Comparison between observed children's tooth brushing habits and those reported by mothers

Information bias can occur in epidemiological studies and compromise scientific outcomes, especially when evaluating information given by a patient regarding their own health. The oral habits of children reported by their mothers are commonly used to evaluate tooth brushing practices and to estimate fluoride intake by children. The aim of the present study was to compare observed tooth-brushing habits of young children using fluoridated toothpaste with those reported by mothers.

A sample of 201 mothers and their children (aged 24-48 months) from Montes Claros, Brazil, took part in a cross-sectional study. At day-care centres, the mothers answered a self-administered questionnaire on their child's tooth-brushing habits. The structured questionnaire had six items with two to three possible answers. An appointment was then made with each mother/child pair at day-care centres. The participants were asked to demonstrate the tooth-brushing practice as usually performed at home. A trained examiner observed and documented the procedure. Observed tooth brushing and that reported by mothers were compared for overall agreement using Cohen's Kappa coefficient and the McNemar test.

Cohen's Kappa values comparing mothers' reports and tooth brushing observed by the examiner ranged from poor-to-good (0.00-0.75). There were statistically significant differences between observed tooth brushing habits and those reported by mothers (p < 0.001). When observed by the examiner, the frequencies of dentifrice dispersed on all bristles (35.9%), children who brushed their teeth alone (33.8%) and those who did not rinse their mouths during brushing (42.0%) were higher than those reported by the mothers (12.1%, 18.9% and 6.5%, respectively; p < 0.001).

In general, there was low agreement between observed tooth brushing and mothers' reports. Moreover, the different methods of estimation resulted in differences in the frequencies of tooth brushing habits, indicative of reporting bias. Data regarding children's tooth-brushing habits as reported by mothers should be considered with caution in epidemiological surveys on fluoridated dentifrice use and the risk of dental fluorosis.