Abstract 10162: Safe and Cost Effective Exclusion of Deep Vein Thrombosis by a Normal Elisa Dimer and Compression Ultrasound: A Prospective Study 1330 Patients
Introduction: A negative compression ultrasonography (CUS) excludes deep vein thrombosis (DVT) with a negative predictive value (NPV) of 98%. A normal rapid ELISA D-dimer test (VIDASR) excludes DVT with a sensitivity and NPV of 99%. The incidence of DVT after a negative CUS and ELISA dimer level above 500 ng/mL is predicted to be increased (5%) indicating the need to repeat CUS. Therefore, we performed a prospective study in 1330 consecutive outpatients with suspected DVT.
Methods: A negative CUS and a VIDASR of <500 ng/L exclude a DVT. A negative CUS with a D-dimer result above 500 ng/L was followed by a second CUS within one week. Patients with DVT on a 1st or 2nd CUS were treated with low molecular weight heparin followed by vitamin K antagonist.
Results: Out of 1330 with suspected DVT a 1st CUS was negative in 1059, and positive for DVT in 271: 20.3%. In 384 with a normal VIDASR (<500ng/mL), the CUS was true negative in 382 and false negative on CUS in 2, indicating a sensitivity of 99.52%, a NPV of 99.48%, and a specificty of 36% irrespective of clinical score . A 2nd CUS was indicated in 675, performed in 641 patients and proved to be positive in 26 (4.0%), indicating a NPV of 96% after a 1st negative CUS. The overall prevalence of DVT after a 1st or 2nd CUS was 297 of 1330 (22.3%): 18% proximal DVT and 2.4% distal DVT. The NPV of a negative 1st CUS and a VIDASR of less than 1000 ng/mL was 99.1% at a specificty of 66.9% irrespective of clinical score. The prevalences of DVT at increasing D-dimer cut of levels of 0 to 500, 500 to 1000, 1000 to 1500, and above 1500 ng/mL were 0.5%, 6.3%, 15.1% and 47.4% respectively irrespective of clinical score.
Discussion: A 1st CUS detected a number of alternative diagnoses in patients with suspected DVT. A 1st CUS and a VIDASR <1000 ng/mL excluded DVT with a sensitivity of 99.1% irrespective of clinical score. The combination of a 1st negative CUS and a normal VIDASR test (<500 ng/mL) irrespective of clinical score, and the combination of low clinical score (asymptomatic for DVT), a normal 1st CUS and an VIDASR of <1000 ng/mL is predicted to safely exclude DVT with a sensitivity of 99.99% and a NPV of 99.99% without the need of repeated CUS testing in two-third of outpatients with suspected DVT. This strategy is predicted to be cost-effective by reduction of repeated CUS by 67%.
- © 2013 by American Heart Association, Inc.