Mp72-16 Not Quite Ready for the Pediatric Urology Playground: An Assessment of Chatgpt and Cryptorchidism

You have accessJournal of UrologyPediatrics VII (MP72)1 May 2024MP72-16 NOT QUITE READY FOR THE PEDIATRIC UROLOGY PLAYGROUND: AN ASSESSMENT OF CHATGPT AND CRYPTORCHIDISM Yitzchak E. Katlowitz, Michael West, Matthew Antonellis, Steven Friedman, Jaime Freyle, and Jacob Khurgin Yitzchak E. KatlowitzYitzchak E. Katlowitz , Michael WestMichael West , Matthew AntonellisMatthew Antonellis , Steven FriedmanSteven Friedman , Jaime FreyleJaime Freyle , and Jacob KhurginJacob Khurgin View All Author Informationhttps://doi.org/10.1097/01.JU.0001009572.60675.69.16AboutPDF ToolsAdd to favoritesDownload CitationsTrack CitationsPermissionsReprints ShareFacebookLinked InTwitterEmail Abstract INTRODUCTION AND OBJECTIVE: ChatGPT is an artificial intelligence available to the general population. It is unique from other developing and available modalities as it has a user interface conducive towards simple conversational input. As a result, use and popularity has recently expanded by clinicians and patients alike for both non-medical and medical queries. Recent publications have shown the validity of the responses vary. Limited information is available pertaining to ChatGPT's capacity to accurately discuss topics related to pediatric urology, in particular cryptorchidism evaluation, diagnostics, and therapeutic interventions. This project is designed to assess ChatGPT's clinical competency as compared to the American Urological Association (AUA) guidelines regarding the evaluation and management of cryptorchidism. METHODS: The AUA guidelines for evaluation and management of cryptorchidism include 16 main points, subdivided into 9 diagnostic key points and 7 treatment key points. Each point was prompted to ChatGPT as a question with further requests for information validation or clinical next-step recommendations. The answers were then sorted into 3 categories relative to the AUA guidelines: complete and accurate (CA), incomplete but accurate (IA), and incorrect or misleading (IM). RESULTS: Of the 16 key AUA guideline points queried, 11/16 answers by ChatGPT were CA, 3/16 were IA, and 2/16 were IM. For the 9 diagnostics points, 6/9 were CA, 2/9 were IA and 1/9 were IM. Of the treatment points, 5/7 were CA, 1/7 were IA and 1/7 were IM. CONCLUSIONS: ChatGPT offered complete and accurate answers on only 69% of the total points. It was accurate yet incomplete in 19% of cases, and inaccurate or misleading in 13% of cases. This general trend remains when analyzing the key points divided by evaluation, diagnostics and treatment. While ChatGPT may be helpful in the hands of an experienced and knowledgeable individual, it is too inaccurate to be relied upon for complex pediatric urological clinical decisions in both diagnostics and therapeutics at this point in time. Download PPT Source of Funding: No funding was acquired © 2024 by American Urological Association Education and Research, Inc.FiguresReferencesRelatedDetails Volume 211Issue 5SMay 2024Page: e1178 Advertisement Copyright & Permissions© 2024 by American Urological Association Education and Research, Inc.Metrics Author Information Yitzchak E. Katlowitz More articles by this author Michael West More articles by this author Matthew Antonellis More articles by this author Steven Friedman More articles by this author Jaime Freyle More articles by this author Jacob Khurgin More articles by this author Expand All Advertisement PDF downloadLoading ...