Setting: Parents of children 12 years old and under whose child had made at least one visit to a study clinic in the years 2000–2003.
Methods: As part of a randomized controlled trial to improve patient provider communication and preventive practices, parents' responses to telephone interview were compared with observations of safety practices during a home visit. Home visits were completed within nine weeks of the telephone interview. Parents were not told that the visit was part of a validation study and home visit observers were unaware of the interview responses. The authors calculated sensitivities, specificities, positive and negative predictive values, and their corresponding confidence intervals.
Results: Sensitivity (0.78 to 0.98) and positive predictive values (0.75 to 1.00) were high for all items. Specificities and negative predictive values were more variable and the highest estimates (specificity 0.95 to 1.00, negative predictive value 0.95 to 0.97) were for car seat types.
Conclusions: The results suggest that parent self report practice of certain injury prevention behaviors (owning a car seat, hot water temperatures) is reliable, whereas self reports on other practices (working smoke detectors, properly fitting bike helmets) may be overstated.