Isolation-room capacity tracked separately from bed counts

System failure identification

A patient needing isolation cannot be placed although beds are free. Isolation capacity is the binding constraint, but it is managed as an infection-control concern rather than a capacity one. Demand for isolation rises at exactly the times general capacity is also tight.
Data pipeline diagnostics

Isolation status lives in the infection control system and bed state in the capacity system. No joined view exists, so isolation-eligible capacity is never computed as a number at all. Two systems each hold half of the answer and neither holds the question.
Predictive computational model

Compute isolation-capable beds as a first-class capacity figure, and forecast demand for it from current admissions and local infection trends, separately from general beds. Infection trend data already exists and is simply not used for capacity planning.
Workflow integration

Add it as a distinct row in the capacity view beside general beds. The change is representational: the same beds, no longer counted as interchangeable. Nothing physical changes, and the reported number stops being wrong.

System failure identification