In Aesthetic surgery journal ; h5-index 35.0
BACKGROUND : In the past 3 months, OpenAI, a San Francisco based artificial intelligence (AI) research laboratory, has released ChatGPT, a conversation large language model (LLM). ChatGPT has the ability to answer user questions, admit to mistakes, and learn from users that are accessing the program.
OBJECTIVES : Due to the importance of producing evidence-based research in plastic surgery, the authors of this study wanted to determine how accurate ChatGPT could be in creating novel systematic review ideas that encompasses the diverse practice of cosmetic surgery.
METHODS : ChatGPT was given commands to produce 20 novel systematic review ideas for 12 different topics within cosmetic surgery. For each topic, the system was told to give 10 general and 10 specific ideas that were related to the concept. In order to determine the accuracy of ChatGPT, a literature review was conducted using Pubmed (National Institutes of Health, Bethesda, MD), CINAHL (EBSCO Industries, Birmingham, AL), EMBASE (Elsevier, Amsterdam, the Netherlands), and Cochrane (Wiley, Hoboken, NJ).
RESULTS : A total of 240 'novel' systematic review ideas were constructed by ChatGPT. We determined that the system had an overall accuracy of 55%. When topics were stratified by general and specific ideas, we found that ChatGPT was 35% accurate for general ideas and 75% accurate for specific ideas.
CONCLUSIONS : ChatGPT is an excellent tool that should be utilized by plastic surgeons. ChatGPT is versatile and has uses beyond research including patient consultation, patient support, and marketing. As advancements in AI continue to be made, it is important for plastic surgeons to consider the use of AI in their clinical practice.
Gupta Rohun, Park John B, Bisht Chirag, Herzog Isabel, Weisberger Joseph, Chao John, Chaiyasate Kongkrit, Lee Edward S