Socio-economic inequalities in survival from screen-detected breast cancer in South West England: population-based cohort study.

BACKGROUND Socio-economic inequalities in breast cancer survival have been reported worldwide, but whether these exist in screen detected as well as symptomatic women has not been established. Making this distinction will allow inferences about the relative contributions of pre- and post-diagnostic delay to these inequalities. METHODS Screening-eligible women diagnosed with breast cancer in South West England (2002-06) were followed-up to 2007. Five-year relative survival ratios (RSRs) were calculated for each deprivation quintile, using deprivation-specific life-tables and a period approach. The 'deprivation gap' in survival was calculated as the slope index of inequality between least and most deprived women. RESULTS The study included 11 018 women, of whom 1176 died during follow-up. Screening status of 54% of women was missing. A clear gradient in survival across deprivation groups ranged from 83.6% [95% confidence interval (CI) 80.0, 86.6] in the most deprived to 90.8% (95% CI 89.0, 92.3) in the least deprived group. Comparing the most deprived to least deprived women, the estimated deprivation gap was -9.42% (95% CI -12.80, -6.04, P = 0.003). Among screen-detected women, inequalities were attenuated, but persisted, ranging from 95.6% (95% CI 90.6, 98.0) in the most deprived to 98.2% (95% CI 95.9, 99.2) in least deprived; the estimated deprivation gap was -3.03% (95% CI -5.75, -0.85, P = 0.023). CONCLUSION The deprivation gap in survival does not appear as marked with screen-detected breast cancer as the other groups, though still apparent. Efforts to eliminate inequalities should consider both increasing breast screening participation and ensuring equal access through secondary care systems for women of lower socio-economic position.