Issues in Multivariate Spatial Analysis of Multiple Diseases Using Complex Health Survey Data

Most spatial analyses of health survey data have employed univariate disease mapping approaches for specific diseases even if the studied diseases are epidemiologically interrelated. Recent extensions have included joint disease mapping of several interrelated health data has both epidemiological and statistical advantages including identifying specific and shared spatial risk patterns and improvement in the statistical power by borrowing strength from related health data variables within the same area and neighbouring areas. Also, despite recent methodological research in spatial analyses of complex health survey data, a majority of the analyses do not account for the survey designs. This chapter brings these two issues in an integrated approach with an application to estimate the geographic distribution of childhood-related diseases in sub-Saharan Africa based on health survey data.