双界点OR值最大法logistic回归在中小学生登革热知识提高影响因素研究中的应用

摘要

Objective To explore the application of two cut-off points and maximum OR values method in logistic model for analyzing impact factors on the improvement of dengue related knowledge of primary and secondary school students.Methods Maximum OR values method in logistic model was applied to the multi-stage stratified cluster random sampling data of the primary and secondary school students in Guangzhou to analyze impact factors on the learning effect of dengue related knowledge.In addition,the model fitting effect of Maximum OR values method in logistic model was compared with continuous variables method,median split method and one cut-off point minimum P values method.Results We enrolled a valid sample of 1311 students in evaluation of learning effect,containing primary 3,primary 5,junior 2 and senior 2 grades students.Maximum OR values method in logistic model retained variables of baseline knowledge scores,intervention method and dengue cases in data of senior 2,which presented the best model fitting effect and performed better in screening out the impact factors related to outcome.Conclusion Two cut-off points and maximum OR values method in logistic model could be used to analyze impact factors on learning effect of dengue related knowledge of primary and secondary school students.There were multiple factors which influenced the learning effect of dengue related knowledge of primary and secondary school students,which are supposed to be developed aim measures of health education.%目的 探讨双界点OR值最大法logistic回归在中小学生登革热知识提高幅度的影响因素研究中的应用.方法 基于多阶段分层整群随机抽样的广州市中小学生登革热相关知识传播效果的干预研究的调查数据,建立中小学生登革热知识提高幅度影响因素的双界点OR值最大法logistic回归模型,与采用连续型变量法、中位数法、单界点P值最小法的logistic回归模型比较模型拟合效果,探讨双界点OR值最大法logistic回归的优势.结果 纳入干预效果评估的共1311人,三年级299名,五年级331名,初二321名,高二360名.以高二数据为例,双界点OR值最大法logistic回归纳入的自变量为知识基线得分、干预方式、登革热病例数;其模型的拟合效果最好,且对于自变量分界点的划分更合理,能够更好地筛选出与结局有关联的影响因素.结论 双界点OR值最大法logistic回归适用于中小学生登革热知识提高幅度影响因素的研究.中小学生的登革热知识的干预效果受多个因素的影响,应开展有针对性的健康教育.