不同检测系统生化指标测定结果的偏倚评估与可比性研究

Comparative study and bias evaluation on the results of serum biochemical marker with different biochemical detecting systems.

目的探讨不同检测系统间生化指标测定结果的可比性,为不同实验室检验结果的互认提供数据。方法以雅培Aeroset生化分析仪,雅培原装试剂和校准品、Bio-Rad质控品组成的检测系统为比较方法;以雅培C8000生化分析仪、上海申能德赛试剂、日本第一化学试剂、上海科华试剂以及德国Trucal校准品为实验方法,分别用病人新鲜混合血清及室内质控血清对丙氨酸转氨酶、碱性磷酸酶、总胆红素、血糖、总蛋白、甘油三酯、钙、钾进行检测。计算实验方法和比较方法之间的系统误差,判断不同检测系统测定结果的可比性。结果雅培Aeroset和C8000两个检测系统的精密度较高(CV均<4%),除钙外,其它项目检测结果的可靠性强(可靠性系数>0.99),系统误差临床可以接受。结论当用不同检测系统检测同一检验项目时,应进行方法比对和偏倚评估,同时应为检验结果加上不确定度赋值,以保证检验结果的可比性,为临床提供一致可靠的检验数据。

Objective In order to make the outcomes which detected by different laboratories to be mutually recognized, we made an approach to the comparability of biochemical marker results which detected by different detecting system. Methods Several biochemical markers（which include alanine aminotransferase, alkaline phosphatase, total bilirubin, Glucose, total protein, glyceryl trioleate, calcium, potassium） in the fresh mixed sera and control sera were detected by comparison method （which includes Abbott Aeroset automatic biochemistry analyzer, reagent and calibrator of Abbott, controller of Bio-Rad） and laboratory method （which includes Abbott C8000 automatic biochemistry analyzer, reagent of Shanghai Diasys and Shanghai Kehua, calibrator of Trucal）. The coefficient and the equation were utilized to estimate relative bias between laboratory method and comparison method, then the comparability of outcomes of different detecting system was able to judged. Results The precisions of both detecting system were perfect（CV〈4 %）. Outcomes of major biochemical markers seemed to be visible reliability（coefficient of reliability〉0. 99） except calcium. The system error could be accepted. Conclusion In order to insure the comparability, it is necessary to do method comparison and bias evaluation when the same test is determined by different detecting system. Meanwhile the outcomes should add the uncertainty to catch the goal of providing reliable data to clinical diagnosis and treatment.