Delving into PubMed Records: Some Terms in Medical Writing Have Drastically Changed after the Arrival of ChatGPT

Abstract

AbstractIt is estimated that ChatGPT is already widely used in academic paper writing. This study aims to investigate whether the usage of specific terminologies has increased, focusing on words and phrases frequently reported as overused by ChatGPT. The list of 118 potentially AI-influenced terms was curated based on posts and comments from anonymous ChatGPT users, and 75 common academic phrases were used as controls. PubMed records from 2000 to 2024 (until April) were analyzed to track the frequency of these terms. Usage trends were normalized using a modified Z-score transformation. A linear mixed-effects model was used to compare the usage of potentially AI-influenced terms to common academic phrases over time. A total of 26,403,493 PubMed records were investigated. Among the potentially AI-influenced terms, 75 displayed a meaningful increase (modified Z-score ≥ 3.5) in usage in 2024. The linear mixed-effects model showed a significant effect of potentially AI-influenced terms on usage frequency compared to common academic phrases (p < 0.001). The usage of potentially AI-influenced terms showed a noticeable increase starting in 2020. This study revealed that certain words and phrases, such as “delve,” “underscore,” “meticulous,” and “commendable,” have been used more frequently in medical and biological fields since the introduction of ChatGPT. The usage rate of these words/phrases has been increasing for several years before the release of ChatGPT, suggesting that ChatGPT might have accelerated the popularity of scientific expressions that were already gaining traction. The identified terms in this study can provide valuable insights for both LLM users, educators, and supervisors in these fields.Author SummaryArtificial intelligence systems have rapidly integrated into academic writing, particularly in the medical and biological fields. This study investigates changes in the frequency of specific terminologies reported as overused by ChatGPT. By analyzing PubMed records from 2000 to 2024, we tracked 118 potentially AI-influenced terms and compared them with 75 common academic phrases. The study’s findings reveal that terms such as ‘delve,’ ‘underscore,’ ‘meticulous,’ and ‘commendable’ saw a marked increase in usage in 2024. However, this trend actually began around 2020. This suggests that while some of these terms were already gaining popularity before the release of ChatGPT, the large language model may have accelerated their adoption in scientific literature. Furthermore, the analysis highlights that the impact of ChatGPT extends beyond new terminologies to altering the frequency and style of commonly used academic phrases. Understanding these trends can help researchers and educators see how AI tools are shaping academic writing.