Abstract
Prior to undergoing total knee arthroplasty (TKA), surgeons are often confronted with patients with numerous questions regarding the procedure and the recovery process. Due to limited staff resources and mounting individual workload, increased efficiency, e.g., using artificial intelligence (AI), is of increasing interest. We comprehensively evaluated ChatGPT's orthopedic responses using the DISCERN instrument. Three independent orthopedic surgeons rated the responses across various criteria. We found consistently high scores, predominantly exceeding a score of three out of five in almost all categories, indicative of the quality and accuracy of the information provided. Notably, the AI demonstrated proficiency in conveying precise and reliable information on orthopedic topics. However, a notable observation pertains to the generation of non-existing references for certain claims. This study underscores the significance of critically evaluating references provided by ChatGPT and emphasizes the necessity of cross-referencing information from established sources. Overall, the findings contribute valuable insights into the performance of ChatGPT in delivering accurate orthopedic information for patients in clinical use while shedding light on areas warranting further refinement. Future iterations of natural language processing systems may be able to replace, in part or in entirety, the preoperative interactions, thereby optimizing the efficiency, accessibility, and standardization of patient communication.