Evaluation And Characterization Of Standard Materials And Study On Results Harmonization For Coagulation Factor And D-dimer Testing

Objective:With the fact that coagulation factor Ⅷ (FⅧ), coagulation factor Ⅸ (FIX) and D-dimer testing are commonly used in clinical hematology laboratories, reference materials are need to promote quality improvement, for the reason that some labs’results are unsatisfactory. So National Center of Clinical Laboratories have prepared 4 lots reference materials for the activity testing of FⅧ and FⅨ and 5 lots reference materials for the quantitative testing of D-dimer. In this study, we evaluate and characterize these reference materials and make the related research on results harmonization to promote FⅧ, FIX and D-dimer detection results consistency and quality improvement.Method:1. Evaluate and characterize the reference materials for FⅧ and FⅨ ctivity testing:according to ISO Guide 35 "Reference materials-general and statistical principles for certification", CNAS-GL03 "Guidance on evaluating the homogeneity and stability of samples used for proficiency testing" and CLSI EP-30 "Characterization and qualification of commutable reference materials for laboratory medicine", the homogeneity, stability and comparability of these reference materials with different factor activity levels were evaluated. And according to ISO Guide 35 "Reference materials-general and statistical principles for certification" and JJG 1006-94 "Technical norm of primary reference material", the reference materials were characterized by eight laboratories using 3 kinds of detection systems (Stago, Sysmex and EL), which were calibrated by the coagulation standard (LOT4) provided by the National Institute for Biological Standards and Control (NIBSC) in UK.2. Detection results consistency and related issues research for FⅧ and FⅨ activity testing:① With reference to foreign literatures, guidelines and the opinions of manufacturers and experts, we made the performance verification of FⅧ and FⅨ activity detection, such as intra-batch precision, between-run precision, trueness and linear range. ② Evaluated the application effect of reference materials as internal quality control materials, though calculated and compared the mean values, standard deviation and the coefficient of variation (CV) with the commercial quality controls. ③According to EP9-A2, compared differences of the FⅧ and FⅨ activity between the Stago and Sysmex; calibrated the detection systems by LOT4, the reference material with the highest value or 4 lots reference materials, elect the best model which could decrease the differences effectively.3. Evaluate and characterize the reference materials for the quantitative testing of D-dimer:according to ISO Guide 35 "Reference materials-general and statistical principles for certification", CNAS-GL03 "Guidance on evaluating the homogeneity and stability of samples used for proficiency testing" and CLSI EP-30 "Characterization and qualification of commutable reference materials for laboratory medicine", the homogeneity, stability and comparability of these reference materials with different values were evaluated. And according to ISO Guide 35 "Reference materials-general and statistical principles for certification" and JJG 1006-94 "Technical norm of primary reference material", the reference materials were characterized by eight laboratories using 3 kinds detection systems whose results units are FEU (Stago, Sysmex and Vidas).4. Detection results consistency and related issues research for the quantitative testing of D-dimer: ①With reference to foreign literatures, guidelines and the opinions of manufacturers and experts, we made the performance verification for the quantitative testing of D-dimer, such as intra-batch precision, between-run precision and linear range. ② Evaluated the application effect of reference materials as internal quality control materials, though calculated and compared the mean values, standard deviation and the coefficient of variation (CV) with the commercial quality controls. ③According to EP9-A2, compared the differences of the quantitative testing between the Stago and Sysmex; calibrated the detection systems by the reference material with the highest value or 5 lots reference materials, elect the best model which could decrease the differences effectively.Results:1. Evaluate and characterize the reference materials for FⅧ and FⅨ activity testing:The results of one-way ANOVA showed that all reference materials had good homogeneity (P>0.05), and the between-bottle homogeneity uncertainties (ubb) of FⅧ and FIX activity were 0.5%～2.9% and 0.1%～3.9%, respectively. All reference materials stored in -80 ℃ remain stable in 9 months by trend analysis, and the long-term stability uncertainties(ults) of FⅧ and FIX activity were 0.5%～5.1% and 0.8%-4.4%, respectively. All reference materials are commutable for the three detection systems. Approved by Grubbs test, Cochran test and normality test, the mean values of reference materials for FVI were assigned by 8,8,8 and 6 labs respectively, and 8,7,8 and 7 labs for FIX respectively. The characterization uncertainties (uchar) of FⅧ and FⅨ activity testing were 0.7%～2.4% and 0.4%-2.4%, respectively. Calculating the combined uncertainties and extended uncertainties (coverage factor k=2), the assigned values of each lot of reference materials for FⅧ activity were (85±13)%, (36±4)%, (20±3)% and (6±2)%, which were (102±13)%, (48±7)%, (29±4)% and (10±2)% for FIX activity, respectively.2. Detection results consistency and related issues research for FⅧ and FⅨ activity testing: ①The intra-batch precision and between-run precision of FⅧ and FⅨ activity detection were all met the manufacture’s requirements. Trueness verification results showed that the bias for FⅧ and FⅨ activity testing were -5.0% and 12.7%. All the correlation coefficient R2 of linear regression equations were over 0.99,10.3%～149.6% intervals of F Ⅷ activity was approved by linear verification, for FⅨ activity that ranges were 16.7%～200.2%.②Internal quality control application results showed that the ranges of CV for FⅧ and FⅨ were 3.1%～4.5% and 4.3%～5.6%, respectively. They were similar to the commercial quality controls (6.4%-6.9% and 5.0%～7.8%)。③40 clinical specimens which activity results range from 1.3% to 123.1% were tested in the detection systems of Stago STA-R Evolution and Sysmex CS 5100, the correlation coefficient R2 of two systems was 0.95. If we set the acceptable value as ±30% of relative deviation (measured value> 12%) or ±3% of absolute deviation (measured valued 12%),85% of the results were acceptable. The range of relative deviation was -18.0%～46.8%, and for absolute deviations were -3.8%～0.37 clinical specimens which activity results range from 3.4% to 147.6% were also tested in the detection systems of Stago STA-R Evolution and Sysmex CS5100, the correlation coefficient R2 was 0.95,84% of the results were acceptable. The range of relative deviations was-17.6%～40.4%, and for absolute deviations the range was 1.5%-2.6%. When we calibrated the detection systems by LOT4, the reference material with the highest value and 4 lots reference materials, for FⅧ the correlation coefficient R2 were all 0.94. The numbers of acceptable results were 95%,88% and 90%, respectively. The ranges of relative deviation were -29.9%-30.4%,-42.1%～21.0% and -33.3%～21.9%, and for absolute deviations were -3.8%～0,-5.4%-0.8% and -5.1%～1.4%, respectively. for FⅨ the correlation coefficient R2 were all 0.96. The numbers of acceptable results were 92%,95% and 95%, respectively. The ranges of relative deviations were -21.6%-36.0%,-33.1%-26.7% and -31.3%～28.5%, and the ranges for absolute deviations were 0.1%～3.1%,-0.6%～2.3% and -0.7%～2.4%, respectively.3. Evaluate and characterize the reference materials for the quantitative testing of D-dimer:The results of one-way ANOVA showed that all reference materials had good homogeneity (P>0.05), and the between-bottle homogeneity uncertainties (ubb) of D-dimer test were 0.001%～0.033%. All reference materials stored in -80℃ remain stable in 11 months by trend analysis, and the long-term stability uncertainties (ults) were 0.011%～0.104%. All reference materials are commutable for the three detection systems. Approved by Grubbs test, Cochran test and normality test, the mean values of reference materials were assigned by 6,6,7,8 and 7 labs respectively. The characterization uncertainties (uchar) were 0.011%～0.096%. Calculating combined uncertainties and extended uncertainties (coverage factor k=2), the assigned values of each lot of reference materials were (0.27±0.04)mg/L, (0.35±0.04)mg/L, (0.58± 0.06)mg/L, (2.01±0.19)mg/L and (2.79±0.30)mg/L.4. Detection results consistency and related issues research for the quantitative testing of D-dimen ①The intra-batch precision and between-run precision were all met the demanded of manufacture and WS/T 477-2015. The correlation coefficient R2 of linear regression equations were over 0.99,0.26mg/L～3.96mg/L intervals was approved by linear verification. ②Internal quality control application results showed that the range of CV was 2.5%-13.7%, which were similar to the commercial quality controls (1.7%-14.8%)。③38 clinical specimens which range from 0.18mg/L to 3.69mg/L were tested in the detection systems of Stago STA-R Evolution and Sysmex CA1500, the correlation coefficient R2 of two systems was 0.95. If we set the acceptable value as ±28% of relative deviation (measured value>0.50mg/L) or ±0.25mg/L of absolute deviation (measured value≤ 0.50mg/L),95% of the results were acceptable. The range of relative deviations was-34.1%～29.6%, and for absolute deviations were -0.22mg/L～0.16mg/L. When we calibrated the detection systems by the reference material with the highest value and 5 lots reference materials, the correlation coefficient R2 were 0.82 and 0.90, respectively. The acceptable proportions were 89% and 89%, respectively. The range of relative deviations were -57.4%～23.5% and -46.6%～36.8%, and the absolute deviations were-0.05mg/L～-0.29mg/L and -0.10mg/L～0.12mg/L, respectively.Conclusion:1. The reference materials which were commutable for the three detection systems showed good homogeneity and stability for activity testing of FVI and FIX. The results of the characterization are accurate and reliable.2. ①Performance verification results of STA-R Evolution for the activity testing of F Ⅷ and FIX meet the requirements of the manufacture and relevant guidelines. ②The reference materials for activity testing of FⅧ and FIX can be used as internal quality controls. ③)40 clinical specimens which activity results range from 1.3% to 123.1% and 37 clinical specimens which activity results range from 3.4% to 147.6% were tested in the detection systems of Stago STA-R Evolution and Sysmex CS5100, the correlation for the activity testing of FⅧ and FIX are good (R2≥095), and for comparison are acceptable (at least 80% of the results satisfied the standard). When we calibrate the detection systems by L0T4, the reference material with the highest value and 4 lots reference materials, for FⅧ and FIX, the correlations of all calibrate models are not changed. But all models could decrease the deviation of comparison.3. The reference materials which were commutable for the three measurement systems showed good homogeneity and stability for the D-dimer quantitative testing. The results of the characterization are accurate and reliable.4. ①Performance verification results of STA-R Evolution for quantitative testing of D-dimer meet the requirements of guidelines. ②The reference materials for quantitative testing of D-dimer can be used as internal quality controls. ③Based on this test data shows,38 clinical specimens which range from 0.18mg/L to 3.69mg/L were tested with the detection systems of Stago STA-R Evolution and Sysmex CA1500, the correlation is good (R2≥095), and for comparison is acceptable (at least 80% of the results satisfied the standard). When we calibrate the detection systems by the reference material with the highest value and 5 lots reference materials, the correlation and comparison don’t improved significantly.