Hospital patient-flow simulation: a worked example

The hospital patient-flow example represents arrivals, triage, diagnostics, treatment, boarding, discharge, and admission. Use it to ask which stage should receive extra capacity under ordinary demand and under a surge, rather than assuming that more capacity anywhere will help equally.

The baseline experiment runs for 24 hours with 40 replications. Record waiting and utilization at triage, diagnostics, treatment, and boarding, plus throughput and time in system for the main outcomes.

Before comparing alternatives, verify routes and totals, then compare the baseline with whatever real patient-flow data the study has available.

Compare alternatives such as one extra triage slot, two extra diagnostic slots, two extra treatment slots, additional boarding capacity, and a coordinated surge-capacity option.

A useful result should show where waiting moved as well as where it fell. Removing a triage queue can simply expose a treatment or boarding constraint downstream.

The model also includes an arrival-surge environment. An alternative that works at ordinary demand may fail when the arrival rate rises, so report both conditions rather than averaging them together.

Do not report one staffing number without the evidence behind it. Show the assumptions, validation checks, run settings, uncertainty, and tradeoffs that produced the recommendation. Replace the example's assumptions with the hospital's own process and data before using it for a real staffing decision.

• Model and data version used for the study.
• 24-hour horizon, replication count, seed policy, and any warm-up rule.
• Waiting, utilization, throughput, and time-in-system measures used for the decision.
• Confidence intervals or paired differences for scenario comparisons.
• Validation checks and known limits of the model.