A method must be provided to support the analysis of security policy rules interdependencies in a (possibly distributed) environment. We propose a Conceptual Graphs based language that will allow us to represent the structure of information and to employ reasoning for consistency checking. We motivate our choice of language by the gained expressivity, the potential for depicting policy associations rigourously and by associated reasoning capabilities. We explain our approach in the context of security requirements for medical systems. We evaluate our work theoretically, by means of an example of a real world policy rule.