[Nationwide statements from regional data: methods of the Center for Cancer Registry Data].

Despite having achieved nationwide registry coverage in addition to substantial improvements in data on the epidemiology of cancer in Germany, the Centre for Cancer Registry Data continues to estimate national statistics on incidence, survival, and prevalence instead of calculating these directly from available data. The methods used for evaluations are based initially on estimates of registration completeness or, for survival analyses, an assessment of the quality of follow-up data. The completeness of incident case registration is estimated on the basis of the mortality/incidence procedure, which assumes a largely constant relationship between the mortality and incidence of a cancer type among people of the same age and sex across federal states. Inclusion criteria for consideration of registry data in national survival analyses are less than 15% of death certificate only (DCO) cases and plausible survival for patients with pancreatic cancer or metastatic lung cancer. Of the 477,300 incident cancer cases estimated for 2010, 429,900 were reported by the cancer registries (90%), and ten federal states contributed data to national survival estimates.