D-dimer assays – pitfalls of analytical comparisons

When a new D-dimer method is evaluated it is typically compared to an established D-dimer method. Usually both an analytical and a clinical comparison are made. The range of clinical specificities between D-dimer assays is wide (from 30 % to above 90 %), which is due to heterogeneity between D-dimer molecules and difference between antibody specificity towards them, especially concerning the preference for high or low molecular weight degradation products and for crosslinked over non-crosslinked degradation products. Discrepancies between studies of the same assays are predominantly related to differences between study populations. Many conditions may give elevated D-dimer concentration without the presence of venous thromboembolism (VTE), which will influence performance. The mixture of these conditions is most often not known or fully disclosed in published studies. Thus the discrepancies between methods and studies mean that you cannot directly compare two analytical results obtained with different methods. We have experienced that D-dimer assay correlation studies are too often presented and interpreted in a misleading way in literature. A hypothetical comparison of 3 assays is presented. A comparator assay with sensitivity 100 % and specificity 45 % is used to compare 2 other assays, with sensitivity 95 % and specificity 45 %, and with sensitivity 95 % and specificity 60 %, respectively. The examples show that the assay that has the better clinical performance has a poorer correlation, a lower overall agreement and a higher negative calculated bias at the cut-off vs. the comparator assay. Thus it must be emphasized that D-dimer assay comparisons should be based on clinical performance where the same samples are measured with all the methods compared. Analytical comparisons should be discouraged, as they do not give any useful information about the potential clinical usability of the assay. Page 1 Article downloaded from acutecaretesting.org Suzanne Ekelund and Maja Eliasen: D-dimer assays – pitfalls of analytical comparisons Maja Eliasen, MSc Technical Product Manager Radiometer Medical ApS Åkandevej 21 2700 Brønshøj, Denmark Phone: +45 38272412 E-mail: [email protected] Suzanne Ekelund, MSc Principal Specialist Clinical Biochemist Radiometer Medical ApS Åkandevej 21 2700 Brønshøj , Denmark Phone: +45 3827 3116 E-mail: [email protected]