCQs, and 240 NNLE Practical MCQs. Then, the translation function of ChatGPT 3.5 was used to translate NCLEX-RN questions from English to Chinese and NNLE questions from Chinese to English. Finally, the original version and the translated version of the MCQs were inputted into ChatGPT 4.0, ChatGPT 3.5, and Google Bard. Different LLMs were compared according to the accuracy rate, and the differences between different language inputs were compared. The accuracy rates of ChatGPT 4.0 for NCLEX-RN practical questions and Chinese-translated NCLEX-RN practical questions were 88.7% (133/150) and 79.3% (119/150), respectively. Despite the statistical significance of the difference (P=.03), the correct rate was generally satisfactory. Around 71.9% (169/235) of NNLE Theoretical MCQs and 69.1% (161/233) of NNLE Practical MCQs were correctly answered by ChatGPT 4.0. The accuracy of ChatGPT 4.0 in processing NNLE Theoretical MCQs and NNLE Practical MCQs translated into English was 71.5% (168/235; P=.92) and 67.8% (158/233; P=.77), respectively, and there was no statistically significant difference between the results of text input in different languages. ChatGPT 3.5 (NCLEX-RN P=.003, NNLE Theoretical P<.001, NNLE Practical P=.12) and Google Bard (NCLEX-RN P<.001, NNLE Theoretical P<.001, NNLE Practical P<.001) had lower accuracy rates for nursing-related MCQs than ChatGPT 4.0 in English input. English accuracy was higher when compared with ChatGPT 3.5's Chinese input, and the difference was statistically significant (NCLEX-RN P=.02, NNLE Practical P=.02). Whether submitted in Chinese or English, the MCQs from the NCLEX-RN and NNLE demonstrated that ChatGPT 4.0 had the highest number of unique correct responses and the lowest number of unique incorrect responses among the 3 LLMs. This study, focusing on 618 nursing MCQs including NCLEX-RN and NNLE exams, found that ChatGPT 4.0 outperformed ChatGPT 3.5 and Google Bard in accuracy. It excelled in processing English and Chinese inputs, underscoring its potential as a valuable tool in nursing education and clinical decision-making.

Wu Z ，Gan W ，Xue Z ，Ni Z ，Zheng X ，Zhang Y ... -  《-》

Can ChatGPT generate practice question explanations for medical students, a new faculty teaching tool?

Tong L ，Wang J ，Rapaka S ，Garg PS ... -  《-》