LUMC admission predictor (SEH to Acute Opname Afdeling)

par LUMC

AI Application

Faible confiance

The LUMC admission predictor ('opnamevoorspeller') is an in-house AI tool that estimates how many patients at the emergency department (SEH) will be admitted to the hospital's Acute Admission Unit (Acute Opname Afdeling, AOA). It is used by the AOA's nurse admission coordinators for bed planning. The LUMC says the model takes variables from the patient record, such as vital-sign measurements and requests for blood tests and radiology; it makes a first prediction within about 10 minutes of a patient's arrival and updates it when new data, such as a lab request, are entered. The model type is not stated. Coordinators see the expected inflow from the SEH on a dashboard next to an existing capacity dashboard, and decide themselves whether to free beds; they also use the figures in the daily bed meeting with other wards. It was built by CAIRELab, the LUMC's AI expertise centre, with the departments and the LUMC capacity centre. Where it runs and how data are stored is not stated.