Shift-change arrival pileups

1. System failure identification

   Arrivals continue at their usual rate during nursing handover, while effective assessment capacity briefly approaches zero. A queue forms twice a day at times that are known months in advance. The resulting backlog persists well past the handover itself.

2. Data pipeline diagnostics

   Staffing is modelled as headcount per shift, a step function. It records that staff are present but not that both the outgoing and incoming teams are occupied with handover rather than with patients. A model with no intra-shift resolution cannot represent a thirty-minute trough.

3. Predictive computational model

   Model effective capacity as a continuous curve that dips through handover, then compare it against forecast arrivals for the same window. The resulting gap is predictable to the minute. The comparison needs no new data collection, only a finer time grain.

4. Workflow integration

   Stagger handover by role so the dip flattens rather than compounds. The change is to the roster template, applied once, rather than a runtime alert that somebody has to act on. A structural fix removes the problem rather than notifying someone about it.

Back to the case study database