Central venous catheter–associated complications in infants with single ventricle

Abstract

Among infants with single-ventricle heart disease who require surgical palliation, central venous access is routinely obtained via the umbilical or femoral veins. Both routes are associated with potential complications, including thrombosis. We sought to analyze the clinical outcomes of patients with umbilical venous catheter vs. femoral central venous catheter placement at the time of initial central venous access in this high-risk patient population. This was a retrospective study, with data collected including demographics, catheter type, duration, complications, and clinical outcomes. Patients were designated as group 1 (initial umbilical venous catheter placed, n = 70) or group 2 (initial femoral central venous catheter placed, n = 19). The study was conducted at a single tertiary care referral institution. We included all 89 patients who underwent single-ventricle palliation at this institution in 2007 and 2008. The overall rates of survival to hospital discharge, thrombosis, and iliofemoral vein occlusion were 82%, 18%, and 21%, respectively. The proportion of thrombosis was 11% in group 1, compared with 42% in group 2 (p < .01). The proportion of iliofemoral vein occlusion was 16% in group 1, compared with 42% in group 2 (p = .02). The proportions of catheter-associated bloodstream infection, need for transhepatic access, and ultrasound-documented thrombus at the inferior vena caval-right atrial junction did not differ significantly between the groups. Patients with non-tunneled femoral central venous catheters for ≥14 days had a higher prevalence of thrombosis (52%) than those with femoral central venous catheters for <14 days (13%) but no difference in the prevalence of iliofemoral vein occlusion. In this population, initial placement of an umbilical venous catheter rather than a femoral venous catheter resulted in significantly lower risks of catheter thrombosis and iliofemoral vein occlusion. For femoral venous catheters, the prevalence of thrombosis, but not of iliofemoral vein occlusion, is proportional to the duration of catheterization.