The validity of determining pancreatic cancer incidence in the United States by the use of mortality data.

To test the validity of using pancreatic cancer mortality as an indicator of incidence, mortality data for 1969-1971 were compared to incidence data for those areas covered by the Third National Cancer Survey, which was conducted during the same time period. Good correlation was found for whites but not for blacks. Small numbers of black study subjects in the Third National Cancer Survey probably influenced the incidence data for blacks. Pancreatic cancer mortality and incidence data for U.S. whites appear to be fairly accurate indicators of each other. The same cannot be said for the corresponding data for blacks. The pancreatic cancer mortality data from 1970 to 1978 for whites in the United States were analyzed to determine the current trend of the disease. In contrast to previously increasing rates, the age-adjusted rates for white females were determined to have leveled off, while the age-adjusted rates for white males appear to have started to decline slightly.