The utilization of canonical analysis in identifying factors influencing hospitalization rates on the regional scale

Abstract

The main goal of presented research is to show the possibilities of ca- nonical analysis utilization in evaluations of intensity and spatial distribution of factors influencing hospital prevalence rates in regions. The setting of this study is the moun- tainous areas of Podkarpackie Province divided into municipalities, forming a total of 217 spatial units. The major reason for utilizing canonical analyses is the large variation of hospital prevalence rates within selected groups of diseases (cardiovascular and respiratory) and between different age groups. As a consequence, hospitalization rates by age groups are included in the analysis as dependent variables, whereas the independent variables used are various socio-demographic as well as standard of living indicators. These independ- ent variables are grouped into three sets (demographic, social and economic), which characterize municipalities under investigation. The mean for the years 2006-2008 for each variable is calculated. The product of the canonical analysis, canonical variables with the highest correla- tion to independent variables, are subsequently presented on cartograms showing spatial distribution of interdependence between the examined phenomena. The final part of this study includes a short evaluation of the method used for the explanation of spatial hospitalization rate inequalities, as well as some possibilities for enhancement of research into spatial variation of ill-health on the regional scale.