Can We Ask ChatGPT About Drug Safety? Appropriateness of ChatGPT Responses to Questions About Drug Use and Adverse Reactions Received by Pharmacovigilance Centers.

Autor:	Pariente, Antoine, Salvo, Francesco, Bres, Virginie, Faillie, Jean-Luc
Předmět:	GENERATIVE artificial intelligence LANGUAGE models CHATGPT DRUG side effects ARTIFICIAL intelligence
Zdroj:	Drug Safety; Sep2024, Vol. 47 Issue 9, p921-923, 3p
Abstrakt:	This document discusses a pilot study that assessed the appropriateness of responses generated by ChatGPT, an AI language model, to questions about drug safety. The study compared the responses provided by ChatGPT to those given by pharmacovigilance experts. Overall, 56.7% of ChatGPT responses were deemed reliable, with slightly higher reliability for patient questions compared to healthcare professional inquiries. The study suggests that ChatGPT could be a valuable tool for drafting responses, but emphasizes the importance of human validation and further investigation. [Extracted from the article]
Databáze:	Complementary Index
Externí odkaz:	Zobrazit plný text záznamu Full text from SpringerLink