Clinical Laboratory Analysis For Quality Control

Objective To study the analytical performanc of test methods quantitatively and identify the quality state of our laboratory, which can provide the evidence for acquiring the effective quality control strategy and accumulate data for continuous improvement of laboratory performance and farther laboratory accreditation.Methods According to NCCLS EP5-A document, we analyzed the with-run, between-run, between-day and total imprecision. According to the external quality assessment results of our laboratory, we analyzed the bias between our measurement value and the target value, then identified the accuracy of the test methods. According to the measured imprecison and accuracy, we decided the method performance characteristics of all the items making use of normalized method decision chart. According to polynomial linear evaluation method and NCCLS EP6 document, we found the best-fit polynomial of the data making use of statistical software, then estimated the imprecision and linear degree of the data combined with the clinical goals. Following the NCCLS EP9-A2 document, we compared the performance of the three types of biochemical autoanalyzers which are used routinely in ourlaboratory.Results The imprecision of every biochemical item is not identical, the variance degree of ALT, Crea, UA, ALP, GGT, HBDH, P is bigger than a quarter of TEa, but smaller than a third of TEa, the imprecision of other items is smaller than a quarter of TEa. The accuracy of every biochemical item is not identical, the bias of TB, Alb, TG, GGT, HBDH, Ca is bigger than a quarter of TEa, but smaller than one half of TEa, the bias of other items is smaller than a quarter of TEa. Known from the normalized method decision chart, the performance of 65% of the items is excellent, only the performance of GGT is poor. There are four kinds of linear results: the linear result of AST, Cl, CO2, GGT, LDH , TP , Na , Urea , Choi is linearity 1, that of TB, ALT, Alb, ALP, AMY, Ca, CK , Crea, Glu, Lip, P, K, LL<\^ TG is linearity 2, that of Mg, LDH is nonlinearity, and the linear result of Pco2 is too imprecision. The results of the three types of Olympus biochemical autoanalyzers are consistent except for Glu between AU5400 and AU2700 and Glu, CO2, Cl, Alb, GGT, UA between AU5400 and AU400. Conclusions The imprecision of all the items is smaller than 1/3 TEa, the methods can be well-managed in routine service if we plan the statistical QC procedure carefully. And the imprecision of TB, DB, AST, TP, Alb, Glu, Urea, Crea, TG, Choi, CK, LDH, Na, K, Cl, Ca, Mg is smaller than 1/4 TEa, which leads to more reliable detection of medically important errors with simpler QC program. The accuracy of every biochemical item is not identical, the bias of TB, Alb, TG, GGT, HBDH, Ca is between a quarter and one half of TEa, the bias of other items is smaller than a quarter of TEa. Only the performance of GGT is poor, the performance of 65% of the items isexcellent, and that of other items meets the clinical requirement, but it can guarantee the quality more efficiently if we improve it ulteriorly. The measurements of most items present linear expression. We ultilize multiple points calibration to guarantee the quality for the nonlinear items. In the definite concentration of most items, the results of the three types of Olympus biochemical autoanalyzers are consistent, or we should adjust the parameters.