Evaluating the Reliability of OpenAI’s ChatGPT-4 in Providing Pre-colonoscopy Patient Guidance

The high degree of guideline adherence by ChatGPT-4.0 underscores its viability as a dependable resource for patient education. Despite its promising results, the study acknowledges limitations such as the structured nature of patient queries and the lack of real patient interactions. The findings suggest a potential role for AI in augmenting patient education and standardizing information dissemination in healthcare.