Is the apparent rise in cancer mortality in the elderly real? Analysis of changes in certification and coding of cause of death in England and Wales, 1970-1990.

The effect of changes in recording and coding of cause of death on trends in cancer mortality in England and Wales in persons aged 45 and over during 1970-1990 is reviewed. During this period, all-cancer mortality rates increased only at ages over 75 in males and over 55 in females. Rises in cancer mortality were largely due to increases in cancer of lung, prostate and unspecified site in men, and of lung, breast and unspecified site in women. Death coding and certification artefacts were much larger in older persons. In those aged 75-84, a change in the position of recording cancer on the death certificate could potentially account for 46% of the recorded increase in prostate-cancer mortality and 28% of the increase in breast-cancer mortality. The decrease in recorded mortality from ill-defined terminal events was far greater than the increase in cancer mortality in this age group. The rise in all-cancer mortality in the elderly was partly due to an increase in lung-cancer mortality, and data artefacts explained a large proportion of the increase in the other common specified cancers in those aged 75-84. The use of routine mortality statistics to chart progress against cancer lacks validity at older ages because of imprecision in certification of cause of death.