Understanding health inequalities in Wales using the Blinder-Oaxaca decomposition method

Abstract

Background: Throughout Wales and the world, health inequality remains a problem that is interconnected with a wider and complex social, economic and environmental dynamic. Subsequently, action to tackle inequality in health needs to take place at a structural level, acknowledging the constraints affecting an individual's (or community's) capability and opportunity to enable change. While the ‘social determinants of health' is an established concept, fully understanding the composition of the health gap is dependent on capturing the relative contributions of a myriad of social, economic and environmental factors within a quantitative analysis. Method: The decomposition analysis sought to explain the differences in the prevalence of these outcomes in groups stratified by their ability to save at least £10 a month, whether they were in material deprivation, and the presence of a limiting long-standing illness, disability of infirmity. Responses to over 4,200 questions within the National Survey for Wales (n = 46,189; 2016–17 to 2019–20) were considered for analysis. Variables were included based on (1) their alignment to a World Health Organization (WHO) health equity framework (“Health Equity Status Report initiative”) and (2) their ability to allow for stratification of the survey sample into distinct groups where considerable gaps in health outcomes existed. A pooled Blinder-Oaxaca model was used to analyse inequalities in self-reported health (fair/poor health, low mental well-being and low life satisfaction) and were stratified by the variables relating to financial security, material deprivation and disability status. Results: The prevalence of fair/poor health was 75% higher in those who were financially insecure and 95% higher in those who are materially deprived. Decomposition of the outcome revealed that just under half of the health gap was “explained” i.e., 45.5% when stratifying by the respondent's ability to save and 46% when stratifying by material deprivation status. Further analysis of the explained component showed that “Social/Human Capital” and “Income Security/Social Protection” determinants accounted the most for disparities observed; it also showed that “Health Services” determinants accounted the least. These findings were consistent across the majority of scenarios modeled. Conclusion: The analysis not only quantified the significant health gaps that existed in the years leading up to the COVID-19 pandemic but it has also shown what determinants of health were most influential. Understanding the factors most closely associated with disparities in health is key in identifying policy levers to reduce health inequalities and improve the health and well-being across populations.