Participation and retention in the breast cancer screening program in New Brunswick Canada

Abstract

New Brunswick (NB) Canada uses its breast cancer screening service program to assess the extent to which eligible NB women are complying with mammography guidelines. While many studies have investigated factors associated with participation in periodic breast cancer screening in Canada and elsewhere, most work has relied on self-reported surveys or smaller scale primary data collection. Using a longitudinal administrative dataset for NB over the period 1996–2011 of 255,789 eligible women aged 45–69, this study examined demographic, socioeconomic and geographic factors associated with initial participation in regular screening at age 50 and ongoing retention in the program. Logistic regression was used to examine correlates of initial screening, while rescreening participation was estimated using survival analysis accounting for rescreening episodes. Initial screening participation was lower for women born outside of NB, many women living farther away from screening centers, women in rural areas, and higher for married women. In contrast, retention was higher for rural women and women recently arrived in NB. For both participation and retention, regional disparities across health zone persisted after controlling for observable personal and locational factors. The analysis highlights important characteristics to be targeted to increase screening but also that how health zones operate their screening programs exerts a very significant effect on the use of screening services by eligible women. This offers lessons for the design and evaluation of any cancer screening program.

Highlights

• •

  We analyze a linked longitudinal dataset on mammography screening from New Brunswick Canada.

• •

  Screening at entry into the screening program at age 50 was lower for immigrants and rural residents.

• •

  Rescreening conditional on previous screening was higher for rural residents and recent arrivals.

• •

  Regional disparities in screening rates across health zone remained after controlling for covariates.