Links to other entries shown thus

event rates

For the purpose of estimating occurrence parameters some diseases are assumed not to remit (for example ischaemic heart disease and stroke). This means that only 'first in lifetime' disease onsets qualify as 'incident events'. However in a proportion of disease onsets there may be insufficient information to classify the event as 'first in lifetime'. This has proved to be a particular problem with ischaemic heart disease where one type of incident event is sudden cardiac death.

For a significant proportion of such deaths there may be insufficient information to know if it really was a first in lifetime manifestation of ischaemic heart disease or not (eg where the person lived alone and medical records could not be traced). If this proportion of unclassifiable events varies across the populations being compared, the comparability of the incidence estimates is lost. In these circumstances an event rate may be deemed a more robust comparator.

Whereas the numerator for the relevant incidence rate is 'first in lifetime disease onsets', the numerator for an event rate is 'all disease onsets'. It was because of the difficulty noted here that the international MONICA study which set out to compare IHD incidence rates in different populations ended up comparing event rates.

See http://www.ktl.fi/publications/monica/manual/part4/iv-1.htm