An unseen danger: Frequency of posterior vessel wall penetration by needles during attempts to place internal jugular vein central catheters using ultrasound guidance

Objectives. To evaluate the frequency of unsuspected posterior vessel wall penetration of the internal jugular vein during ultrasound-guided needle cannulation. Design: Prospective, single-blinded observational study. Setting: Urban level I emergency department with an annual census of 80,000. Patients. Residents who had previously completed a 2-day ultrasound course including a 3-hr didactic and hands-on session on ultrasound-guided central venous cannulation. Interventions Residents were asked to place an ultrasound-guided catheter on a human torso mannequin. Residents used a short-axis approach for ultrasound guidance. During the procedure, an 8-4 MHz convex (endocavity) transducer was used to observe the path of the resident's needle without interference with the placement procedure. Measurements and Main Result Unknown to residents, researchers tracked the frequency of posterior wall penetration and the final needle location when the resident felt that optimal needle placement was achieved in the lumen of the internal jugular. Residents were also asked to rate their confidence regarding appropriate final needle position on a 10-point Likert scale. Statistical analysis consisted of descriptive statistics and Spearman correlation analysis. A total of 25 residents partici- pated. All had placed at least one ultrasound-guided central catheter previously. The median number of previous ultrasound-guided cannulations was 8.0. Sixteen (64%) residents accidentally penetrated the posterior wall of the internal jugular vein during cannulation. The median number of posterior wall penetrations was 1.0 for all residents. In six cases the final location of the needle was through the posterior wall and deep to the venous lumen. In five of these cases the carotid artery was actually mistakenly penetrated. Median confidence by residents-regarding appropriate needle placement was 8.0 out of 10. More training and more ultrasound-guided catheters placed were associated with fewer posterior wall penetrations (p = .04). Conclusions. In this study, residents accidentally penetrated the posterior vessel wall of the internal jugular in a lifelike vascular access mannequin in the majority of cases. These results suggest that care must be taken even with ultrasound-guided central catheter placement and that alternative ultrasound guidance techniques, such as visualization of the vein and needle in longitudinal axis, should be considered. (Crit Care Med 2009; 37:2345-2349)