Routine surveillance ultrasound for the management of central venous catheters in neonates

Objectives To evaluate the frequency of central venous catheter (CVC)-related thrombi detected by routine surveillance ultrasound, and to assess whether positive findings had an impact on management or outcomes. Study design All neonates in a tertiary neonatal intensive care unit who had a CVC inserted for >14 days underwent routine surveillance ultrasound biweekly between January 2003 and December 2009. Data were reviewed retrospectively. Results Although all neonates were asymptomatic at time of surveillance ultrasound, 645 of the total 1333 CVCs inserted in 1012 neonates underwent surveillance ultrasound, and thrombi were detected in 69 (10.7%). The CVCs with thrombi were more likely to be removed for nonelective reasons compared with CVCs without thrombi (59% vs 38%; P =.001; OR, 2.4, 95% CI 1.4-3.9). A total of 955 surveillance ultrasounds were performed to detect and monitor 69 CVCs with thrombi. The majority of thrombi were nonocclusive and nonprogressive. A change in management occurred in 8 cases of CVC-related thrombi (12%), or 1% of all screened cases. An average of 14 ultrasounds were required to detect and monitor 1 CVC with thrombus, at a cost of $951 per CVC with thrombus and $8106 per case of CVC-related thrombi with a change in treatment. Conclusion Asymptomatic thrombi were detected in a significant proportion of CVCs by routine surveillance ultrasound. There were significant costs, but infrequent changes to patient management.