Objectives: ChatGPT is a large language model that is able to generate human-like text. The aim of this study was to evaluate ChatGPT as a potential supplement to urological clinical practice by exploring its capacity, efficacy and accuracy when delivering information on frequently asked questions from patients with bladder cancer. Methods: We proposed 10 hypothetical questions to ChatGPT to simulate a doctor–patient consultation for patients recently diagnosed with bladder cancer. The responses were then assessed using two predefined scales of accuracy and completeness by Specialist Urologists. Results: ChatGPT provided coherent answers that were concise and easily comprehensible. Overall, mean accuracy scores for the 10 questions ranged from 3.7 to 6.0, with a median of 5.0. Mean completeness scores ranged from 1.3 to 2.3, with a median of 1.8. ChatGPT was also cognizant of its own limitations and recommended all patients should adhere closely to medical advice dispensed by their healthcare provider. Conclusions: This study provides further insight into the role of ChatGPT as an adjunct consultation tool for answering frequently asked questions from patients with bladder cancer diagnosis. Whilst it was able to provide information in a concise and coherent manner, there were concerns regarding the completeness of information conveyed. Further development and research into this rapidly evolving tool are required to ascertain the potential impacts of AI models such as ChatGPT in urology and the broader healthcare landscape.
Full Text Finder