Vena caval filter placement by intravascular ultrasound

Objective: Placement of vena caval filters under fluoroscopic surveillance incurs significant expense and potential risks associated with the transpor tation of critically Ill patients. Intravascular ultrasound (IVUS) allows d irect intraluminal visualization of the vena cava and the renal veins. The purpose of this study is to evaluate the accuracy of vena caval filter plac ement under IVUS in an animal model. Methods: Fifteen Simon-Nitinol venal cava filters (C.R. Bard, Inc., Covingt on, GA) were placed under IVUS guidance into four anesthetized sheep. Twelv e were placed transfemorally, and three were placed transjugularly. Accurac y of placement was confirmed with fluoroscopy by measurement between the fi lter tip and the targeted side branch. Results: The vena caval filters placed femorally averaged 0.33 +/- 0.42 cm distance from the target vein side branch. Jugular approach filter placemen t was less accurate. Although two out of three filters placed from the jugu lar vein were correctly positioned, the distance from the target vein side branch was much greater averaging 2.5 +/- 1.04 cm. Conclusion: Femoral placement of vena caval filters under IVUS is extremely accurate. The transjugular route, however, was technically challenging and standard fluoroscopic vena caval filter placement appears to be more appro priate. Our success with the femoral approach merits further clinical inves tigation in the use of IVUS for critically ill patients that would benefit from bedside vena caval filter placement. (C) 2001 The International Societ y for Cardiovascular Surgery. Published by Elsevier Science Ltd. All rights reserved.