Use of D-dimer for the exclusion of new or recurrent proximal deep vein thrombosis of the lower limb in anticoagulated patients: A multicentre retrospective study
Elisabeth Gagnon, Cassandra Larivière, Mathieu Allard, Anne-Audrey Lapointe, Isabelle Viens, Shaima Kaka, Geneviève Le Templier, Dominique Toupin, Étienne Beaudoin, Samuel Lemaire-Paquette, Mandy Malick, Pierre-Nicolas Perron, Sarah Higgins, Venkatesh Thiruganasambandamoorthy, Alexandre Mutchmore, Gregoire Le Gal, Bobby GouinIntroduction:
D-dimer (DD) testing is widely used to exclude proximal deep vein thrombosis (DVT) in patients not receiving anticoagulation who have low or intermediate pretest probability, in whom it demonstrates high sensitivity and negative predictive value (NPV). However, its diagnostic accuracy in patients already receiving direct oral anticoagulants (DOACs) remains uncertain. This study aimed to evaluate the diagnostic accuracy of DD testing for excluding proximal DVT in patients receiving DOACs.
Methods:
A retrospective cohort study was conducted including adults receiving DOACs who presented to the emergency department with suspected DVT between January 2010 and September 2022. All included patients underwent both DD testing and compression ultrasonography. The diagnostic performance of the conventional DD threshold of 500 µg/L was assessed, focusing on sensitivity and NPV. Exploratory analyses of the age-adjusted DD threshold and sensitivity analyses restricted to definite proximal DVT were also performed.
Results:
Seventy-one patients were included (mean age, 74.6 years; 52% women). Nine patients were diagnosed with new thrombosis, of which 7 were proximal. Using the conventional threshold, DD sensitivity was 71% (95% CI, 29–96) and NPV was 89% (95% CI, 67–99). The age-adjusted threshold had a lower sensitivity of 57% (95% CI, 18–90), with a similar NPV (91%; 95% CI, 76–98). In sensitivity analyses restricted to definite proximal DVT ( n = 5), the sensitivity of the conventional DD threshold was 80% (95% CI, 28–99), with an NPV of 94% (95% CI, 73–100).
Discussion:
In contrast to patients not receiving anticoagulation, DD testing showed low sensitivity and NPV in patients receiving DOACs who had suspected DVT. These findings suggest that DD testing cannot safely exclude proximal DVT in this population. Compression ultrasonography should remain the preferred diagnostic strategy. Larger prospective studies are warranted to explore the role of DD testing in patients receiving anticoagulation and to develop a new diagnostic algorithm.