Commentary: Inequalities in cancer screening programmes

Abstract

Screening for cancer is intuitively attractive. Well-run cancer screening programmes can save lives, reduce morbidity, provide reassurance to individuals about their health and encourage a focus on prevention and early detection. Despite the intuitive appeal, the harmful effects of screening (both potential and actual) are well documented. Harm to an individual includes over-diagnosis and treatment of questionable abnormalities, anxiety for those with false positive results and false reassurance for those with false negative results. One of the important, but often not well articulated, harmful effects at a population level is the potential for cancer screening to increase health inequalities between population groups. Screening can be divided into two types. First, there are ‘organized screening’ programmes, which work within pre-agreed structures, policies and standards, and typically focus on mortality reduction for screened individuals. Secondly, there is ‘opportunistic screening’ where screening occurs either as a result of a request from an individual or from contact with a health professional who offers the screening test. In this issue, Palencia et al. combine individual-level data from the WHO World Health Survey covering 22 European countries, with information about the organizational structures of breast and cervical cancer screening programmes within countries, to examine in multi-level analyses whether socioeconomic disparities in breast and cervical screening participation are affected by the type of screening offered. They found that for breast cancer screening, there was higher participation in screening in countries with organized programmes, but no such pattern was seen for cervical cancer screening. For both breast and cervical cancers, socio-economic inequalities in participation measured both on relative and absolute scales were more likely to be found in countries without organized screening programmes. One of the arguments for organized screening is that participation rates are likely to be optimized. Therefore, the finding that cervical cancer screening participation was not affected by the type of screening was surprising. It is worth noting that, while in the study by Palencia et al. participation in cervical screening was measured using a cut-off point of screening in the previous 3 years, all five countries with national (organized) cervical screening programmes recommend intervals of 5 years for at least some women (in the UK, Denmark and Sweden recommendations vary by age). In contrast, almost all of the countries included in this study with opportunistic screening recommend screening intervals of 1–3 years. As a consequence, it is possible that screening participation was underestimated in countries with organized screening. This is not true for breast cancer screening as almost all countries, regardless of how screening is organized, recommend screening at either a 2or 3-year interval. Four other important points should be made. First, while the paper by Palencia et al. focuses on socioeconomic inequalities determined by education level, such inequalities can occur on many axes e.g. ethnicity, gender, geography, age, sexuality and disability. Although most social determinants of health are not controlled by the health sector, there is increasing evidence that health-care systems, including screening programmes, can actively alter inequalities in health outcomes. The ‘inverse equity hypothesis’ describes how inequalities in health care occur as new health interventions are introduced. This theory suggests that as a new intervention is introduced into a population, those that are most deprived, and usually most in need of the intervention, are most likely to take it up later and at a lower rate than more privileged groups. An example of this hypothesis in action occurred with the introduction of the breast cancer screening programme in Published by Oxford University Press on behalf of the International Epidemiological Association