ABSTRACT
Objective To explore the verification methods for the performance of quantitative detection kit of anti-dsDNA antibodiy with enzyme-linked immunosorbent assay (ELISA).Methods The precision was verified according to the EP15-A2 document approved by the American Clinical and Laboratory Standards Institute(CLSI).The accuracy was verified by detecting the samples of previous external quality evaluation(EQA),compared with the comparative kits and recovery test.The lower limit of detection(LLD) was calculated by the results of blank samples.The cut-off value was verified according to the C28-A3C document approved by CLSI and CNAS-CL39:Guidance on the Application of Accreditation Criteria for the Medical Laboratory Quality and Competence in the Field of Clinical Qualitative Immunology respectively.The improved Doumas method was used to verify the range of linearity.Results The measured intra-assay and inter-assay coefficients of variation were lower than those announced by the manufacturer or the calculated values according to the EP15-A2 document.The coincidence rates for negative and positive EQA samples between detected and expected values were 98.4％ (63/64) and 100％ (20/20) respectively.The total coincidence rate was 98.8％ (83/84).The coincidence rate for negative and positive samples between the results from candidate and comparative kits were 91.2％ (52/57) and 87.0％ (40/46) respectively.The total coincidence rate was 89.3％ (92/103) and the Kappa value was 0.783 (P =0.062),which implied excellent consistency between the two kits.The mean recovery rate was 99.65％.The measured LLD was 0.5 IU/mL which was lower than 1 IU/mL as claimed by the manufacturer.The measured cut-off value according to the CNAS-CL39 document was 18.51 IU/mL,which was close to 20 IU/mL announced by the manufacturer.Based on the C28-A3C method,the cut-off value could be approved.The linear regression equation was Y =0.978 8X-3.125 4,r2 =0.996 1.There was no statistical difference between the intercept (-3.125 4) and 0 (t =-0.772,P =0.483).The range of linearity was from 1.6 to 212.5 IU/mL,which was consistent with the values declared by the manufacturer.All the verifications of the five performances above-mentioned could be passed.Conclusion The precision,accuracy,LLD,cut-off value and range of linearity of the candidate quantitative ELISA kit for anti-dsDNA antibody were consistent with the statement of the manufacturer,which indicated the performance of the kits may meet the requirements of clinic diagnosis and treatment.A series of methods used in this study provided a simple protocol for verifying the performance of quantitative ELISA kits.
ABSTRACT
Objective To investigate the application value of serum pancreatic isled autoantibodies and biochemical indicators in classification diagnosis of type 1 diabetes mellitus (T1DM ) and type 2 diabetes mellitus (T2DM ) .Methods The clinical data and laboratory detection results in 99 cases of T1DM and 577 cases of T2DM were retrospectively analyzed .The levels of pancreatic isled autoantibodies and biochemical indicators were compared between the two groups and their characteristics were analyzed .Re‐sults The positive rates of single detection and combine detection of glutamic acid decarboxylase autoantibodies (GADA) ,insulino‐ma‐associated antigen‐2 autoantibodies (IA‐2A ) ,islet cell autoantibodies (ICA ) and ZnT8 autoantibodies (ZnT8A ) in the T1DM group were higher than those in the T2DM group ,the differences were statistically significant (P0 .05) .Moreover ,the fasting and postprandial 2 h CP levels in the T1DM group showed decreasing trend as the T1DM course extending ,and the difference had statistical difference among different disease courses ;but the fasting and postprandial 2 h CP levels in the T2DM group had no obvious decreasing trend .The areas under the receiver operating characteristic(ROC) curve of fasting and postprandial 2 h CP for differential diagnosis of T1DM and T2DM in the patients with the disease course < 2 year were 0 .902(95% CI:0 .850-0 .954) and 0 .905(95% CI:0 .852-0 .958) respective‐ly .The suitable threshold value of fasting CP was 0 .283 nmol/L ,its sensitivity and specificity were 82 .6% and 89 .2% ,respective‐ly ,which of postprandial 2 h CP was 0 .421 nmol/L ,its sensitivity and specificity were 84 .8% and 89 .2% respectively . Conclusion T1DM and T2DM are different in onset age ,BMI value ,serum GADA ,IA‐2A ,ICA ,ZnT8A ,insulin ,CP ,glucose , HbA1c ,TG and HDL‐C levels ,which may assist clinic in their classification diagnosis .