Validation of cause-of-death certification for outpatient cancers: the contrasting cases of melanoma and mycosis fungoides.

The reported incidence of cancers typically diagnosed in the outpatient setting may underestimate their true frequency, but the validity of mortality estimates from many of these causes has not been studied in detail. The authors sought to evaluate the validity of mortality estimation from death certificates for two such cancers, melanoma and mycosis fungoides (a cutaneous lymphoma), using routinely collected cancer registry data for 1973-1994 from the Surveillance, Epidemiology, and End Results program. Their method estimates that 93% of the deaths attributable to melanoma were certified as due to melanoma but that only 60% of the deaths attributable to mycosis fungoides were so certified. Evaluation of the accuracy of cause-of-death certification in this manner is helpful in the interpretation of mortality statistics.