Value of a multinational approach in determining the causation of cancer.

MacMahon and Pugh define epidemiology as the use of knowledge on the frequency and distribution of disease to search for determinants. This paper demonstrates that a multinational approach in cancer epidemiology can be of great value in at least four circumstances: namely, the compilation and standardization of data, the assessment of risk, the pooling of study populations to obtain interpretable results, and the provision of resources for specific epidemiologic investigations. One or two examples are given for each category--the determination of international cancer incidence patterns, the evaluation of the risk posed by chemicals to man, assessment of the effects of low doses of ionizing radiation, determination of the long-term effects of exposure to asbestos substitutes, and studies on the influence of diet on esophageal cancer.