加权非线性回归用于不同生化分析仪之间的比对

摘要

Objective To make a comparison study on results from different biochemical analyzers using weighted nonlinear regression. Methods Daily 5 blood samples with high, normal and low concentration values from patients were collected and each sample was repeatedly measured twice respectively by 3 biochemical analyzers, Bayer1650, AU600 and VITROS 950. The measured results of BUN,CR,UA and GLU from different biochemical analyzers were weighted based on the standard deviation of repetitive measurements, and then were fitting through 11 regression models( including linear model) from SPSS Curve Estimation Program. Results The coefficient of determination ( R2 ) of all best regression models found out was more than 0. 99 ( the correlation coefficient was more than 0. 995 ), and then method comparison and bias estimation between different biochemical analyzers could be made more accurate in a wider comparative range. Conclusions Weighted nonlinear regression is suitable for method comparison and bias estimation for clinic laboratory.%目的 利用加权非线性回归对不同牛化分析仪的测定结果进行比对.方法 每天收集5份高、中、低不同浓度水平的病人血液样本.每份样本分别使用Bayer1650、AU600和VITROS 950生化分析仪进行2次重复测定,根据这2次重复测定的标准差对测定结果进行加权处理,然后利用SPSS曲线拟合(Curve Estimation)程序的11种回归模型(包括线性模型)对不同仪器BUN、CR、UA和GLU项目的 测定结果进行拟合.结果 所选出的最佳模型的决定系数值(R2)均>0.99(相关系数>0.995),这使生化分析仪之间的方法比对和偏差估计更为准确可靠,且可比对范围更广.结论 加权非线性回归适用于临床检验的方法比对和偏差估计.