LAUSR

BACKGROUND Venous thrombosis (VT) in children is often associated with a central venous catheter (CVC). We aimed to determine the incidence of VT associated with percutaneous non-tunnelled CVCs in a general paediatric population, and to identify risk factors for VT in this cohort. METHODS Observational, prospective study enrolling consecutive patients at a tertiary multi-disciplinary paediatric hospital. A total of 211 percutaneous, non-tunnelled CVCs were analysed. Data regarding potential risk factors for CVC-related VT were collected. Compression ultrasonography with colour Doppler was used to diagnose VT. RESULTS Overall, 30.3% of children developed CVC-related VT, with an incidence rate of 29.6 (confidence interval, 22.5-36.9) cases/1000 CVC days. Upper body CVC location, multiple lumen CVCs, and male gender were independent risk factors for VT in multivariate analysis. All upper body VTs were in the internal jugular vein (IJV). The occurrence of CVC-related VT did not affect length of paediatric ICU or hospital stay. In patients with VT, femoral CVCs, young age, paediatric ICU admission, and a ratio of CVC/vein diameter >0.33 were associated with VT being symptomatic, occlusive, or both. IJV VT was often asymptomatic and non-occlusive. CONCLUSIONS Paediatric non-tunnelled CVCs are frequently complicated by VT. Avoiding IJV CVCs and multiple lumen catheters could potentially reduce the overall risk of VT. However, IJV VT was more likely to be smaller and asymptomatic compared with femoral vein VT. More data are needed on the risk of complications from smaller, asymptomatic VT compared with the group of VT with symptoms or vein occlusion. Femoral vein CVCs and CVC/vein diameter >0.33 could be modifiable risk factors for VT with larger thrombotic mass. CLINICAL TRIAL REGISTRATION ACTRN12615000442505.