The transformative impact of large language models on medical writing and publishing: current applications, challenges and future directions.

English summary

This review examines the impact of large language models (LLMs) on multiple stages of medical research and scholarly publishing, with a focus on experimental evidence. LLMs are being used for literature searches, research design, writing assistance, quality assessment, citation generation, and data analysis, as well as for manuscript screening, generation of review comments, and identification of potential biases in peer review and publication. Their adoption also raises concerns about accuracy, reliability, transparency, bias, and research integrity.

Key points

• Global surveys indicate a high prevalence of LLM use in scientific writing, with both potential benefits and challenges.
• Reported applications include literature search, research design, writing support, quality assessment, citation generation, and data analysis.
• In peer review and publication, LLMs have been used for manuscript screening, generating reviewer comments, and identifying potential biases.

Audience

Medical researchers, scholarly authors and reviewers, editors, medical educators, and advanced learners interested in LLM-based research and publishing tools

Scope and limitations

The supplied text is a summary of a review on LLM applications in medical writing and publishing. It does not provide the review's search strategy, eligibility criteria, study-level samples, or quantitative findings; therefore, the magnitude and quality of the reported evidence cannot be independently assessed from this excerpt. It also explicitly notes the need to address current LLM limitations and biases and to update policies and practices as the technology evolves.

Research context

The topic has a neutral thematic connection to medical AI and data science through its focus on LLM use in research and publishing, human-AI collaboration, algorithmic bias, and scholarly integrity.