Background: Most studies of tooth brushing behaviors rely on self-report or demonstrations of behaviors conducted in clinical settings. This study aimed to determine the feasibility of objective assessment of tooth brushing behaviors in the homes of high-risk children under three years old. We compared parent self-report to observations to determine the accuracy of self-report in this population.

Methods: Forty-five families were recruited from dental and medical clinics and a community social service agency. Research staff asked questions about oral health behaviors and observed tooth brushing in the homes. Brushing was also video-recorded. Video-recordings were coded for brushing behaviors by staff that did not collect the primary data; these abstracted data were compared to those directly observed in homes.

Results: Most families were Hispanic (76%) or Black (16%) race/ethnicity. The majority of parents had a high school education (42%) or less (24%). The mean age of children was 21 months. About half of parents reported brushing their child's teeth twice a day (58%). All parents tried to have their children brush, but three children refused. For brushing duration, 70% of parents reported differently than was observed. The average duration of brushing was 62.4 s. Parent report of fluoride in toothpaste frequently did not match observations; 39% said they used toothpaste with fluoride while 71% actually did. Sixty-eight percent of parents reported using a smear of toothpaste, while 61% actually did. Brushing occurred in a variety of locations and routines varied. Abstracted data from videos were high in agreement for some behaviors (rinse with water, floss used, brushing location, and parent involvement: Kappa 0.74-1.0). Behaviors related to type of brushing equipment (brushes and toothpaste), equipment storage, and bathroom organization and clutter had poor to no agreement.

Conclusions: Observation and video-recording of brushing routines and equipment are feasible and acceptable to families. Observed behaviors are more accurate than self-report for most components of brushing and serve to highlight some of the knowledge issues facing parents, such as the role of fluoride.