Estimating Determinants of Healthcare Utilisation: A Nonparametric Approach

Abstract

The goal of this paper is to propose a nonparametric analysis applied to medical schemes data. Because medicalschemes data are aggregated and have heavy tailed densities, we consider nonparametric density estimation,which overcomes limitations related to aggregation bias and heavy tails in the data. We demonstrate theusefulness of this class of models by using an empirical example that seeks to predict healthcare use. This casestudy can be used by researchers in health-care as well as in intermediate course in data analysis in order tohighlight the importance of distribution-free models in the context of a real data set.