Social gradient in child oral health: individual, school and area variation

Abstract

This thesis describes the oral health of New South Wales (NSW) children aged 5-12 years by socioeconomic (SES) characteristics utilising the individual-, school- and area-level socioeconomic indicators. It also quantifies the usefulness of SES indicators for targeting of dental services. Methods A cross-sectional study of NSW 5–12 year-olds was conducted in 2007 using a multi-stage, stratified, cluster sample approach. Explanatory SES variables were explored at three levels: individual, school and area. Caries prevalence, caries severity and significant caries were calculated. Bivariate analysis was undertaken. Prevalence ratios (PR) of caries prevalence and SiC₁₀ were modelled by Poisson regression (PROC LOGLINK, SUDAAN 10.0). Rate ratios (RR) of caries severity were modelled using Poisson regression (PROC GENMOD, SAS 9.2). Multi-level analysis (SAS PROC GLIMMIX) was undertaken accounting for the nested structure. Use of SES variables to target dental services was examined using number of cases, relative risk and population attributable fraction (PAF%). Results Just under 40% of NSW children had a prevalence of deciduous caries with mean dmfs of 3.18 surfaces and just over 22% had experienced permanent caries with mean DMFS of 0.61 surfaces. Variation in oral health by SES indicators There was significant variation in caries prevalence, caries severity and SiC₁₀ by socioeconomic characteristics; children from the lowest SES category had significantly higher caries prevalence and severity compared to the highest SES category for all SES indicators in both the deciduous and permanent dentition. Membership of the SiC₁₀ group showed lower SES groups had a higher proportion of children who formed part of the SiC₁₀ group. Associations across individual, school and area-level SES indicators In the final models, income was significant for all three caries measures for both dentitions. The children from the lowest income category had significantly higher odds of caries, more severe caries and membership of the SiC₁₀ group. School type as an explanatory factor was not significant for caries prevalence and SiC₁₀ in the multi-level model, although the children attending a disadvantaged public school had significantly higher odds of permanent caries severity. Effectiveness of targeting by SES indicators In both the deciduous and permanent dentition there were fewer cases of caries and SiC₁₀ cases in the designated SES target group, the lowest SES group, than outside the designated target group. SES demonstrated a low population attributable fraction for deciduous and permanent caries prevalence, caries severity and significant caries. Conclusions The study demonstrated that caries was higher among lower SES groups whether measured by individual, school or area characteristics. In many instances there were three and five-fold differences among those in the lowest SES categories providing a consistent association with poor oral health. Income was independently associated with variation in child oral health when adjusting for the nested structure. Low SES categories did not identify the majority of those with caries or the highest levels of caries and would therefore be limited as a basis for a targeted oral health strategy and a population health focus that uses a social determinants approach would be more appropriate.