Evaluating primary prevention programmes against cancer.

Based on current knowledge, roughly one third of all cancers worldwide are preventable, and primary prevention is increasingly seen as an important cancer control strategy. Interventions to reduce the exposure to known causes can be effected through legislation or education, or by means of vaccination or chemoprevention. Since primary prevention actions can be costly and will compete for resources needed for other disease control activities, and since there is no guarantee that they will be successful, they should not be introduced haphazardly but on the basis of scientific evaluations. This paper presents the main principles to be followed in designing such evaluations; the illustrations often, of necessity, come from other diseases (particularly cardiovascular disease), where there is considerably more experience. Because the interventions involve changes in lifestyle and behaviour, and because a long time is necessary to observe the ultimate endpoints, controlled intervention studies against cancer present many scientific and logistical difficulties. Some interventions, such as vaccination and chemoprevention (to test suspected protective agents) may be evaluated by traditional clinical trial methodology, using intermediate as well as final (cancer incidence and/or mortality) endpoints. Active, target-directed and preferably controlled health service research studies will definitely be needed to assess community or population interventions based on legislation or education.