Abstract
Background: Exclusion of deep venous thrombosis (DVT) using a pretest probability score (PTP) and a normal D-dimer concentration has proven to be safe and feasible. However, the high frequency of false positive D-dimer results in the elderly hampers its use, necessitating additional compression ultrasound (CUS) in these patients. We evaluated an alternative approach for the exclusion of DVT, especially in the elderly, using a thrombin generation assay (TGA) in combination with a PTP and a D-dimer assay.
Patients and methods: In symptomatic outpatients suspected for DVT, the Endogenous Thrombin Potential (ETP) assay (Dade Behring Marburg GmBH, a Siemens Company, Germany) was tested. The results of this assay were expressed as ETP % NV (normalized value) [%], ETP AUC (area under the curve) [mE], ETP tlag (lagtime) [sec], ETP tmax (time to peak) [sec] and ETP Cmax (peak value) [mE/sec]. In all patients, a Tina-Quant D-dimer assay (Roche, Manheim, Germany) and a PTP according to Wells were performed. DVT was diagnosed or excluded using B-mode CUS.
Results: A total of 443 patients were included (38% men, mean age 60 y). Total DVT prevalence was 39%. In the total group, the Tina-Quant D-dimer assay had a sensitivity (sens) of 98.9%, a negative predictive value (NPV) of 98.3% and a specificity (spec) of 41.9%. Combining D-dimer with a PTP < 2, these values were 100%, 100% and 52.2% respectively. With a spec of 51.8%, a large number of false positive results remain.
All ETP parameters had a low AUC of the receiver operation curve (ROC), range 0.53–0.64.
For all ETP parameters, there was a significant difference (p < 0.001) between patients with and without DVT, except ETP tmax. ETP parameters were tested for additional value to Tina-Quant and PTP. Only ETP tlag demonstrated an additional value, using an optimal cut-off value of 23.0 sec.
In patients with a low PTP and an abnormal D-dimer value, that might be false positive, adding ETP tlag improved specificity to 87.5%, but sensitivity decreased to 26.5%. The percentage false positive results decreased from 38 to 8%, with an increase of false negative results. However, in the fourth age quartile cohort (patients >75 y), sensitivity reached 100% with a specificity of 94.1% and a decrease of false positive results from 58 to 5% (Table).
Discussion and conclusion: In this study, a safe exclusion of DVT is possible with a low PTP and a normal D-dimer result as first step. Combining ETP tlag with a low PTP and an abnormal Tina-Quant D-dimer as second step (figure), the sens and NPV reached values of 100% in the fourth age quartile cohort. This algorithm should be confirmed by a new management study.
Table. Performance of Tina-Quant (TQ) in relation to DVT, < 500 μg/l, low PTP + TQ < 500 μg/l and low PTP + TQ ≥ 500 μg/l + ETP tlag < 23.0 sec in the total cohort and the fourth age quartile; values (95% lower confidence limit)
. | Sens . | Spec . | NPV . | false pos . |
---|---|---|---|---|
Sens = sensitivity, Spec = specificity, NPV = negative predictive value and false pos = false positive | ||||
Total | ||||
Tina-Quant < 500 | 98.9 (96.1) | 41.9 (36.2) | 98.3 (94.2) | 35.4 |
PTP + TQ < 500 | 100 (92.6) | 51.8 (43.8) | 100 (94.9) | 37.7 |
PTP + TQ > 500 + tlag | 26.5 (12.9) | 87.5 (76.8) | 69.1 (57.9) | 8.1 |
Elderly | ||||
Tina-Quant < 500 | 100 (89.4) | 23.9 (14.6) | 100 (80.5) | 51.9 |
PTP + TQ < 500 | 100 (15.8) | 37.5 (18.8) | 100 (66.4) | 57.7 |
PTP + TQ > 500 + tlag | 100 (15.8) | 94.1 (65.3) | 100 (84.6) | 5.2 |
. | Sens . | Spec . | NPV . | false pos . |
---|---|---|---|---|
Sens = sensitivity, Spec = specificity, NPV = negative predictive value and false pos = false positive | ||||
Total | ||||
Tina-Quant < 500 | 98.9 (96.1) | 41.9 (36.2) | 98.3 (94.2) | 35.4 |
PTP + TQ < 500 | 100 (92.6) | 51.8 (43.8) | 100 (94.9) | 37.7 |
PTP + TQ > 500 + tlag | 26.5 (12.9) | 87.5 (76.8) | 69.1 (57.9) | 8.1 |
Elderly | ||||
Tina-Quant < 500 | 100 (89.4) | 23.9 (14.6) | 100 (80.5) | 51.9 |
PTP + TQ < 500 | 100 (15.8) | 37.5 (18.8) | 100 (66.4) | 57.7 |
PTP + TQ > 500 + tlag | 100 (15.8) | 94.1 (65.3) | 100 (84.6) | 5.2 |
Disclosures: No relevant conflicts of interest to declare.
Author notes
Corresponding author
This feature is available to Subscribers Only
Sign In or Create an Account Close Modal