Reducing emergency department crowding through predictive data analytics – flow|ER

Crowding is a known contributor to mortality, morbidity and poor patient experience in emergency departments (EDs). ED crowding means ambulances are unable to offload patients, and patients have significant delays for initial assessment and subsequent admission.

The current way of managing ED workload is to react to demand. This creates bottlenecks and stress for staff.

This intervention, to be implemented at the ED at Addenbrooke’s Hospital in Cambridge, will involve using a software solution, flow|ER, to predict which parts of the ED will be under stress, and require support and intervention (for example paediatrics, minor injuries, resuscitation room), and which patient pathways (for example admitted or discharged, self-presenter or ambulance arrival). The software can predict the level of crowding that the ED can expect up to eight hours ahead.

Management interventions will then be implemented to mitigate the predicted surge or lull in workload. The interventions are a set of structured response suggestions to address demand. These interventions have been developed in conjunction with a wide range of ED staff.

The impact the intervention has on crowding will be monitored over the course of nine months, and data on crowding will be compared to historic data at Addenbrooke’s and comparable hospitals. Measurements will include using the international crowding measure in emergency departments (ICMED) tool, the four-hour waiting target and qualitative feedback from staff and patients.

It is estimated that flow|ER will solve 15% of crowding in the department, which means over 2,000 patients per year would benefit.

Contact information

For more information about this project, please contact Dr Adrian Boyle, Consultant Emergency Physician, Addenbrooke’s Hospital, Cambridge.