No representation of the causal chain is likely to be applicable to every circumstance. Whilst one version of the chain may work for environmental contaminants (e.g. air or water pollution), others may be necessary to represent issues such as health impacts of natural hazards, insect-borne diseases or traffic accidents.  In most cases, also, we will need to add detail to the chain to show the specific elements that are of interest.  The example below shows a causal chain for road traffic injuries. Some elements of the basic (pollution-based) chain are evident: road traffic flows, for example, are akin to environmental concentrations, vehicle-to-vehicle and person-vehicle interactions might be thought of as a form of exposure, and injuries clearly represent health effects.  However, neither sources nor releases are so clearly identifiable.

We therefore have to be willing to customise the causal chain to the issue being assessed. In the process, we will often find that the original, simple linear structure becomes less apparent, and the chain becomes more web-like in structure.  This should not be a cause for concern: the causal chain is only a ‘model’ and should be adjusted to match our perception of reality, not the other way round!