Harris E K, Yasaka T
Clin Chem. 1983 Jan;29(1):25-30.
We describe a statistical method for calculating a "reference change," defined as that difference between two consecutive test results in an individual that is statistically significant in a given proportion of all similar persons. By allowing for variation in within-person variances, this procedure computes a reference change that is more specific (i.e., less prone to false positives) than that obtained directly from the distribution of observed differences between measurements. Moreover, the method may easily be extended to a test for trend in three successive measurements. The method has been applied to semi-annual measurements of serum calcium and alkaline phosphatase in 698 men and women enrolled in a large health-maintenance program. We believe that these ideas may also be usefully applied to successive laboratory tests in carefully defined patient populations--but this introduces special problems, which are discussed briefly.
我们描述了一种统计方法，用于计算“参考变化”，其定义为个体两次连续检测结果之间的差异，该差异在所有相似个体的给定比例中具有统计学意义。通过考虑个体内方差的变化，此程序计算出的参考变化比直接从测量值之间观察到的差异分布中获得的参考变化更具特异性（即，更不易出现假阳性）。此外，该方法可轻松扩展为对连续三次测量的趋势进行检验。该方法已应用于参与大型健康维护计划的698名男性和女性的血清钙和碱性磷酸酶的半年期测量。我们认为，这些理念也可有效地应用于精心定义的患者群体的连续实验室检测——但这会带来一些特殊问题，将简要讨论这些问题。
