Determination of umbilical venous catheter tip position with radiograph.

OBJECTIVES: To compare the cardiac silhouette method with the vertebral body method in predicting the umbilical venous catheter tip position on ultrasound; to measure the length of the target zone for the umbilical venous catheter tip; and to determine the time taken for a neonatologist to ascertain position of the umbilical venous catheter tip with ultrasound.

DESIGN: Prospective cohort study.

SETTING: Neonatal ICU.

PATIENTS: Newborn infants with an umbilical venous catheter.

INTERVENTIONS: Ultrasound scans to determine the umbilical venous catheter tip position were performed within an hour of corresponding anteroposterior chest-abdominal radiograph.

MEASUREMENTS AND MAIN RESULTS: Two hundred paired radiograph and ultrasound scans in 82 newborn infants were analyzed. Each radiograph was reviewed independently by an experienced neonatologist who recorded the position of the umbilical venous catheter tip by vertebral level and by the cardiac silhouette method. For each method, the sensitivity, specificity, and positive and negative predictive values were calculated for the prediction of the true position of the catheter tip on ultrasound. The umbilical venous catheter tip was well positioned in just 28 of 200 scans. The cardiac silhouette method was superior to the vertebral level method for all test variables, with a sensitivity and specificity of 86% and 94% compared with 61% and 74%. The length of the target zone approximates to a single T8 vertebral body height on radiograph.

CONCLUSIONS: For radiograph and ultrasound scans performed within an hour of each other, the cardiac silhouette method more accurately predicts umbilical venous catheter tip than vertebral body level and methods described in previous studies. Catheters are frequently malpositioned. The length of the target zone for optimal umbilical venous catheter tip position is short. Ultrasound assessment of umbilical venous catheter tip position is quick.