Dodd R Y, Bastiaans M J, Nath N, Ellis F R, Barker L F
Vox Sang. 1980 Dec;39(6):309-17. doi: 10.1111/j.1423-0410.1980.tb02967.x.
The American Red Cross collects blood from a number of defined subsets of the donor population and teh proportion of blood collected from each subset varies widely from center to center. A large part of the variation in prevalence of HBsAg may be related to variations in the proportion of blood collected from plants and factories, military units and schools or colleges. We have derived a regression equation, significant at the p less than 0.001 level, which links HBsAg prevalence with these collection parameters. Using this equation, we were able to predict the prevelance of HBsAg among first-time donors in 6 of the 9 geographic divisions of the United States with an accuracy exceeding 10%. The predictions for the remaining division were within 35% of the actual value. Correlation studies were supported by measurements of true donor prevalence in three blood centers.
美国红十字会从特定的献血者群体子集中采集血液，每个子集采集的血液比例在不同中心之间差异很大。乙肝表面抗原（HBsAg）流行率的很大一部分差异可能与从工厂、军队单位以及学校采集的血液比例变化有关。我们推导出了一个回归方程，在p小于0.001水平上具有显著性，该方程将HBsAg流行率与这些采集参数联系起来。使用这个方程，我们能够预测美国9个地理区域中6个区域首次献血者的HBsAg流行率，准确率超过10%。其余区域的预测值与实际值相差在35%以内。相关性研究得到了三个血液中心对真实献血者流行率测量的支持。
