Remarks

In order to be comparable between different populations, relative survival figures must be either age-specific or age-standardised. In EUROCARE-3 relative survival tabulated by country is presented as age-standardised survival. Age standardised country-specific survival was calculated by the direct method, using five age classes and, as standard, the age distribution of the whole set of cases analysed for each site. The same standard distribution was used for both sexes.

Difficulties in ascertaining the vital status of incident cases generally result in an overestimation of survival as deaths are missed. The relative survival data for poor prognosis cancers, such as lung, pleura and liver cancer are indirect indicators of follow-up quality. High survival for such cancers suggests (but does not prove) inadequate follow up procedures. Inadequate follow-up is likely in Spain, Austria and, for 10-year survival, in Wales (see survival for lung cancer). The survival data for these countries for breast and cervical cancer should therefore also be considered less reliable (Capocaccia et al., 2003).

Screening or early diagnostic activity (for breast and cervical cancers) must be taken in account in the interpretation of differences in survival among countries when such activities are not similar in all countries (Capocaccia et al., 2003).