A Randomized, Controlled Trial of Varying D-Dimer Cutpoints According to Clinical Pretest Probability Compared with a Uniform D-Dimer Cutpoint for Diagnosis of Deep Vein Thrombosis

D-dimer assays are currently used as a one-size-fits-all diagnostic test for DVT where the same cutpoint is used to define a positive result for all patients. This approach ignores the relationship between the diagnostic properties of the assay and the pretest probability of the disease.

We conducted a randomized, allocation concealed, controlled, assessor-blinded trial to compare the safety and efficiency of two diagnostic strategies for D-dimer testing for first acute DVT of the lower limb: Selective use of D-dimer testing, using different D-dimer cutpoint levels according to clinical pretest probability (C-PTP) and the traditional Uniform approach.

1723 consecutive eligible inpatients and outpatients at 5 clinical centres (4 in Ontario, 1 in Montreal) had their C-PTP assessed using the Wells' Score and were randomized to either: Uniform D-dimer testing, in which all patients had a D-dimer test and a level of <0.5 μ g fibrinogen equivalent units (FEU)/mL was classified as negative (the control group); or Selective D-dimer testing, in which D-dimer testing was (a) performed in patients with Low and Moderate C-PTP only, (patients with a High C-PTP had an ultrasound, but no D-dimer testing), and (b) was classified as negative if the D-dimer level was <0.5 μ g FEU/mL with Moderate C-PTP or <1.0 μ g FEU/mL with Low C-PTP (the experimental group). The D-dimer assays used in this study were the MDA D-Dimer (bioMerieux, Inc and Trinity Biotech) and STA Liatest D-Dimer(Diagnostica Stago). All patients with a positive D-dimer had an ultrasound of their leg. Symptomatic venous thromboembolic events (VTE) during 3 months of follow-up were objectively confirmed and independently adjudicated.

The incidence of VTE during 3 months of follow-up in patients who were not diagnosed with DVT at initial presentation was 1.1% (9/811) with Uniform testing, and 1.1% (9/814) with Selective testing (difference 0.004%, 95% CI:-1.122, 1.111). The incidence of VTE during 3 months of follow-up in patients in the Low C-PTP Select arm (D-dimer cutpoint 1.0 μ g FEU/mL) was 0% (0/356), and in the Low C-PTP Uniform arm (D-dimer cutpoint 0.5 μ g FEU/mL), it was 0.3% (1/336).

This is the first randomized controlled clinical trial to show that Selective use of D-dimer testing, with a higher D-dimer cutpoint in patients with Low C-PTP, is safe and potentially more efficient than current practice. Funded by the Heart & Stroke Foundation of Ontario.

Linkins:Trinity Biotech: in-kind donation of D-dimer kits; BioMerieux: in-kind donation of D-dimer kits; Stago Diagnostica: in-kind donation of D-dimer kits. Bates:bioMerieux: in-kind donation of D-dimer kits; Trinity Biotech: Consultancy, in-kind donation of D-dimer kits. Kearon:BioMerieux: in-kind donation of D-dimer kits; Trinity Biotech: in-kind donation of D-dimer kits; Stago Diagnostica: in-kind donation of D-dimer kits.