Morgenstern H, Winn D M
Stat Med. 1983 Jul-Sep;2(3):387-96. doi: 10.1002/sim.4780020311.
This paper presents a new method for determining the optimal sampling ratio and sample size in different types of study designs involving binary exposure and disease variables. The sampling ratio is optimized by maximizing cost efficiency, which is the ratio of the precision in effect estimation to the total sampling cost. One may easily compute the optimal sampling ratio with a hand calculator, and it is independent of the sample sizes of the compared groups. Optimal sample sizes obtain from use of the optimal sampling ratio in the appropriate asymptotic power function for comparing two proportions or rates with unequal sample sizes. We illustrate the method with a case-control design, compare it with other methods for optimizing the sampling strategy, and discuss it in a practical context.
本文提出了一种新方法，用于确定涉及二元暴露和疾病变量的不同类型研究设计中的最佳抽样比例和样本量。通过最大化成本效率来优化抽样比例，成本效率是效应估计精度与总抽样成本的比率。人们可以使用手持计算器轻松计算出最佳抽样比例，并且它与比较组的样本量无关。通过在适当的渐近功效函数中使用最佳抽样比例来获得比较两个不等样本量的比例或率时的最佳样本量。我们用病例对照设计来说明该方法，将其与其他优化抽样策略的方法进行比较，并在实际背景下进行讨论。
