Determination of D-dimer by different quantitative assays-A harmonization exercise

Background: D-dimer determined by different assays shows considerable variability due to lack of standarization. We describe a harmonization exercise performed in an attempt to diminish this variability using plasma samples with high D-dimer concentration. Materials and methods: Five laboratories participated with several different quantiative D-dimer assays: Vidas D-dimer Exclusion (Biomerieux), Auto-Dimer (Biopool), D-dimer Plus, Acute Care D-dimer (both Dade Behring) and Hemosil D-Dimer (Instrumentation Laboratory). For harmonization a set of six plasmas was prepared: normal pooled plasma and samples A-E, prepared by mixing normal pooled plasma with increasing parts (1,2,3,8 and 14.7) of patient pooled plasma with a high D-dimer concentration. For validation 15 plasma samples from patients with venous throm beombolism were utilized. Results: A reference regression line through the mean values of D-dimer in samples A-E was calculated and used to harmonize the results of the 15 validation samples. After harmonzation the coefficients of variation improved considerably form 89% to 19% (mean values) for validation samples with high D-dimer conentration. For validaton samples with low D-dimer concentration (normal or around cut-off) coefficients of variation even increased from a mean value of 86% before harmonization to a mean value of 224% after harmonzation. Conclusions: The harmonziation procedure considerably decreased variability between different D-dimer assays in samples with a high D-dimer concentration. However, for samples with D-dimer concentration around cut-off, which is critical for clincal application, no improvement was observed.