Variable Performance of D-dimer Testing by Hemostasis Laboratories: The Australasian/Asia-Pacific Experience

Semin Thromb Hemost. 2024 Mar 1. doi: 10.1055/s-0044-1781450. Online ahead of print.


D-dimers represent the breakdown products of fibrin. Thus, elevated plasma D-dimers will arise following a thrombotic event, such as a deep vein thrombosis or a pulmonary embolism, and therefore, a nonelevated D-dimer is used to effectively exclude such events. D-dimers are also elevated in a range of other conditions, for example, during disseminated intravascular coagulation. D-dimer levels may also be associated with prognostic value. For example, highly raised D-dimer levels can be associated with worsening clinical features in coronavirus disease 2019. Thus, D-dimer testing represents a commonly requested hemostasis test, often performed in 24/7 laboratories. Unfortunately, D-dimer testing is neither standardized nor harmonized across manufacturers or laboratories. Indeed, considering reporting units and the magnitude of units, up to 28 different combinations may be reported by laboratories. We provide updated findings for D-dimer testing in our geographic region, using recent data from the Royal College of Pathologists of Australasia Quality Assurance Programs, an international external quality assessment program, currently with over 450 participants in the D-dimer module. Data show a wide variety of assays in use and variable outcomes in reported numerical values when assessing proficiency samples. D-dimer testing mostly comprised reagents from three main manufacturing suppliers, with a small number of users of reagents from other manufacturers. Reported results showed important differences in numerical values for the same homogeneous tested samples when normalized to a single reporting unit (e.g., mg/L). Nevertheless, despite using different test reagents and reporting, most participants uniformly identified D-dimer values as below or above a "detection" cut-off for samples that were constructed to be below or above most cut-off values. As expected, mixed findings were reported for samples containing levels around expected cut-off values. We hope that our findings, reflecting on the heterogeneity of test reagents and test data, help improve diagnostic testing for D-dimer testing and facilitate harmonization and standardization, in the future.