Spatial and temporal variation in socio-economic and environmental factors relevant to mortality and morbidity in Hong Kong

Abstract

Worldwide studies demonstrate that living in socio-economically disadvantaged areas, commonly indicated by income level, is associated with poorer health. Although researchers in Hong Kong have examined the impact of income inequality on mortality at area level, an income inequality indicator is not appropriate for use in this context, because the territory is partitioned into small, heterogeneous areas regarding income. A better understanding of the relationship between mortality and income is required.

The first study examined temporal variations (1981-1985 and 2006-2010) in mortality by median monthly household income at the Small Tertiary Planning Unit level. It showed that residents of wealthy areas had lower mortality than those in poor areas in terms of standardised mortality ratios (SMRs), excess death and total years of life lost, but not life expectancy. Improvements in social policy formulation, universal education and public health care may improve life expectancy. The results suggested a need for a broader investigation of neighbourhood health.

Accordingly, in the second study, the first multiple deprivation index for use in Hong Kong was developed by Factor Analysis in order to be able to measure and monitor health inequalities. The impact of deprivation on mortality was examined by using Bayesian hierarchical approach. The generated multiple deprivation index had 7 domains and 22 socio-economic and environmental indicators with appropriate weighting. The most important domain was barriers to services, while income-related domains accounted for more than 50% of the weighting. The spatial pattern of mortality was random. Deprivation did not have a significant impact on neighbourhood SMRs, but people living in areas characterised by lower education levels and more people working in elementary occupations were more likely to die prematurely.

The unexpectedly weak relationship between deprivation and mortality prompted an investigation into whether deprivation influences other health outcomes. Thus, the third study examined the impact of deprivation on neighbourhood public hospitalisation. Areas with higher neighbourhood standardised hospitalisation ratios were located in the central city and urban areas. Children aged under five living in more deprived areas and those with pneumonia as a principal diagnosis were more likely to be hospitalised. Deprivation in terms of the housing and income factors was positively associated with hospitalisation, while the factors of barriers to services, demographic and living environment were negatively associated.

The strong relationship between deprivation and hospitalisation suggested that resource allocation and the efficiency of the public hospital system is also an important issue. Hence, the last study investigated the change of the number of total hospitalisation days per person in 2000-2010 using a decomposition method. The mean duration of stay contributed most to the reduction in total hospitalisation days. The side effects of the Hong Kong health care service, such as increasing public hospitalisation rates and days, gradually emerged from the data.

The findings reported in this PhD thesis will draw policymakers’ attention to spatial heterogeneity of health outcome, the need to evaluate the hospitalisation system, develop community-based and universal public health care policies and tackle service provision and allocation issues in the near future.