Survival rates

• What is the difference between absolute and relative survival rates?
• What does “period method” mean?
• Why was only the data of certain registries used?

What is the difference between absolute and relative survival rates?

The results of survival analyses in "Cancer in Germany” and in the interactive database describe the average survival prospects of adults after a certain cancer diagnosis. Absolute and relative survival rates were calculated for this purpose.

Absolute survival rates represent the proportion of persons who are still alive after a certain amount of time after their diagnosis. For example, an absolute 5-year survival rate of 80 percent means that 80 out of 100 people with a certain type of cancer have survived the first five years after their diagnosis.

Relative survival takes into account the fact that not all deaths among cancer patients are due to cancer. For this purpose, lifetables from the German Federal Statistical Office are used. These reflect the survival probabilities in the general population in Germany according to age, sex and calendar year. Relative survival rates approximate cancer-related mortality as the ratio of the absolute survival of cancer patients to the survival in the general population of the same age and sex (i.e., the expected survival).

For example, a relative 5-year survival of 80 percent means that five years after a cancer diagnosis, the proportion of survivors is 80 percent of the proportion expected from the lifetables of the general population of the same age and sex over the same period. The relative survival is always higher than the corresponding absolute survival. The difference between absolute and relative survival increases with the mean age at diagnosis of the cancer in question. The expected survival was calculated using the so-called Ederer II method.

to the top

What does “period method” mean?

In order to estimate up-to-date survival chances, the so-called "period method“ developed by Brenner was used. This method takes into account information on the survival of persons who have lived during a certain period of time (e.g., 2015-2016). The calculation of 5-year survival for the period 2015 to 2016 thus includes data from all persons diagnosed between 2010 and 2016 and who had not died before the beginning of 2015. The result can be interpreted as an estimate of the survival prospects of people who were diagnosed with cancer between 2015 and 2016. Naturally, 5-year survival for these individuals cannot be directly determined until 2021. Assuming a continuous improvement in survival rates over time, this method would slightly underestimate the chances of survival.

to the top

Why was only the data of certain registries used?

Because the accuracy of survival analyses depends on the quality of the underlying data, registries must meet two criteria in order to be included in the current survival calculations. On the one hand, the DCO proportion over all primary cancers (C00-C97 excluding C44) should not exceed 15% during the period considered. The unknown diagnosis date of a DCO case, i.e., a case for which a death certificate is the sole source of information, usually leads to the exclusion of this case from survival analysis. Studies suggest, however, that this leads to a potential overestimation of survival rates, as DCO cases are associated with patients with shorter survival times on average.

The second criterion reflects the quality of vital status follow-up of patients with regularly registered cancers. International studies show that people diagnosed with pancreatic or metastasized lung cancer have a very poor prognosis. Moreover, the average survival rate for these diagnoses has not changed substantially over a long period of time. Therefore, cancer registries demonstrating a high proportion of survivors with these cancers may have deficiencies in data quality, leading to a relevant number of "missed" deaths. Therefore, survival analyses only included data from registries in which patients diagnosed with pancreatic cancer or metastasized lung cancer had an average 5-year relative survival of a maximum of 8 percent.

According to the two criteria above, the cancer registries of Bavaria, Bremen, Hamburg, Lower Saxony, North Rhine-Westphalia (only the Münster administrative district), Saarland and Schleswig-Holstein were included in the current analyses.