ABSTRACT
I. INTRODUCTION
The bulk of the material in this book describes statistical methods useful in studying problems arising in the field of cancer. The methods assume that human cancer can be sufficiently well defined (1) to identify a case of one of these diseases when it occurs in a population; (2) to permit assignment of mortality due to cancer to one of the cancer rubrics with sufficient accuracy to be useful; or (3) to provide unambiguous operational time of occurrence data. Whether these assumptions are justified in a particular application of the methods will depend upon a number of circumstances. One purpose of this chapter will be to describe some of the characteristics of these diseases which have particular relevance to these circumstances. A second objective of the chapter will be to set forth in some detail the sources of information concerning the frequency of occurrence of human cancer. So-called tumor registries are an important element in this history and a forerunner of current large data gathering operations in this field. Consequently registries and their sequallae will be described in a separate section. A final section will describe the strengths and shortcomings of available information on cancer relative to the purpose of this book.
