van den Besselaar AMHP, van Rijn CJJ, Abdoel CF, Cobbaert CM. Bias and uncertainty of the International Normalized Ratio determined with a whole blood point-of-care prothrombin time test device by comparison to a new International Standard for thromboplastin.
Thromb Res 2021;
202:1-7. [PMID:
33706049 DOI:
10.1016/j.thromres.2021.02.018]
[Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/26/2020] [Revised: 02/04/2021] [Accepted: 02/10/2021] [Indexed: 10/22/2022]
Abstract
BACKGROUND
Whole blood point-of-care PT/INR test devices, e.g. CoaguChek XS, are calibrated by their manufacturers. In the Netherlands, each new lot of test strips for CoaguChek XS is validated by a group of anticoagulant clinics collaborating with a Coagulation Reference Laboratory. In 2017, a new International Standard for recombinant human thromboplastin (coded rTF/16) has been established by the World Health Organization.
AIM
To assess uncertainty of the validation procedure and the magnitude of the INR bias of a series of consecutive lots of test strips imported in the Netherlands.
METHODS
CoaguChek XS test strip INR results were compared to INRs determined with the new International Standard rTF/16. Comparisons were made with variable numbers of blood samples obtained from patients treated with vitamin K-antagonists. Relationships between CoaguChek XS and rTF/16 results were determined with orthogonal regression analysis. The relationships were used to assess bias and uncertainty of bias.
RESULTS
Average bias between CoaguChek XS test results and rTF/16 depends on the INR level. Overall, there was a trend of increasing bias and increasing uncertainty with increasing INR values. Along the sequence of 47 consecutive lots, a temporary fluctuation of bias was observed. At an INR level of 3.0 the average bias was less than 10% in all cases, but at an INR of 4.0 there were 5 lots with average bias between 10 and 15%.
CONCLUSION
Validation of test strips is useful to assess bias but depends on availability of fresh patients' samples and traceability to an accepted Reference Measurement System.
Collapse