Breast Cancer Incidence and Socioeconomic Status in Canada

Background

Cancer surveillance data are critical to inform public policy. Differences in cancer incidence and mortality by indicators of socioeconomic status can be used to detect regional differences and may be helpful in determining the effects of public health interventions. National data on breast cancer incidence by socioeconomic status would be useful to inform public policy, and would allow calculation of case-fatality rates in addition to population-based incidence. Breast cancer is particularly interesting because of the numbers of Canadians affected, as well as the possibly inverse association with socioeconomic status.

Objectives

1. Pilot methods for national Canadian Cancer Registry data access via a Statistics Canada Research Data Centre.

2. Apply postal code linkage techniques to develop a novel method for measuring Canadian breast cancer incidence by location and by socioeconomic status.

3. Develop a standardized methodology for subsequent development of similar tools for analysis of data on colorectal cancer, lung cancer, cervical cancer, and other cancers.

Methods

From the Canadian Cancer Registry data within the Research Data Centre, we will select all cases of invasive breast cancer from 1985 (earliest year when postal code missing values are no more than 5%) to 2005 (latest year when complete outcomes data are available), excluding those with missing or invalid postal code. Ethical review was carried out by the BCCA/UBC Research Ethics Board. We will modify our previously-developed software to link postal code at diagnosis from these data with enumeration area / dissemination area (EA/DA) and then to socioeconomic status variables (SES) available by EA/DA. Population data on SES by EA, census year (1986, 1991, 1996, 2001, 2006), and 5-year age groups will be obtained from Statistics Canada. We then plan to report incidence of breast cancer by year, province, urban/rural categories, and quintile of socioeconomic status as estimated by neighborhood income. We will calculate the relative risk of breast cancer by quintile of socioeconomic status. Finally we plan to leverage this project by developing tools and a handbook for conceptually similar projects to be done in the future through access to the cancer incidence data at this or other RDCs: colorectal cancer with respect to screening program implementation; cervical cancer with respect to screening program or vaccine program implementation; lung cancer with respect to regional differences in workplace or public place smoking, regulation and taxation.

Research Team

• John Spinelli
• Zenaida Abanto
• Amy MacArthur
• Lydia Xu