Global Healthcare Professionals’ Perceptions of Large Language Model Use In Practice (Preprint)

BACKGROUND Chat Generative Pre-Trained Transformer (ChatGPTTM) is a large language model (LLM)-based chatbot developed by OpenAITM. ChatGPT has many potential applications to healthcare, including enhanced diagnostic accuracy and efficiency, improved treatment planning, and better patient outcomes. Healthcare professionals’ perceptions of ChatGPT and similar artificial intelligence tools are not well known, and understanding these attitudes is important to inform the best approaches to explore their use in medicine. OBJECTIVE Our aim was to evaluate the healthcare professional's awareness and perception about potential applications of ChatGPT in the medical field, including potential benefits and challenges of adoption. METHODS We designed a 33-question online survey that was distributed amongst healthcare professionals via targeted emails, professional Twitter and Linkedin accounts. The survey included a range of questions to define respondent demographic characteristics, familiarity with ChatGPT, perceptions of this tool’s usefulness and reliability, and opinions on its potential to improve patient care, research, and education efforts. RESULTS One hundred and fifteen healthcare professionals from 21 countries responded to the survey, including physicians, nurses, researchers, and educators. 101 (87.8%) had heard about ChatGPT mainly from peers, social media, and news, and 77 (76.2%) had used ChatGPT at least once. Participants found ChatGPT to be helpful writing manuscripts (n=31, 45.6%), e-mails (n=25, 36.8%), and grants (n=12, 17.6%), accessing the latest research and evidence-based guidelines (n=21, 30.9%), providing suggestions on diagnosis or treatment (n=15, 22.1%), and improving patient communication (n=12, 17.6%). Respondents also felt that the ability of ChatGPT to access and summarize research articles (n=22, 46.8%), to provide quick answers to clinical questions (n=15, 31.9%), and to provide patient education materials (n=10, 21.3%) were also helpful. However, there are concerns regarding utilization of ChatGPT such as the accuracy of responses (n=14, 29.8%), limited applicability to its practice (n=18, 38.3%), legal and ethical considerations (n=6, 12.8%) mainly regarding plagiarism or copyright violations. Participants stated safety protocols such as data encryption (n=63, 62.4%), access control (n=52, 51.5%) may assist in ensuring patient privacy and data security. CONCLUSIONS Our results show that ChatGPT use was widespread among healthcare professionals in daily clinical, research, and educational activities. The majority of our participants have found ChatGPT to be useful however there are concerns about patient privacy, data security, and its legal and ethical issues as well as the accuracy of the information ChatGPT provides. Further studies are required to understand the impact of ChatGPT and other LLMs on clinical, educational, and research outcomes and concerns of its utilization have to be addressed systematically and through appropriate methods.