Health status at the turn of the century varies widely across nations. To elucidate policies that can have a real impact on improving health outcomes such as life expectancy and child mortality is of particular interest to many researchers and decision makers. Our study builds on and extends beyond existing literature in the range of conditions considered as well as in the statistical procedures used to derive explanatory models. In fact we address the problem from a health ecology perspective, since many factors, such as demographics, economics, politics, education, and culture, are believed to influence health. Our study includes 161 countries representing all possible economical and demographical conditions in the world, and thus is not limited to only affluent nations as most previous studies are.
We gathered data from several reliable sources, and encountered large amounts of missing data. We dealt with this problem using multiple imputation techniques. Each imputed data set was derived through hierarchical multivariate means models. These models appear adequate since countries are naturally grouped by economic and geographical-cultural conditions.
Three country groups were derived in the health outcome (life expectancy/child mortality) space. At the two extremes of health outcomes are a group consisting of mostly sub-Saharan African countries, and a group of mostly affluent countries. The three groups are reflected and also detected in demographics (death rate/population dependency) space. This finding hints a tight relationship between health outcomes and demographics conditions. The extend of this relation, as well as the strength of the relationship between other factors such as economics, education, and health performance, and health outcomes, was confirmed through Bayesian hierarchical multiple linear regression models.
This work is joint with Sue Thomas Hegyvary and Devon Berry.
