Implementing Bayesian Learning in Assessing Performances of Healthcare System and Population Health in Three Transitory European Countries

Abstract

This paper analyses the effect of key healthcare system indicators on the population health measured with the proxy variable, life expectancy at age 65, within three transitory European countries. The analysis is implemented in terms of the Bayesian TVCVAR model. The data sources are mainly from WHO, World Bank and United Nations. This study is based on the period from 1991 to 2020. The model results for Croatia correspond to the notion that the LE65+ and NM, CHEGDP, PH move in the same direction while the LE 65+and HB move in the opposite direction. The findings for Hungary reflect that NM, HB, CHEGDP and the LE65+ move in the same direction except LE65+ and PH which move in the opposing direction. The results for Romania reveal that CHEGDP, PH and LE65+ move in the same direction and that direction of NM, HB and LE65+ is time-varying. The BTVCVAR output found time-varying effect on LE65+ only for Romania. In Croatia and Hungary there was found a time-invariant effect on LE65+ by the parameters in the model. These results confirm the presence of strong relationship between population health and the key healthcare indicators that are representing the healthcare system performances within these three European countries.