Petersen H, Larsen M L, Hörder M, Blaabjerg O
Department of Clinical Chemistry, Odense University Hospital, Denmark.
Scand J Clin Lab Invest Suppl. 1990;198:66-72.
In screening programmes one should distinguish between the traditional bimodal distributions of results and a unimodal distribution as the basis for interpretation. A model for evaluation of the effects of biological within-subject and preanalytical variation as well as analytical variation is described. It is concluded that bias from blood sampling and analytical performance influences the outcome of screening programs significantly. At least three blood specimens are needed for estimating the homeostatic set point of cholesterol in individuals.
在筛查项目中，应区分传统的双峰结果分布和单峰分布，以此作为解释的基础。本文描述了一个评估生物体内变异、分析前变异以及分析变异影响的模型。得出的结论是，采血偏差和分析性能会显著影响筛查项目的结果。估计个体胆固醇稳态设定点至少需要三份血液标本。
