Hospital Capacity Simulator

What you just saw

The epidemic curve itself is familiar — rise, peak, fall. The new thing here is the horizontal capacity line. When infection demand crosses the bed or ICU line, the excess patients cannot be treated at the full standard of care, and mortality for that cohort jumps sharply. The total mortality is not just the R₀ of the disease — it is R₀ plus the peak-to-capacity ratio.

The science behind it

Simple SEIR models treat mortality as a fixed rate per infected case. Real-world hospital data from COVID-19 (and earlier from H1N1 and SARS-CoV-1) show mortality as a function of local ICU utilization: once you pass ~90% ICU occupancy, case-fatality climbs nonlinearly. This simulator captures the mechanism — healthcare capacity as a threshold that converts infection surge into excess mortality.

Try these experiments

1. 1. Flattened curve

   Settings: r0=1.3 total_beds=500

   What to look for: Long epidemic, never exceeds capacity. Mortality stays at baseline.

2. 2. Overwhelmed

   Settings: r0=3.0 total_icu=50

   What to look for: ICU saturates at the peak. Excess mortality visible as a clear step-up on the death curve.

3. 3. Surge capacity

   Settings: r0=3.0 total_icu=200

   What to look for: Same epidemic, 4× ICU — mortality drops dramatically. Capacity planning matters enormously.

from scirouter import SciRouter
client = SciRouter(api_key="sk-sci-...")
result = client.simulation.run(
    model="hospital",
    params={"population": 1_000_000, "r0": 2.5,
            "total_beds": 50_000, "total_icu": 5_000},
    steps=365, seed=42,
)
print(result.metrics["excess_mortality"])