Inaccuracies in oral cavity–pharynx cancer coded as the underlying cause of death on U.S. death certificates, and trends in mortality rates (1999–2010)

•Data on U.S. mortality from oral cavity–pharynx cancer (OCPC) are usually based on cause of death from death records.•For age < 65 years, mortality data may underestimate OCPC deaths by 22–35%, and by 70–80% for the subgroup of tonsillar cancers, compared to cancer registry diagnosis.•Temporal declines in death rates were evident, however, even after modification for inaccuracies in cause of death.•A lack of decline in death rates was evident for tonsillar cancer.

SummaryObjectiveTo enhance surveillance of mortality from oral cavity–pharynx cancer (OCPC) by considering inaccuracies in the cancer site coded as the underlying cause of death on death certificates vs. cancer site in a population-based cancer registry (as the gold standard).MethodsA database was used for 9 population-based cancer registries of the Surveillance, Epidemiology and End Results (SEER) Program, including deaths in 1999–2010 for patients diagnosed in 1973–2010. Numbers of deaths and death rates for OCPC in the SEER population were modified for apparent inaccuracies in the cancer site coded as the underlying cause of death.ResultsFor age groups <65 years, deaths from OCPC were underestimated by 22–35% by using unmodified (vs. modified) numbers, but temporal declines in death rates were still evident in the SEER population and were similar to declines using routine mortality data for the entire U.S. population. Deaths were underestimated by about 70–80% using underlying cause for tonsillar cancers, strongly associated with human papillomavirus (HPV) infection, but a lack of decline in death rates was still evident.ConclusionRoutine mortality statistics based on underlying cause of death underestimate OCPC deaths but demonstrate trends in OCPC death rates that require continued surveillance in view of increasing incidence rates for HPV-related OCPC.