Comparison of Bayesian estimates of health care costs for Czech and Slovakian insurance companies

Abstract

This paper presents an application of the Bayesian inference to estimate health care expenses by health insurance companies in the Czech Republic and Slovakia. In particular, the Bühlmann-Straub model is applied onto real data, covering time period 2010-2015, from all seven Czech and three Slovakian health insurance companies in order to calculate the credible health care costs per person and the corresponding credibility factors for each company. Computed credible health care costs per person are compared with the real estimations of health care costs per person quoted by analyzed health insurance companies. These real estimates are not available for all the Slovakian health insurance companies and two Czech health insurance companies. The results of Bühlmann-Straub model show that while the largest insurance companies can reliably predict the health care costs for the subsequent time period based on its own statistical data, the smaller companies can find out credible estimation of health care costs per person, taking into account data from all comparable healthcare insurance companies.