Cancer Registration in England and Wales: Some Aspects Relevant to Interpretation of the Data

Abstract

SUMMARY Aspects of the England and Wales cancer registration system are discussed in the light of biases and other artefacts which can arise from them. Some are specific to England and Wales data, but many apply to cancer registration data generally. Considerations relevant to interpretation of the data are: the effects of clinical diagnostic practice; the completeness and accuracy of registry data collection; new registrations, deletions and alterations made to the file of registrations after data for a publication have been extracted; unintended duplication of registrations and the effects of rules intended to prevent this; the ways in which use is made of death certificates in cancer registration; the geographic coverage of the individual registries within the national scheme; inaccuracies in population denominators; differences between the definition of place of residence used for cancer registration and that used for population estimates; the coding systems used for the data; and errors. Several of the artefacts identified are of relatively small moment overall but of considerable importance for some sub-group analyses, particularly geographic analyses within England and Wales. Further specification is needed of such effects in England and Wales and other registry data sets.