RESUMEN
Background: Early childhood fall is a pressing global public health problem and one of the leading causes of child injury. China has a high proportion of children and a high burden of illness from falls. Therefore, educational interventions to prevent childhood fall would be beneficial. Methods: We used the outcome of knowledge, attitude and practice questionnaire, which was conducted by Pudong New District of Shanghai Municipal Government, to summarize demographic and baseline characteristics grouped by intervention or not, and analyzed descriptive statistics of continuous and categorical variables. A logistic stepwise function model was established to study the influence of different covariables on the degree of injury, and AIC/BIC/AICC was used to select the optimal model. Finally, we carried out single-factor analysis and established a multifactor model by the stepwise function method. Results: Attitude and actual behavior scores had significant differences. The intervention and control groups had 20.79 ± 3.20 and 20.39 ± 2.89 attitude scores, respectively. Compared to the control group (5.97 ± 1.32), the intervention group had higher actual behavior scores (5.75 ± 1.50). In the univariate analysis results, fathers' education level, mothers' education level, actual behavior and what cares for children had a significant influence on whether children got injured. In multivariate analysis, attitude had a positive influence on whether injured [odds ratio: 1.13 (1.05-1.21), P < 0.001]. Conclusion: Educational intervention for children and their guardians can effectively reduce the risk of childhood falls, and changes in behavior and attitude are the result of educational influence. Education of childhood fall prevention can be used as a public health intervention to improve children's health.