Mathematical epidemiology has, now more than ever, attracted great attention in the scientific community. However, little exploration has been done in the direction of modelling human behavioural response when facing an epidemic outbreak. In this context, individuals may employ containment measures to prevent the disease spreading, i.e. providing a benefit to the entire population, at a cost that is heterogeneously perceived by each individual. In our work we study the coevolution of cooperative behaviour and individual perception through the lens of evolutionary game theory, considering the adoption of a disease containment measure as a cooperative act. We introduce this game-theoretical framework as the "Norm Compliance Dilemma", where the evolution of agents’ behaviors depends on their distinct, time-evolving perceptions.