Swiss Cheese Model

I thank Jessica for pointing out this podcast and inviting a discussion of Reason's model.

I agree with Pete and Ralph that the model is simplistic and not useful for determining causation. There are a variety of more robust protocols for examining an incident and determining the direct, contributing, and root causes. There are even better graphical methods of presenting causation analysis, such as a fishbone diagram.

The value of the Swiss Cheese graphic is its ability to clearly present layer of protection. Ralph does this with respect to Covid. I have done this in discussing risks associated with working on high-pressure systems and on systems using toxic gases. I have found it very useful when I am asking senior management to fund engineering parameters for system design or upgrade.

All models are reliant on the assumptions used to build the model. If a presentation uses the Swiss Cheese model, it is reasonable to dissect the assumptions underlying each layer, to understand the strengths and weaknesses of the layer and the model. The same should be said for any model.

Just like our current safety paradigm "RAMP", this model must be used with recognition of its virtues and its limitations.

Neal