The incidence of catheter-associated venous thrombosis in noncritically ill children.

OBJECTIVE: Previous studies estimated the incidence of catheter-associated venous thrombosis to be between 2% and 81%. Our goals were to define the incidence rate of catheter-related thrombosis in a hospitalized, noncritically ill, pediatric population and to determine modifiable factors that alter the risk of thrombosis.

METHODS: A retrospective cohort study was performed at the North Carolina Children's Hospital from 2009 to 2012. Chart review was performed with extraction of patient characteristics and line-related variables. Presence of symptomatic catheter-associated venous thrombosis was the primary outcome. Bivariable analysis and multivariable logistic regression were used to explore associations between line-related variables and thrombosis.

RESULTS: A total of 1135 lines were placed in 815 patients for 118,023 catheter-days. Thirty-six were complicated by venous thrombosis (3.2%) yielding a rate of 0.3 events per 1000 catheter-days. In multivariable analysis, increasing age (odds ratio [OR] 1.08, 95% confidence interval [CI] 1.03-1.13; P=.002), renal dialysis (OR 3.2, 95% CI 1.09-9.66; P=.035), and a diagnosis of inflammatory bowel disease or short bowel syndrome (OR 4.3, 95% CI 1.2-15.0; P=.02) were associated with increased risk of thrombosis. Modifiable risk factors, such as line site, size, and lumens, were not significantly associated with thrombosis. No thromboembolic events were observed.

CONCLUSIONS: We observed a lower incidence rate of catheter-associated venous thrombosis than in most previous reports. No modifiable characteristics altered the risk of thrombosis. Additional investigation of measures to prevent thrombosis is warranted in higher-risk populations, such as patients undergoing dialysis or patients with inflammatory bowel disease.