Identifying patients at risk (PDWS)

Patient monitoring in emergency departments typically includes basic physiological vital signs as well as assessments of pain and consciousness.

All projects

Project period

Start: 2019
End: 2022

Automated monitoring provides a continuous stream of data that enable real-time tracking of patient state, but reliable projection of trajectory is challenged by the sparseness of data and variation in patterns. The project sought to address both the challenges of device utilisation and patient risk stratification.

Aim

The project aimed to reduce the number of unforeseen deteriorations of patients in medical emergency departments. The project’s approach was looking at how data from the information systems that hospitals already use could be utilised in novel ways through machine learning based models. Ideally, the models would be able to span the diversity of patients and medical device utilisation.

The project built on the system Patient Deterioration Warning System (PDWS), which was developed in a previous project. The PDWS is currently integrated into the emergency department’s monitoring systems.