Early Childhood Dental Problems: Classification Tree Analyses of 2 Waves of an Infant Cohort Study.

Abstract

Investigations into the wider bioecological understanding of dental problems in early childhood are limited in national surveys. Classification tree analysis (CTA) was used to explore multilevel interactions among key aspects of child and primary caregiver (PCG) psychosocial and physical health affecting dental problems in preschool children. Data were derived from the Growing Up in Ireland study, a nationally representative sample of 9-mo-olds (N = 11,134) in 2007/2008 followed up at age 3 y (N = 9,793) in 2010/2011. Analysis included PCG reports of children's dental problems, general health, temperament, emotional and behavioral difficulties, and their own general health, stress and depression, relationship, and sociodemographic variables. Misclassification costs were specified for the model by applying a higher penalty for misclassifying those with a dental problem (minority class). Logistic regression analyses were carried out for comparison. Dental problems were reported among 302 infants (2.7%) at 9 mo of age and 493 children (5.0%) at 3 y. CTA identified infant temperament (Infant Characteristics Questionnaire unpredictable) as the primary predictor of dental problems at 9 mo and child global health at 3 y of age. First-level predictors were PCG depression score and use of a soother at 9 mo and PCG ethnicity and unscheduled hospital visits at 3 y of age. Regression analyses results supported the most important predictors at 9 mo and 3 y of age. The CTA model for 9-mo-old infants had a specificity of 90.4%, sensitivity of 31.2%, and overall accuracy of 88.8% while that for 3-y-olds had a specificity of 58.5%, sensitivity of 66%, and overall accuracy of 59%. Key aspects of infant/child and PCG health, as well as psychosocial characteristics associated with reported dental problems, should be considered in future multidisciplinary approaches to child health. Knowledge Transfer Statement: The results of this data analysis should help raise awareness among clinicians of how primary caregiver and child psychosocial and general health factors are associated with early childhood dental problems, even before the primary dentition is complete. Classification tree analysis visually demonstrates how factors such as infant temperament (9 mo) and child global health (3 y) can interact at multiple levels and affect different subgroups of the child population. Future intervention strategies for oral health should involve consideration of the psychological and general health characteristics of the young child and PCG at both the patient and population levels. This knowledge could assist decision makers adopt an integrated multidisciplinary approach in formulating a coherent oral health policy for preschool children.