Diagnosis of lower limb deep venous thrombosis in emergency department patients: performance of Hamilton and modified Wells scores.

Abstract

We validate and compare the Hamilton score for assessment of lower limb deep venous thrombosis with the modified Wells score. Consecutive patients presenting to the emergency department of a tertiary center for suspected lower limb deep venous thrombosis were prospectively recruited. Hamilton score and modified Wells score calculations, D-dimer, and complete (calf veins included), single lower limb ultrasonographic examination were performed for all patients. All patients with a negative ultrasonographic examination result for deep venous thrombosis were followed up for 3 months. The study population consisted of 116 men and 193 women, with an average age of 55.6 years (SD 20.1). A total of 67 (21.7%) patients were diagnosed with deep venous thrombosis. Forty (59.7%) of these patients had isolated calf deep venous thrombosis, and the other 27 (40.3%) patients had proximal deep venous thrombosis. The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), positive likelihood ratio (LR+), and negative likelihood ratio (LR-) for an unlikely Hamilton score (score < or = 2) and D-dimer were 99% (95% confidence interval [CI] 92% to 99.96%), 42 % (95% CI 36% to 49%), 32% (95% CI 26% to 39%), 99% (95% CI 95% to 99.98%), 1.7 (95% CI 1.52% to 1.9%), and 0.04 (95% CI 0.01 to 0.25), respectively. The sensitivity, specificity, PPV, NPV, LR+ and LR- for an unlikely modified Wells score (score < or = 1) and D-dimer were 99% (95% CI 92% to 99.96%), 33 % (95% CI 27% to 39%), 29% (95% CI 23% to 35%), 99% (95% CI 93% to 99.97%), 1.47 (95% CI 1.34 to 1.62), and 0.05 (95% CI 0.01 to 0.32), respectively. An unlikely probability of Hamilton score and a negative D-dimer may effectively exclude a lower limb deep venous thrombosis. Hamilton and modified Wells scores have similar performance characteristics.