Real-time ultrasonography for placement of central venous catheters in children: A multi-institutional study.

Abstract

Recommendations for the use of real-time ultrasonography for placement of central venous catheters in children are based on studies involving adults treated by nonsurgeons. Our purpose was to determine the frequency of use of real-time ultrasonography use by pediatric surgeons during central venous catheter placement, patient and procedure factors associated with real-time ultrasonography use, and adverse event rates. Using data gathered from 14 institutions, we performed a retrospective cohort study of patients <18 years old who underwent central venous catheter placement. Patient demographics and operative details were collected. We used a logistic regression model to evaluate factors associated with real-time ultrasonography use. Real-time ultrasonography was used in 33% of attempts (N = 1,146). The subclavian vein (64%) was accessed preferentially for first site insertion. Real-time ultrasonography was less likely to be used for subclavian vein (odds ratio = 0.002; P < .0001) and more likely to be used when coagulopathy (international normalized ratio >1.5) was present (odds ratio = 11.1; P = .03). The rate of mechanical complications was 3.5%. Real-time ultrasonography use was associated with greater procedural success rates on first-site attempt, but also with a greater risk of hemothorax. Pediatric surgeons access preferentially the subclavian vein for central venous access, yet are less likely to use real-time ultrasonography at this site. Real-time ultrasonography was superior to the landmark techniques for the first-site procedure success, yet was associated with greater rates of hemothorax. Prospective trials involving children treated by pediatric surgeons are needed to generate more definitive data.