Coronavirus Protection - Swiss Cheese Model

Stack vaccination, masking, ventilation/distancing, and testing effectiveness against a baseline exposure risk to estimate the residual coronavirus infection risk that gets through every layer.

Quick Facts

Core idea
Multiple imperfect layers, stacked
Vaccination, masking, ventilation/distancing, and testing each block some but not all exposure - like slices of Swiss cheese with randomly placed holes.
Combining layers
Residual risk = baseline x product of (1 - effectiveness)
The model assumes each layer's failures are independent of the others.

Your Results

Calculated
Residual infection risk
-
After all active layers
Combined layer effectiveness
-
Risk cut by stacking layers
Absolute risk reduction
-
Percentage points below baseline
Weakest layer
-
Best candidate to reinforce

Ready

Enter a baseline exposure risk and each layer's effectiveness, then calculate the residual risk.

About the Swiss cheese model of coronavirus protection

The Swiss cheese model is a way of visualizing layered defense against an infectious hazard. It was adapted from psychologist James Reason's model of how accidents happen in complex systems, and it was popularized for COVID-19 by virologist Ian Mackay, who drew each protective measure as a slice of cheese full of random holes. No single slice - a vaccine, a mask, better ventilation, a rapid test - blocks every possible route of exposure on its own. But when several imperfect slices are stacked together, it becomes much less likely that a hole in one layer lines up with a hole in every other layer at the same time.

How the calculator combines the layers

  • You set a baseline exposure risk: the estimated chance of infection from a specific exposure (say, sustained indoor contact with an infectious person) if none of the protective layers below were in place.
  • Each layer - vaccination, masking, ventilation and distancing, testing and isolation - is entered as a percent effectiveness, the share of exposure risk that layer removes on its own.
  • The layers are combined by multiplying their failure probabilities: residual risk equals the baseline risk times the product of (1 minus each layer's effectiveness), assuming the layers act independently of one another.

Putting the result in context

The residual risk is only as good as the baseline exposure estimate and the effectiveness values you enter - both are judgment calls that should reflect the specific setting, the prevailing variant, local case rates, and how well each layer is actually being used (a mask worn loosely is not the same as a well-fitted one). Treat the output as a way to compare scenarios - what happens if I drop one layer, or add another - rather than as a precise personal probability.

When to consult a professional

This calculator is an educational arithmetic tool, not a medical risk assessment. For guidance on vaccination, testing, or protecting someone at higher risk of severe illness, consult a healthcare provider or your local public health authority, and follow current official guidance rather than the default values shown here.

Frequently Asked Questions

What is the Swiss cheese model of pandemic protection?
The Swiss cheese model, adapted from James Reason's accident-causation theory and popularized for COVID-19 by virologist Ian Mackay, pictures each protective measure - vaccination, masking, distancing, ventilation, testing - as a slice of cheese with random holes representing its imperfections. No single slice blocks every exposure, but stacking several slices together makes it far less likely that the holes in every layer line up at once.
How is the combined risk calculated?
Each active layer is treated as an independent filter with its own effectiveness. The residual risk equals the baseline exposure risk multiplied by the product of (1 minus each layer's effectiveness) across all layers. The combined effectiveness of the stack is 1 minus that same product, expressed as a percentage - this is the standard independent multiplicative layers model used to reason about defense in depth.
Does stacking every layer mean zero risk?
No. This model assumes each layer fails independently of the others, which is a simplification - some layers address the same route of exposure and are not perfectly independent in reality, and reported effectiveness for any given layer varies with variant, fit, timing, and setting. Treat the output as an educational estimate, not a guarantee, and follow current public health guidance for your situation.

Related calculators