An AI-based method has been developed to predict relapses in A&E patients

This software predicts that 7 per cent of people treated in A&E departments in the Region of Murcia will return within a month of being discharged, and can calculate an individual’s risk of readmission.

A multidisciplinary research project at UCAM, led by Dr Juan José Hernández and Dr Horacio Pérez, has developed a methodology based on artificial intelligence (specifically machine learning) capable of predicting, with 95 per cent accuracy, relapse in patients who have been treated in A&E departments in the Region of Murcia.

This achievement has been made possible thanks to data provided by the Murcia Health Service, which has enabled the creation of a system that estimates that around 7 per cent of patients treated in A&E – and who meet certain criteria defined by the algorithm – will require readmission within 30 days of being discharged. Furthermore, the model can calculate the individual risk of readmission for each patient.

The system has been built using a wide range of information, including clinical and demographic data. This includes procedures carried out during hospitalisation, the time of admission and discharge, medical history, postcode, gender, age and lifestyle factors such as smoking or alcohol consumption. The combination of all these factors enables the model to identify patterns that influence the likelihood of a relapse following treatment in A&E.

The UITA, BIO-HPC and Hydro MRLab research groups have developed this tool, which will be made available to the Murcia Health Service or any other health service wishing to use it to improve hospital management and care for the public.

You can find the full article here: https://www.mdpi.com/2504-4990/6/3/80