摘要
目的 在满足需探查的最小干预效益及检验把握度的条件下 ,给出群随机试验重复横断面抽样的费用效益设计方法。方法 利用条件极值———拉格朗日乘法导出可使研究费用最少时的每组群数J ,每群个体数K。结果 获得了以保证检验把握度为条件 ,以费用最小为目标的最优化设计方案 ,同时给出了在固定J或K时的研究费用计算方法。结论 这一结果为涉及大量观察对象的社区干预研究提供了费用效益设计的定量分析方法。
Objective To establish approaches of cost-efficient design of repeated cross-sectional sampling of cluster randomization trails for specified values of the statistical power and treatment effect.Methods Application of the Lagrange multiplies to solving the number of cluster per group,J,and of individual per cluster,K,for the minimizing overall cost.Results The formula of the optimum sample size to minimize the cost of taking the samples for specified values of the statistical power were deduced.The algorithm of J for specified K and the algorithm of K for specified J were also deduced.Conclusion A new method of cost-efficient design of cluster randomization trails is established for community intervention,which involves many subjects and large expenditures of time and money.
出处
《中国卫生统计》
CSCD
北大核心
2000年第6期328-330,共3页
Chinese Journal of Health Statistics
关键词
群随机试验
重复横断面抽样
费用效益设计
Cluster randomization trails
Cross-sectional sampling
Cost-efficient design
