Individual and contextual factors influencing dental health care utilization by preschool children : a multilevel analysis

The effect of contextual factors on dental care utilization was evaluated after adjustment for individual characteristics of Brazilian preschool children. This cross-sectional study assessed 639 preschool children aged 1 to 5 years from Santa Maria, a town in Rio Grande do Sul State, located in southern Brazil. Participants were randomly selected from children attending the National Children’s Vaccination Day and 15 health centers were selected for this research. Visual examinations followed the ICDAS criteria. Parents answered a questionnaire about demographic and socioeconomic characteristics. Contextual influences on children’s dental care utilization were obtained from two community-related variables: presence of dentists and presence of workers’ associations in the neighborhood. Unadjusted and adjusted multilevel logistic regression models were used to describe the association between outcome and predictor variables. A prevalence of 21.6% was found for regular use of dental services. The unadjusted assessment of the associations of dental health care utilization with individual and contextual factors included children’s ages, family income, parents’ schooling, mothers’ participation in their children’s school activities, dental caries, and presence of workers’ associations in the neighborhood as the main outcome covariates. Individual variables remained associated with the outcome after adding contextual variables in the model. In conclusion, individual and contextual variables were associated with dental health care utilization by preschool children.


Introduction
Unequal dental health care utilization by preschool children has been demonstrated in developing countries. 1,2,3,4Previous studies have suggested that some individual factors, such as being older, having a higher income, having one's own house, having access to preventive oral health information, having better educated parents, and being white children, could explain the reasons for seeking dental care services. 3,4,5,6,7evertheless, recent studies have demonstrated the importance of contextual factors for different oral health outcomes, including the use of dental services. 8,9,10,11,12n Brazil, the latest official publication showed regional inequality in dental care utilization (SBBRASIL  2010).It is suggested that the economic and social attributes of the community might influence dental care utilization by people who reside in it. 12Some recent evidence suggests that contextual determinants, such as a higher dentist-to-population ratio, influence dental care utilization by the population. 12aking into account, risk for several conditions in individuals living in a community could be influenced by community context.Therefore, contextual and individual determinants have been used together in multilevel analyses.This allows adjusting individual characteristics by the contextual determinants of communities. 5,13,14Presently, few studies use multilevel analysis to assess individual and contextual factors for dental care utilization. 5,10Moreover, only one study was carried out with preschool children, showing that the use of dental care utilization was associated with individual factors (demographic, dental health, habits, and socioeconomic factors) and with contextual variables (dental care model). 5However, new studies are necessary to investigate other contexts in other regions.
Therefore, the aim of this study is to investigate the effect of contextual factors on dental care utilization after adjustment for individual characteristics of Brazilian preschool children.

Methodology Ethics approval
This study was approved by the Research Ethics Committee of the Federal University of Santa Maria and of the School of Dentistry of University of São Paulo, Brazil.A written informed consent was obtained from all parents.

Study design and data collection
A cross-sectional study was performed with preschool children aged 1 to 5 years from Santa Maria, a town in Rio Grande do Sul State, located in southern Brazil.The sample size was calculated as follows: 5% standard error, 80% power, 95% confidence interval, 10% non-response rate, 2:1 ratio of unexposed to exposed subjects, and a prevalence ratio to be detected of at least 1.8.Moreover, the sample design was adjusted to 1.4 (design effect).The minimum sample size was estimated at 456 children. 4 As the present study was part of a survey in which other outcomes were considered, the final sample size was greater than the minimum required.
Participants were randomly selected from children attending the National Children's Vaccination Day.Fifteen health centers were selected for this research and all of them were equipped with a dental chair.These were the largest health centers in town and more than 85% of the children were vaccinated in these centers during the study.
Fifteen examiners and 30 assistants performed the clinical examination and administered the questionnaire.They had been trained and calibrated before data collection.More information about the methodology and data obtained in this epidemiological survey can be accessed in previous publications. 4,15,16hildren were examined for dental caries in a dental chair.Visual examinations followed the ICDAS criteria. 17,18As the ICDAS criteria are comparable with standard criteria (WHO) in an epidemiological survey of preschool children, 19 we used the ICDAS cutoff point of 3 (0-2 sound, 3-6 carious) to calculate the decayed/missing/filled teeth (dmft) index.Only severity criteria were used for the analysis.In other words, only teeth with frank cavitation were considered.The sample was dichotomized into decayed teeth > 4 and decayed teeth ≤ 4.

Individual-level independent variables
A structured questionnaire was used to obtain individual data such as demographic variables, socioeconomic factors, and data on dental health care utilization.The parents answered questions about age, children's sex, race (black or white), 20 family income (less than 1 Brazilian minimum wage (BMW), between 1 and 2 BMW, between 2 and 3 BMW, and more than 3 BMW) parents' schooling (≥ 8 years of formal education or < 8 years of formal education) and mothers' participation in their children's school activities (yes or no).

Contextual independent variables
Two community-related variables were used to assess the contextual influences on children's dental care utilization: presence of dentists and presence of workers' associations in the neighborhood (yes or no).These covariates have been previously used as community factors and community-related variables. 12,13

Outcome variable
The outcome was measured by a question from the questionnaire: "Has your child ever visited a dentist?" (Yes or No).

Data analysis
The STATA 12.0 software (Stata Corporation, College Station, USA) was used for data analysis.Unadjusted and adjusted multilevel logistic regression models were used to describe the association between outcome and predictor variables.The odds ratio (OR) and the respective 95% confidence intervals (95%CI) were calculated.
In the analysis, children (first level) were nested in neighborhoods (i.e., one of the 15 health centers) (second level).In the first stage, an unconditional model ('null' model) estimated the basic partition of data variability between two levels before individual and community characteristics were taken into account. 21The second model (Model 2) added covariates at the individual level; the ''full'' final model (Model 3) included individual factors and contextual covariates.Variables with a p-value ≤ 0.20 in the unadjusted analyses were entered into the adjusted models.They were retained and considered statistically significant in the final models only if they had a p-value ≤ 0.05 after adjustment.

Results
A total of 639 children, 321 males (50.2%) and 318 females (49.8%), participated in this study (91.3% participation rate).Failure to participate was mostly attributed to denied permission by guardians.
Participants were predominantly white and more than half of the caregivers had a high level of education.Most parents had a household income ≤ 3 BMW.A prevalence of 21.6% was found for regular use of dental services (Table 1).Table 2 shows the unadjusted assessment of the associations of dental health care utilization with individual and contextual variables.The analysis included children's ages, household income, parent's schooling, mothers' participation in their children's school activities, dental caries, and presence of workers' associations in the neighborhood as the main outcome covariates.
After adjusting for the individual covariates (Table 3 -Model 1), children's ages, household income, mother's schooling, mothers' participation in their children's school activities, and dental caries were identified as individual determinants of dental health service utilization.These variables remained associated with the outcome after adding the contextual variables in the model (Model 3).A higher use of dental health services was observed among those living in neighborhoods with workers' associations.

Discussion
International guidelines recommend that infants have an initial oral health evaluation in the first year of life. 22,23Evidence suggests that an early visit to the dentist (in the first year of life) should help prevent dental disease. 6,24Moreover, early educational oral health programs for mothers could increase the use of dental services by preschool children. 25However, previous studies showed that dental care utilization by preschool children is low and is associated with socioeconomic factors. 3,4,5,26n our study, socioeconomic factors were related to the use of dental care, as reported by other authors. 3,4,5,26he current study considered household income and parents' schooling as proxies for socioeconomic status, which were identified as individual determinants for the outcome.Notwithstanding, only mother's schooling was fitted into the final model after adjustment for contextual variables.Children whose mothers did not complete primary education were less likely to use dental services.It is known that poor educational level may lead to low income, unemployment, and poor occupational status.Financial costs and low level of information about the importance of oral health could be a hindrance to dental care. 27Table 3 shows that socioeconomic variables affected the use of oral health services at both individual and contextual levels.This finding confirms previous reports on the importance of investigating the impact of contextual factors on oral health outcomes. 8,9,10,28oreover, most children who participated in our study had their first dental visit after the age of 4 years.Only 14 children went to the dentist before their first year of life, as recommended by international guidelines. 22,23These findings could be related to socioeconomic status and availability of dental care. 29,30In other words, a social gradient in health is present in dental service utilization.
T h i s st udy a l s o ve r i f ie d t h at mot he r s' participation in their children's school activities positively influenced dental care utilization.Participative mothers were more likely to take their children to the dentist.We suggest that mothers who keep track of the school activities are more concerned with everything that involves their children, including oral health.Socioeconomic factors could be part of this relation.Generally, mothers with a higher family income are better informed and can afford to monitor their children in all aspects.Therefore, participation in children's school activities could be strongly modulated by socioeconomic status.Nevertheless, further studies are needed for comparison of our results.To our knowledge, no earlier study had investigated this association in the literature.
Furthermore, a significant effect of social context on dental care utilization by preschool children was verified in our analysis.In the final model, the number of workers' associations located in the neighborhood could be noted at the contextual level, as children who lived in areas with workers' associations were more likely to go to the dentist.This contextual variable could be related to social support.Neighborhood connection, such as time spent with friends that attend the meetings of an association can be a proxy for social support and social capital. 31,32We suggest that frequency of contact with friends may reduce social isolation, which plays an important role in maintaining oral health.In other words, oral health behaviors can be influenced by social controls.Participation in workers' associations implies interaction between individuals, which may suggest the transmission of good ideas and healthier choices. 13,33egarding contextual variables, in this study, the prevalence of dental health care utilization was not influenced by the presence of dentists in the neighborhood (Table 3), which is in disagreement with a previous study 12 that analyzed data from the six largest metropolitan counties in Ohio, USA.Our study was conducted only in Santa Maria, Brazil, where there are a large number of dentists per capita.Almost all the participants in our study lived in areas where there were dentists.This could possibly explain our findings.
The cross-sectional design could be a possible limitation of our study.Therefore, our results should be interpreted along with the limitations of this type of study.Moreover, the number of workers' associations was provided by the local authorities and it is possible that it might not be perfectly correct, despite the fact that this information has been used in official local publications.
In our analysis, we used the multilevel approach to verify the extent to which the outcome is accounted for by contextual and individual variables.We adjusted the models according to the region where the children lived.Recent research has reported that contextual variables, such as city-level variables (suburbs with the worst social exclusion indices; socially-deprived areas), can negatively influence oral health outcomes. 34,34Therefore, assessing the characteristics of the neighborhood in which children live is important for the investigation into oral health inequalities.

Conclusion
In conclusion, individual and contextual variables are associated with dental health care utilization by preschool children.Therefore, these variables must be taken into consideration for redirection of resources allocated to public health and for the formulation of oral health policies.

Table 1 .
Individual and contextual level characteristics of the sample.

Table 2 .
Unadjusted assessment of the association of dental health care use among preschool children with individual and contextual-level variables.

Table 3 .
Multilevel Adjusted Assessment of dental health care use among preschool children associating individual and contextual variables.