生化分析系统淀粉酶测定结果的偏倚评估

目的 探讨不同生化分析系统检测血清淀粉酶(Amy)结果偏倚.方法 先检测三台仪器重复性及线性分析后,再以强生V350干式生化仪测定结果作为比较方法的测定值(X),以我院7170A全自动生化分析仪测定结果作为实验方法的测定值(Y),对46例测定结果进行相关分析及偏倚的评估.结果 各检测系统测定Amy的精密度变异系数均小于10%,3个不同牛化分析系统新鲜血清淀粉酶测定结果相关系数均大于0.975,可靠性系数为0.9856.两台7170A生化分析仪测定46例血清Amy在3个医学决定水平(50,120,200 U/L)处相对偏差分别为39.08%、8.95%、0.34%和40.58%、10.39%、1.76%.结论 不同生化分析系统血清Amy检测结果存在偏倚.其相对偏差在医学决定水平50 U/L时,临床不可接受;而在120和200 U/L时,临床可接受.

Bias estimation on amylase reports by chemical analysis systems

Objective To investigate the bias of amylase results measured by three analyzers. Methods A total of 46 blood samples was examined for amylase with three biochemistry analyzers.Johnson Vit ros 350 and two Hitachi 7170A analyzers. The coefficient and the equation were used to evaluate the bias among three systems. Results The correlation coefficiency of each detection system was above 0. 975 and their reliability coefficiency was 0. 9856. The CV value of all detection systems did not exceed 10% by precise assessment Johnson Vitros 350 was regarded as a standard detection system. The deviation of three medical decision levels of 50,120 and 200 U/L were 39. 08 %, 8. 95 % and 0.34% for one Hitachi 7170A analyzer and 40.58%, 10.39% and 1.76% for another one. Conclusion The bias of amylase results measured by two Hitachi 7170A analyzers exists,which could be acceptable at the medical decision levels of 120 and 200 U/L, but not at 50 U/L.