Les séminaires de recherche de l'Irdes constituent un outil de présentation, de discussion scientifique et de diffusion de travaux de recherche en cours de réalisation dans le domaine de l'économie de la santé. Ces séminaires se tiennent dans les locaux de l'Irdes et sont ouverts à l'ensemble de la communauté scientifique.

Durée d'une séance : 1h30

- 40 minutes de présentation par les auteurs

- 25 minutes de discussion par un chercheur senior

- 25 minutes de discussion avec la salle

Vendredi 17 mars 2017

Social Inequality in Cancer Survival in Canada

Auteurs : Grignon M. (McMaster University Hamilton, Ontario, Canada)

Discutant : Or Z. (Irdes)

A social gradient in the survival period following a cancer diagnosis has long been observed, even in countries with universal publicly funded access to health care services. However, the evidence is based mostly on neighbourhood or area rather than on individual level observations; moreover, it is quite dated or refers to specific cancer sites.
Our estimates are based on an extraordinary Canadian database in which records from 20 percent of respondents in the 1991 Census of Population are linked to 326,000 longitudinal records of ever-diagnosed individuals from the Canadian Cancer Registry; mortality is tracked to the end of 2010.
We estimate Cox proportional hazard regression models to assess the effects of SES on survival after diagnosis, while controlling for a variety of other factors.
We find that survival rates after diagnosis vary with SES: in the lowest educational category, these rates are 15 percent below those in the highest; in the lowest income quintile, they are 17 percent below those in the highest.
The study provides strong evidence that cancer survival is associated with SES: those with higher levels of education and income live longer even after accounting for age, sex, occupation, immigrant status and ethnic origin, region of residence, marital status, tumor site, and cancer stage at the time of diagnosis. The other factors that might explain some of the differences - but which we were unable to take into account - could be lifestyle differences, including smoking, exercise and nutritional practices, as well as treatment differences.