[How reliable are our cancer statistics? Cancer cases in Hamburg's autopsy material]

Among the causes of death, cancer ranks second after cardiovascular diseases. If a cancer patient dies before the tumor is diagnosed the malignancy may be unreported among post-mortem findings and thus is often not included in cancer and cause of death statistics.

All 8844 autopsies carried out at the Institute of Forensic Medicine for the City of Hamburg from 1994 to 2001 were analysed and those cases with malignant tumor identified. The cases were then included in the cancer registry if this had not been done previously. The postmortem diagnosis of malignancy was made histologically on formalin-fixed tissue. In these cases attempts were made to find out if the initial diagnosis of malignancy had been made before death.

A malignancy was found in 519 autopsies (5.9 %). In 67.3 % of these (349/519) the patients had been aware of the disease. 27.2 % of all malignant tumors (141/519) were discovered during the post-mortem examination. In 5.6 % of cases (29/519) the question of whether the tumor had been diagnosed before death could not be answered because of lack of data or an incomplete case history. In 52.2 % (271/519) of the autopsies the tumor had definitely been the cause of death. Of these 271 tumors 17 % (46/271) were discovered only during the forensic post-mortem examination. The Hamburg Cancer Registry had been notified of only 58 cases of the 519. As a result of this study 451 new cancer cases (86.9 %) were notified to the Hamburg cancer registry.

The registration of cancer deaths in Hamburg is incomplete and the notification of cancer to the Cancer Registry is inadequate. The estimated number of unreported cases is probably greater and cancer as the cause of death is higher than reported in the statistics.