A learning health system approach to avoiding acute occupancy crises

Acute NHS providers currently face severe problems due to high bed occupancy rates. This leads to overcrowding, worse outcomes for patients and poor patient experience. Patients may be admitted to inappropriate wards, which can affect the care they receive and add to lengths of stay.

Occupancy peaks, when the hospital is at or above bed capacity, can have critical consequences for services. Without means of accurate early warning, the problem often only becomes apparent when it has started to have an impact, which leaves little time for escalation response.

This innovative project, led by NIHR CLAHRC North West London and hosted by Chelsea and Westminster NHS Foundation Trust, will build a bespoke local occupancy prediction model with a 48-hour horizon. It will use local data and advanced informatics, and will be used to alert formal discharge escalation.

By giving a warning that the hospital is heading towards an occupancy crisis in advance, appropriate discharge processes can begin while there is still time for the crisis to be averted.

The prediction model will take into account factors that might influence length of stay, as identified from interviews with staff and patients, and will take into account local practices. The escalation response itself will be developed using improvement principles.

The prediction model will be developed by NIHR CLAHRC North West London. Trust analysts will run the model, and responses to alerts will involve front line staff. The working group for the project will include patients who have used the service, to ensure wide engagement and ownership.

Contact information

For more information about this project, please contact Paul Sullivan, Consultant Physician and Honorary Senior Lecturer in Improvement Science, NIHR CLAHRC North West London.