G Ekbäck1, C Persson. 1. Department of Dentistry, Orebro County Council, Orebro, Sweden. gunnar.ekback@orebroll.se
Abstract
AIM: This study assessed the prevalence of socio-demographic clusters in a Swedish county and the relationship of socio-demographic clusters and caries. METHODS: All 2-19-year-olds (n = 58,573) who attended a routine check-up in Orebro County in 2005-2007 were involved in this study. Initially, two-stage cluster analyses were used to identify outliers. Secondly, the Ward method which is a hierarchical clustering method was used to conduct the final analysis. Bivariate logistic regression was also used to study the relationship between cluster membership and caries. The smallest study unit used in the initial analysis for geographical area is known as key code area, which is a geographical entity defined by the municipalities themselves. Decayed surface (DS/ds) has been used as a measure of dental caries. RESULTS: The county of Orebro clustered in five different socioeconomic clusters. Each cluster was defined by proportion of people over 75 years, native-born, single parents, and those with low incomes and low level of education. Odds ratio (OR) for having DS/ds > 0 in the last dental check-up during 2005-2007 was 1.5 (cluster 1), 1.3 (cluster 2), 1.4 (cluster 3) and 3.8 (cluster 4) compared with the most socioeconomically favoured cluster (cluster 5). CONCLUSION: Cluster analysis of socioeconomic data is a useful tool to identify neighbourhoods with different socio-economic conditions.
AIM: This study assessed the prevalence of socio-demographic clusters in a Swedish county and the relationship of socio-demographic clusters and caries. METHODS: All 2-19-year-olds (n = 58,573) who attended a routine check-up in Orebro County in 2005-2007 were involved in this study. Initially, two-stage cluster analyses were used to identify outliers. Secondly, the Ward method which is a hierarchical clustering method was used to conduct the final analysis. Bivariate logistic regression was also used to study the relationship between cluster membership and caries. The smallest study unit used in the initial analysis for geographical area is known as key code area, which is a geographical entity defined by the municipalities themselves. Decayed surface (DS/ds) has been used as a measure of dental caries. RESULTS: The county of Orebro clustered in five different socioeconomic clusters. Each cluster was defined by proportion of people over 75 years, native-born, single parents, and those with low incomes and low level of education. Odds ratio (OR) for having DS/ds > 0 in the last dental check-up during 2005-2007 was 1.5 (cluster 1), 1.3 (cluster 2), 1.4 (cluster 3) and 3.8 (cluster 4) compared with the most socioeconomically favoured cluster (cluster 5). CONCLUSION: Cluster analysis of socioeconomic data is a useful tool to identify neighbourhoods with different socio-economic conditions.