Large language models (LLMs), a type of artificial intelligence, have gained significant attention in healthcare for their ability to generate human-like text. Researchers and healthcare professionals are exploring ways to integrate LLMs into clinical tasks, such as note generation and diagnostic support. Various LLMs, including GPT and others fine-tuned for medical use, have shown promise in tasks like predicting hospital length of stay and readmissions. Effective patient communication is crucial in healthcare, and text-based AI models offer potential for enhancing this. Organizational health literacy, which involves making health information easier for patients to understand, is central to this. Studies show that improving readability of communications is linked to better patient outcomes, and LLMs could be a useful tool here. A study by Zaretsky et al used GPT-4 to transform hospital discharge summaries into patient-friendly formats. They reviewed existing summaries and patient preferences to create a concise, one-page summary. The results showed that the LLM-generated summaries were shorter, more readable, and more understandable than original summaries. However, 18 summaries were flagged for potential safety risks, with omissions and hallucinations being the main issues. While the study highlights the potential of LLMs in enhancing patient education, safety concerns and practical challenges in implementation remain. The study also notes that LLMs may not be ready for widespread use due to safety risks and technological barriers. However, with improvements in safety and automation, LLMs could become important tools in healthcare communication. The study underscores the need for careful oversight and further research to address these challenges.Large language models (LLMs), a type of artificial intelligence, have gained significant attention in healthcare for their ability to generate human-like text. Researchers and healthcare professionals are exploring ways to integrate LLMs into clinical tasks, such as note generation and diagnostic support. Various LLMs, including GPT and others fine-tuned for medical use, have shown promise in tasks like predicting hospital length of stay and readmissions. Effective patient communication is crucial in healthcare, and text-based AI models offer potential for enhancing this. Organizational health literacy, which involves making health information easier for patients to understand, is central to this. Studies show that improving readability of communications is linked to better patient outcomes, and LLMs could be a useful tool here. A study by Zaretsky et al used GPT-4 to transform hospital discharge summaries into patient-friendly formats. They reviewed existing summaries and patient preferences to create a concise, one-page summary. The results showed that the LLM-generated summaries were shorter, more readable, and more understandable than original summaries. However, 18 summaries were flagged for potential safety risks, with omissions and hallucinations being the main issues. While the study highlights the potential of LLMs in enhancing patient education, safety concerns and practical challenges in implementation remain. The study also notes that LLMs may not be ready for widespread use due to safety risks and technological barriers. However, with improvements in safety and automation, LLMs could become important tools in healthcare communication. The study underscores the need for careful oversight and further research to address these challenges.