Role of participation in organized screening on the survival of women with breast cancer according to socioeconomic disparities: A population-based study

Abstract

Several observational studies have suggested the impact of socioeconomic disparities in breast cancer diagnosis. Living in a deprived area was associated to advanced tumor stage at diagnosis and poor prognosis. The participation in organized breast cancer screening (OBCS) program can help to detect cancers at an early stage and to reduce mortality. This study aimed at describing survival of women with cancer according to their participation in OBCS and socioeconomic characteristics in the department of Gironde. All cases of breast cancer in situ or invasive diagnosed over the period 2008–2010, in women aged 50 to 74 years were extracted from Gironde Cancer Registry. The participation in OBCS program was obtained from the Gironde Screening Coordination Structure (AGIDECA). Women were categorized in 4 groups according their participation in the OBCS. The socioeconomic environment was measured by an aggregated indicator of deprivation: the French European Deprivation Index (EDI). EDI was categorized in quintiles (quintile 1: women most affluent and quintile 5: women most deprived). We estimated net survival rates at 1 and 5 years using the Pohar-Perme method according to the participation in OBCS and socioeconomic deprivation. A correction of the advance bias at diagnosis (or lead time bias) was taken into account by modifying the survival time of the screened women according a method established by Duffy et al. During the 2008–2010 period, 2,244 women were diagnosed with first breast cancer in Gironde department. In this population, 297 deaths were reported. Among all these women, 57% had participated in OBCS, 24% were non-screened, 14% had interval cancer (cancer diagnosed between 2 screening test) and 5% were diagnosed more than 2 years after the last screening test. Net survival was higher in the group of screen-detected women compared to non-screened women (97.3% vs. 83.8%, P < 0.0001). Non-screened women had lower net survival when they were deprived (quintile 1: 91.1% vs. quintile 5: 78.1%, P < 0.0106). Taking into account socioeconomic disparities, survival was higher for screened women than for non-screened women, but the difference was significant only in most deprived areas (quintiles 3 to 5). After correction for the bias in advance of diagnosis, the survival rates of screen-detected women were lower than without correction, but they remained higher than those of women who were not screened. Women with breast cancer participating in OBCS in Gironde had better net survival at five years compared to women who were non-screened. The role of OBCS on breast cancer survival seemed to be stronger for women in most deprived areas. The implementation of incentive actions in deprived areas in order to foster women to participate in OBCS could improve their survival in case of breast cancer diagnosis.