Non-Simultaneously Diagnosed Multiple Primary Malignant Neoplasms: General Considerations

Abstract

In the case of non-simultaneously occurring multiple primary malignancies, available methods allow one to compare the observed and expected number of second primary cancers. Without the presence of an adequate control group, it is impossible to evaluate whether a reported association between two cancers simply represents a chance occurrence. It is imperative always to question whether a reported association is real or artifactual. In examining results of such investigations, one must constantly consider certain possibilities for bias. Problems inherent in the use of data from selected populations, establishing adequate statistical techniques, and utilizing a proper control population have already been discussed in Chapter 1.