Guiding a large language model via specific prompts is critical for producing desired outputs and ensuring healthcare systems make the most out of generative artificial intelligence.