Share This

This short communication describes programming the artificial intelligence (AI) algorithm for systematic review of current literature about guidelines on the clinical efficacy and safety of biologics in the treatment of chronic rhinosinusitis with nasal polyps (CRSwNP). The authors first performed a systematic literature search for the latest guidelines and included 12 studies in the review. Then ChatGPT was prompted to develop its own decisional algorithm to select the right patients for biologic therapy and to choose the biologic therapy that best suits their needs. The authors report that the ChatGPT algorithm generated human-like responses to their queries and prompts. However, the drawback of ChatGPT is that it may produce biased or inaccurate responses based on inaccuracies in its underlying training data. As ChatGPT relies on statistical patterns in its training data, it is unable to understand the meaning of the questions and can struggle with concepts which require complex understanding. The authors conclude that more research is needed to train ChatGPT and optimise the AI algorithms to select the right patients for biologic therapy and to choose the biologic therapies that best suits their needs.

Is generative pre-trained transformer artificial intelligence (Chat-GPT) a reliable tool for guidelines synthesis? A preliminary evaluation for biologic CRSwNP therapy.
Maniaci A, Saibene AM, Calvo-Henriquez C, et al.
EUR ARCH OTORHINOLARYNGOL
2024;281(4):2167–73.
Share This
CONTRIBUTOR
Gauri Mankekar

Department of Otolaryngology-Head Neck Surgery, Louisiana State University Shreveport, Louisiana, USA.

View Full Profile