Accuracy of diagnosis of cancer among autopsy cases: JNIH--ABCC population for Hiroshima and Nagasaki.

The accuracy of death certificate diagnoses of cancer in the fixed population of about 100,000 samples in Hiroshima and Nagasaki was determined for the period 1961 approximately 1970 by comparison with autopsy findings. In general, when the death certificate listed cancer as a cause of death it was found at autopsy in a high proportion of cases. However, cancer was not always reported on death certificates, indicating that cancer occurs more frequently than recorded by official mortality statistics. Older persons, persons who die at home, and persons with certain cancers are more likely not to have cancer named on their death certificates. It is estimated that in the 10,749 deaths occurring at home or in hospital, there were 32% more deaths due to cancer than certified on death certificates (3,095 vs. 2,345) and for persons aged 70 or more dying at home it is estimated there were 55% more stomach cancer (269 estimated vs. 174 listed) and 244% more lung cancer (141 estimated vs. 41 listed) than were certified on death certificates. The death certificate is not a good source of information for cancer of the cervix because many cases of this disease reported on death certificates as cancer of the uterus. This practice needs to be taken into account in the use of mortality data for cervical cancer in Japan.