An analysis of the relationship between healthcare spending and health outcomes: a data analytics perspective using the theory of production functions

This research investigates the relationship between per capita spending on healthcare and population health outcomes at the provincial level in Canada using data from 1980 to 2010. The health outcomes examined include life expectancy at birth and at age 65, number of infant deaths, and potential years of life lost from treatable causes, all of which are separated by gender. Using analytics methods as an application of the theory of growth accounting, the study evaluates the performance of the provincial health care systems in terms of their ability to efficiently produce longevity. The study also specifies the categories of healthcare spending which are most influential in determining the efficient production of longevity and measures the contribution of healthcare spending to the determination of infant mortality and deaths from treatable causes. The methods employed include Data Envelopment Analysis, Decision Tree Induction, and Multivariate Adaptive Regression Splines. The results of the analysis point to the fact that Canada’s provinces operate inefficiently in their production of health outcomes and confirm the importance of healthcare spending to determining health outcomes in Canada.