论文部分内容阅读
目的对某国产艾滋病病毒1型(HIV-1)病毒载量检测试剂盒的应用性能进行评估。方法收集200份血液样品,分别用该国产待评估试剂(简称待测试剂)与进口参比试剂(简称参比试剂)进行盲试检测。然后采用阴阳性符合率和卡方检验等统计方法,来判定两种试剂定性检测结果的一致性;用相关性、回归分析和Bland-Altman模型等统计方法,来判定两种试剂定量检测结果的一致性。结果两种试剂总体符合率达100%,Kappa值为1.000(P<0.001)。两种试剂盒检测结果的回归分析表现出较强的相关性(R~2=0.933)。运用Bland-Altman模型比对,显示两种试剂检测结果有97.06%(165/170)的样本在95%一致性界限内,说明两种试剂检测结果的一致性较好。文章还将经治疗样本和未经治疗样本的检测结果进行定量检测结果分析,结果显示,治疗样本两种试剂检测结果的相关性为0.968,回归分析表现出强相关性(R~2=0.937);Bland-Altman模型显示,有96.64%的样本在95%一致性界限内,说明两种试剂在治疗样本中的一致性较好。在未治疗样本中,两种试剂检测结果的相关性为0.952,回归分析表现出强相关性(R~2=0.906);Bland-Altman模型显示,有98.04%的样本在95%一致性界限内,说明两种试剂在未治疗样本中的一致性较好。从经过治疗样本和未经治疗样本的定量检测结果可以看出,无论是否接受治疗,国产试剂与进口试剂在检测在检样本的测试效果方面具有强相关性。结论该国产试剂和进口试剂检测结果具有较高的定性符合率、较强的相关性以及较强的定量一致性。
Objective To evaluate the application performance of a domestic HIV-1 viral load detection kit. Methods A total of 200 blood samples were collected and tested blindly using the domestic reagent to be evaluated (referred to as test reagent) and imported reference reagent (referred to as reference reagent) respectively. Then using the positive and negative coincidence rate and chi-square test and other statistical methods to determine the qualitative consistency of the two reagents test results; correlation, regression analysis and Bland-Altman model and other statistical methods to determine the two reagents quantitative test results consistency. Results The overall coincidence rate of the two reagents was 100% with Kappa value of 1.000 (P <0.001). Regression analysis of the two kits showed a strong correlation (R ~ 2 = 0.933). Using the Bland-Altman model comparison, 97.06% (165/170) of the two reagents showed a 95% agreement within the limits of agreement, indicating a good agreement between the two reagents. The results of the quantitative analysis of the treated and untreated samples were also analyzed. The results showed that the correlation between the two tested samples was 0.968 and the regression analysis showed a strong correlation (R ~ 2 = 0.937) The Bland-Altman model shows that 96.64% of the samples are within the 95% agreement limits, indicating that the two reagents are well-matched in the treated samples. In untreated samples, the correlation between the two reagents was 0.952, and the regression analysis showed a strong correlation (R ~ 2 = 0.906). The Bland-Altman model showed that 98.04% of the samples were within the 95% agreement limit , Indicating that the consistency of the two reagents in the untreated sample is better. It can be seen from the quantitative test results of the treated samples and the untreated samples that the domestic reagents and the imported reagents have a strong correlation in testing the test results of the tested samples whether or not receiving the treatment. Conclusion The test results of domestic and imported reagents have high qualitative coincidence rate, strong correlation and strong quantitative consistency.