EzaPredictive

by Expertisecentrum Zorgalgoritmen

AI Application

Low Confidence

EzaPredictive is a set of machine-learning prediction models for hospital emergency departments (SEH) and wards, made by Expertisecentrum Zorgalgoritmen (EZA), a company set up by the Samenwerkende Algemene Ziekenhuizen. EZA says EzaPredictive 1.0 predicts how long a patient will stay at the SEH, the chance that the patient will be admitted, and the expected length of stay on the ward; alongside it EZA supplies an SEH crowding predictor ('Code Rood Voorspeller') that shows current and expected crowding for the coming hours. EZA says the models are trained with federated learning across hospitals, so patient data stay in each hospital, and that clinicians see which factors contribute to each prediction. The predictions are shown in the hospital's electronic patient record (EPD) to support patient flow and capacity planning; in EZA's report of the first deployment, the coordinating SEH doctor and nurse judge each prediction and decide what to do. EZA registered EzaPredictive 1.0 as an MDR class I device, which EUDAMED has listed as no longer on the market since 16 July 2026; hosting is not stated beyond installation in each hospital.