Appropriate Use of Venous Imaging and Analysis of the D-Dimer/Clinical Probability Testing Paradigm in the Diagnosis and Location of Deep Venous Thrombosis.

Background: The D-dimer (DD) level combined with the pretest Wells criteria probability (WCP) score can safely exclude deep venous thrombosis (DVT). The objective of this study was to examine the correlation between DD results alongside WCP score with findings on venous duplex ultrasound (VDU). The hypothesis is that VDU remains overutilized in low-risk patients with negative DD and that higher DD levels may correlate with thrombus burden and location.

Methods: Patients who presented to a high-volume tertiary care center with lower limb swelling with or without associated pain were retrospectively examined through June and July for 4 consecutive years (2012 to 2015). After calculating WCP, patients were divided into low-, moderate-, and high-risk categories. Electronic DD results utilizing enzyme linked immunosorbent assay, WCP data, and VDU analysis data were merged and analyzed based on receiver operator characteristic curve to determine the DD cutoff point for each WCP. Abnormal DD with an average value ≥ 0.6 mg/L fibrinogen equivalent units (FEUs) was correlated to positive DVT to differentiate proximal DVT (above popliteal vein) from distal DVT (below popliteal vein).

Results: Data of 1,909 patients were analyzed, and 239 (12.5%) patients were excluded secondary to serial repeat visits or follow-ups, surveillance screens, and if they had a previous history of DVT. The average age was 62.1 ± 16.3 years with more women (55.7%) and the majority presented with limb pain and edema (87%). DD studies were ordered and completed in 202 patients and correlated with all positive and negative DVT patients (100% sensitivity and negative predictive value, with specificity and positive predictive value of 14.9% and 15.9%, respectively). Twenty-six of 202 patients had DD that were in the normal range 0.1-0.59 mg/L (FEU), all of which were negative for DVT (100% sensitive). Fifty one of 202 patients had DD values of 0.6-1.2 mg/L FEU, of which only 3 DVTs were recorded, and all of them were distal DVTs. In addition, 685 patients with WCP <1 and negative DD were sent for VDU. Thus, 762 patients had an unnecessary immediate VDU (Wells ≤1 and -DD) study during their initial presentation. Potential charge savings for VDU for all patients are 762 × $1,557 = $1,186,434 and DD for all patients are 762 × $182 = $138,684, with total potential savings of $1,047,750 (USD 2016).

Conclusions: This study suggests that DD is still underutilized, and DD in conjunction with WCP could significantly reduce the number of unnecessary immediate VDUs. Higher value of DD (>1.2 mg/L FEU) may raise concern for proximal DVT. Concern on cost-effectiveness exists and raises the demand for a proposed algorithm to be followed.