What is it about?
Hospitals often struggle to decide quickly whether patients arriving at the emergency department (ED) need to be admitted. This study introduces a new AI method called LEGOLAS that uses large language models (LLMs)—the same technology behind tools like ChatGPT—to make early predictions about hospital admissions. Instead of relying on complex data formats, LEGOLAS turns patient records into readable stories and trains the AI to understand and learn from them.
Featured Image
Photo by Graham Ruttan on Unsplash
Why is it important?
Making accurate decisions early in a patient’s ED journey helps hospitals manage resources better, reduce overcrowding, and improve patient care. LEGOLAS can predict outcomes faster and more accurately than traditional methods, and it also explains how it reaches its decisions, making it more trustworthy for healthcare professionals. This is the first use of LLMs for early hospital admission prediction in Emergency Departments
Perspectives
LEGOLAS could be integrated into hospital systems to support real-time decision-making. It may help doctors and administrators prepare for incoming patients, allocate staff more efficiently, and improve overall hospital operations.
Prof. Donato Malerba
Universita degli Studi di Bari Aldo Moro
Read the Original
This page is a summary of: Leveraging a large language model (LLM) to predict hospital admissions of emergency department patients, Expert Systems with Applications, August 2025, Elsevier,
DOI: 10.1016/j.eswa.2025.128224.
You can read the full text:
Contributors
The following have contributed to this page
