Inflammatory bowel disease, peptic ulcer and diverticular disease as certified causes of death in an English population 1979-2003.

BACKGROUND: When gastrointestinal diseases are certified as causes of death, they are often not selected as the underlying cause. Until recently, only one underlying cause of death has been coded and analysed in official national statistics in England and many other countries. AIMS: To report on the total 'burden of mortality' from some common gastrointestinal diseases, and whether it has changed over time, including all certified causes of death as well as underlying causes, (i) in the Oxford region from 1979 to 2003, (ii) in England from 1996 to 2003; and to quantify the under-ascertainment of cause-specific mortality when based on underlying cause alone. METHODS: We searched death certificate data from the Oxford Record Linkage Study database, and from English national data, for specified gastrointestinal diseases certified as underlying or contributory causes of death. RESULTS: For all the conditions studied, underlying-cause-coded mortality missed a substantial percentage of all certified deaths. The extent of underestimation varied according to the periods in which different criteria were used for the selection of the underlying cause. For example, in Oxford, in the latest period 1993-2003, underlying-cause-coded mortality identified only 37% of all death certificates with ulcerative colitis, 47% of Crohn's disease, between 62 and 68% for the different types of peptic ulcer and 66% of diverticular disease. CONCLUSIONS: Studies of mortality for these diseases should take account of all certified causes as well as underlying-cause mortality. This is particularly important for analyses that go across periods of change to the rules for selecting the underlying cause of death.