Estimating the Crude Probability of Death due to Cancer using French Registry Data

Tuesday, 19 August 2014
Exhibit hall (Dena'ina Center)
Nadine Bossard, PhD , Université Claude Bernard Lyon 1, VILLEURBANNE, France
Laurent Remontet, MS , CNRS, UMR5558, VILLEURBANNE, France
Florence Binder-Foucard, PhD , Reseau des Registres Francais des Cancers FRANCIM, TOULOUSE, France
Laetitia Daubisse, PhD , Reseau des Registres Francais des Cancers FRANCIM, TOULOUSE, France
Aurelien Belot, PhD , Université Claude Bernard Lyon 1, VILLEURBANNE, France
Hadrien Charvat, PhD , CNRS, UMR5558, VILLEURBANNE, France
INTRODUCTION: Though net survival is an important epidemiological tool allowing temporal or geographical comparisons, it cannot give information on the “real” survival of cancer patients. Indeed, net survival is the survival that would be observed if cancer were the only possible cause of death. In this work, we provide estimates of the “real”, or “crude”, probabilities of death due to cancer and due to other causes as well as the probability of being alive up to ten years after diagnosis according to the age and year of diagnosis.

METHODS:  We used data from the French network of cancer registries and studied five common cancers: head and neck, breast, prostate, lung and colorectal cancers.Crude probabilities for a particular cause were obtained by combining the hazard specific of that cause and the overall survival. Cancer-specific hazard was estimated from a flexible excess hazard model, also known as relative survival model, with département, age and year of diagnosis as covariates. Other-cause-specific hazard was obtained from general population life tables.

RESULTS:  For breast, prostate and colorectal cancer, the impact of other causes of death on the total probability of death becomes more important as the age at diagnosis increases while it stays negligible for lung and head and neck cancer whatever the age at diagnosis. For breast, prostate and colorectal cancer, there is a strong decrease in the probability of death due to cancer for patients diagnosed more recently.In elderly patients with prostate cancer, few patients die from their cancer although the cancer-specific hazard is high; indeed, they die from other causes.

CONCLUSIONS:  The crude probability of death is an intuitive concept that may prove particularly useful in the appraisal of the real impact of the cancer as well as in the choice of an appropriate treatment or in the indications of a screening strategy by allowing the clinician to estimate the proportion of cancer patients who really die from their disease.