自建检测系统测定血清总蛋白测量不确定度的评估

目的评估我科白建检测系统测定血清总蛋白的测量不确定度。方法根据血清总蛋白的医学决定水平（MDL）配制两个浓度水平的试验样本．分别定义为低值MDL样本、高值MDL样本。先评估血清总蛋白在低值MDL及高值MDL的批内重复性、批间重复性、方法偏倚等因子，再根据上述因子确定血清总蛋白在上述两个MDL水平处的扩展不确定度（EU）。结果（1）自建检测系统测定血清总蛋白在低值MDL的批内变异系数（CVw）、批间变异系数（CVB）、方法偏倚变异系数（CVBiaw）分别为0．86％、1．37％、2．31％，EU为5．64％。（2）自建检测系统测定血清总蛋白在高值MDL的CVW、CVB、CVBias。分别为0．74％、1．41％、3．29％，EU为7．31％。结论本研究评估测量不确定度的方法简单、易行，试验所得的EU可以准确地反映临床检验结果的分散性。

Evaluation of measurement uncertainty of serum total protein detected by serf-developed detecting system

Objective To evaluate the measurement uncertainty of serum total protein detected by the self-developed detecting system in our department. Methods Two groups of samples were prepared according to medical decision levels（MDL） of serum total protein, which were defined as low-value MDL sample and high-value MDL sample respectively. The within-run repeatability, between-run repeatability and method biases of serum total protein at the 2 levels of MDLs were evaluated firstly, and then the corresponding expanded uncertainties （EU） were confirmed. Results （1）The within-run coefficient of variation（CVw）, between-run coefficient of variation（CVB） and methods bias coefficient of variation（CVBias） of serum total protein at the low-value MDL were 0.86%, 1.37% and 2.31% respectively, and the EU was 5.64%. （2）The CVw , CVB and CVBias of serum total protein at the high-value MDL were 0.74%, 1.41% and 3.29% respectively, and the EU was 7.31%. Conclusion This method for evaluating measurement uncertainty in this study is of convenience and feasibility, and the EU can reflect the dispersion of clinical test results accurately.