Could ChatGPT Be a Tool Capable of Providing Qualified, Empathetic, and Assertive Answers to Patients After Bariatric Surgery? A Comparative Analysis of Its Versions.

Abstract

Introduction: Artificial intelligence (AI) is revolutionizing healthcare by providing advanced tools that enhance clinical practice. Bariatric surgery, a key treatment for severe obesity, requires consistent postoperative care to ensure optimal outcomes, which can be challenging. AI, including ChatGPT, has been employed in various medical fields, but its role in postoperative bariatric surgery follow-up remains underexplored. This study evaluates and compares the effectiveness of ChatGPT 3.5 and 4 in responding to common postoperative questions.

Methods: A cross-sectional pilot study, where a multidisciplinary team of bariatric surgery experts identified ten frequently asked questions related to the postoperative period. These questions were presented to ChatGPT 3.5 and 4 under a standardized prompt simulating a real patient interaction. The responses were evaluated by a group of specialists using criteria such as response quality, empathy, safety, and clinical acceptance. Statistical analyses were conducted to compare the performance of the two versions.

Results: The analysis revealed that ChatGPT 4 showed a trend toward better performance compared to GPT-3.5, particularly in accuracy and safety, although several comparisons did not reach statistical significance. Additionally, both versions demonstrated a high level of empathy, though ChatGPT 4 showed a superior ability to provide clinically acceptable and safe responses.

Conclusion: ChatGPT has demonstrated the ability to provide qualified, empathetic, and clear responses to most questions. ChatGPT-4, in particular, offered more satisfactory responses with fewer errors. The tool shows great potential in the therapeutic approach for bariatric surgery patients, but further studies are needed to enhance its safe and effective use by both physicians and patients.