Multiple primary cancers at memorial hospital 1949-1962

Abstract

The Memorial Hospital, New York, surveys of multiple primary cancer risks undertaken during 1965-1971 are summarized. After re-examination the methodological problems noted were: 1) that prospective studies tell us more than prevalence or retrospective studies; 2) that risks must be related to the observation period, patient's ages, etc.; 3) that hospital series probably are selected but that this may be overcome by internal comparisons; 4) that the search for explanatory variables is difficult, to the extent data on these variables were not explicitly collected; and 5) that computing multiple risk ratios for each of many cancer types guarantees a number of accidentally statistically significant results. It was also concluded that most cancer patients have excess risk for specific later cancers, depending on their original cancer type and that knowledge of these risks is of true clinical importance.