We enrolled 85 of 121 eligible children. Overall, 37 subjects had DVT of which 25 were incident and 12 were prevalent (Fig. 1). Of the prevalent DVT, one was in the lower extremity unrelated to a CVC. Two subjects did not have repeat ultrasonography because of unexpected discharge from the hospital. The agreement in the readings between radiologists was 93.9%. We did not find any difference in the characteristics of the subjects and of the CVC, cointerventions, and laboratory results between subjects with and without incident DVT (Table 1). In contrast, subjects with prevalent DVT were older and had bigger CVCs than those without DVT (Table 2). The frequency of factor V Leiden and prothrombin G20210A mutations were not different between those with and without incident or prevalent DVT. Factor VIII activity was measured in all 85 subjects except for one who had incident DVT whose specimen clotted. G value and other TEG variables were measured in all subjects. Because the values measured with and without heparinase were not significantly different, we presented those measured without heparinase. The average time from insertion of the CVC to blood draw was 19.8 ± 8.4 hours (mean ± SD).
Association of Factor VIII Activity, G Value, and Incident DVT
Factor VIII activity was associated with incident CVC-related DVT. For every SD increase in factor VIII activity measured in the study population, the OR for incident DVT was 1.98 (95% CI, 1.10–3.55) (Fig. 2A). The AUC of factor VIII activity was 0.66 (95% CI, 0.52–0.79) (Fig. 3A). Combining factor VIII activity and G value did not significantly change the AUC (0.67; 95% CI, 0.54–0.80; p = 0.58). At the optimal threshold of more than 100 IU/dL, the sensitivity of factor VIII activity was 92.0% (95% CI, 74.0–99.0%), whereas its specificity was 41.3% (95% CI, 27.0–56.8%). The sensitivity and specificity of factor VIII activity at the previously reported threshold of more than 150 IU/dL were 36.0% (95% CI, 18.0–56.5%) and 80.4% (95% CI, 66.1–90.6%), respectively. When adjusted for age, center effect, and recent surgery, the association remained significant (OR, 1.93; 95% CI, 1.10–3.39). The OR for incident DVT also remained significant when the two subjects with no exit ultrasonography were analyzed as having (OR, 1.93; 95% CI, 1.08–3.43) or not having (OR, 1.96; 95% CI, 1.11–3.47) incident DVT. G value and the other TEG variables were not associated with incident DVT (Figs. 2A and 3B).
Association between (A) incident and (B) prevalent deep venous thrombosis (DVT) and factor VIII activity and thromboelastography variables. Coagulation index, a global measure of coagulation, is a combination of reaction time, kinetics time, angle, and maximum amplitude.
Receiver operating characteristic curves showing the performance of (A and C) factor VIII activity and (B and D) G value, expressed as area under the curve (AUC), in identifying (A and B) incident and (C and D) prevalent deep venous thrombosis (DVT).
Association of Factor VIII Activity, G Value, and Prevalent DVT
Unlike with incident DVT, factor VIII activity was not associated with prevalent DVT (Figs. 2B and 3C). G value, on the other hand, was associated with prevalent DVT. For every SD increase in G value measured in the study population, the OR for prevalent DVT was 2.21 (95% CI, 1.19–4.10) (Fig. 2B). The AUC was 0.68 (95% CI, 0.504–0.87) (Fig. 3D). Combining G value and factor VIII activity did not significantly change the AUC (0.69; 95% CI, 0.51–0.88; p = 0.46). The sensitivity of G value at the optimal threshold of more than 11.0 kdyn/cm was 50.0% (95% CI, 21.1–78.9%) with a specificity of 87.7% (95% CI, 77.9–94.2%). The sensitivity and specificity of G value at the previously reported threshold of more than 12.4 kdyn/cm were 33.3% (95% CI, 9.9–65.1%) and 94.5% (95% CI, 86.6–98.5%), respectively. When adjusted for age, center effect, recent surgery, and size of the CVC, the OR for prevalent DVT was 2.46 (95% CI, 1.18–5.13). The other TEG variables were not associated with prevalent DVT (Fig. 2B).