Can ChatGPT pass the Turkish Orthopedics and Traumatology Board Examination? Turkish orthopedic surgeons versus artificial intelligence

Abstract

BACKGROUND: Artificial intelligence has been shown to achieve successful outcomes in various orthopedic qualification examinations worldwide. This study aims to assess the performance of ChatGPT in the written section of the Turkish Orthopedics and Traumatology Board Examination, compare its results with those of candidates who took the exam, and determine whether ChatGPT is sufficient to achieve a passing score. METHODS: This retrospective observational study evaluated whether ChatGPT achieved a passing grade on 400 publicly available questions from the Turkish orthopedics qualification exam over the past four years. ChatGPT's performance was compared with the mean scores of the candidates who took the exam. RESULTS: A total of 627 candidates participated in the four exams included in the study, of whom 292 (46.5%) passed. ChatGPT received higher scores than 619 (98.7%) of the candidates. In all exams conducted between 2020 and 2023, ChatGPT achieved significantly higher scores than the mean exam success rate (p=0.012, p=0.012, p=0.002, p=0.005, respectively). Of the 400 questions analyzed, 36 (9%) included figures. CONCLUSION: This is the first study to evaluate the performance of ChatGPT in the Turkish orthopedics proficiency exam. Our findings indicate that ChatGPT demonstrated high success in the Turkish Orthopedics and Traumatology Board Examination (TOTBE) written exam, achieving higher scores than the vast majority of candidates taking the exam (98.7%). ChatGPT performed well in the first part of the proficiency exam, where only theoretical knowledge is assessed. However, the human factor, which synthesizes both theoretical and practical knowledge, remains essential in daily medical practice.

Ähnliche Arbeiten