Clinical evaluation of eight different D-dimer tests for the exclusion of deep venous thrombosis in primary care patients

Abstract

D-dimer tests are an essential element in the diagnostic work-up of deep venous thrombosis (DVT). However, the poor standardization amongst assays necessitates clinical validation before implementation in daily practice. We therefore evaluated the analytical and diagnostic performance of eight D-dimer tests in a representative group of 290 prospectively identified consecutive primary care patients with suspected DVT. Seven quantitative D-dimer assays, and a qualitative test, Simplify, were evaluated. Correlation between assays was generally poor and several assays showed a significant bias in the method comparison. Nevertheless, the Vidas D-dimer, Innovance D-dimer (CA1500 and BCS), Pathfast D-dimer, and HemosIL HS500 (ACL TOP), all displayed 100% (95% CI: 85–100%) sensitivity. Tina-quant (Modular), AQT90 D-dimer, and Liatest (STA^®) D-dimer tests showed a slightly lower sensitivity of 95% (78–100%). and the Simplify test reached a sensitivity of 91% (72–99%) that was further improved in combination with a clinical decision rule to 95% (76–100%). In concert with the low (8.2%) prevalence of proximal DVT, diagnosed by compression ultrasonography, in our study, all test reached a negative predictive value (NPV) of at least 99%. The user friendliness of the assays differed mainly by stability of reagents, calibration frequency, time required to obtain a test result and costs of a test. In conclusion, despite considerable analytical differences, in our low-risk population all tests evaluated displayed an excellent NPV. In combination with a validated clinical decision rule to identify low-risk patients, even a straightforward POC solution could safely and cost-efficiently rule out DVT.

Key Words::

• Venous thrombosis
• fibrin fibrinogen degradation products
• primary health care
• humans
• clinical laboratory techniques
• immunoassay
• predictive value of tests
• point-of-care systems

Acknowledgements

We are grateful to Gert Schmaal, Willy Doornbos, Antonia Meijer, Vicky Sebo, Bert de Leeuw, Henk Bakker, and Bert Nienhuis, technicians and sonographists at Certe. The measurements of ACL TOP D-dimer were performed in collaboration with Medical Centre Alkmaar (Dr P.C.M. Bartels). The Liatest measurements were performed in collaboration with Medical Centre Leveste, location Scheper Hospital Emmen (J.G.J. Pouwels); and Theo Hofman (Roche). We thank Siemens Healthcare Diagnostics, Sysmex Netherlands, Biomerieux, Instrumentation Laboratory, Roche, Radiometer and Alere Health for providing the reagents. The study was performed independently of these companies and without any direct financial support.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.