Duplex-directed vena caval filter placement: report of initial experience.
BACKGROUND: Fluoroscopy, cost, and patient transport contribute to difficulties occasionally associated with the placement of vena caval filters. Follow-up data in the literature document the use of duplex ultrasonography in visualizing the filter and determining caval patency. Filter placement at the bedside or in the vascular laboratory with duplex ultrasonography may simplify this common procedure. We have attempted to define the feasibility of this method. METHODS: Patients referred to the vascular surgery service for vena caval interruption were evaluated for ability to visualize the renal veins and inferior vena cava. Location of renal veins, maximum diameter of the vena cava, and presence or absence of thrombus were documented. If visualization was adequate, placement was performed at the bedside for patients in intensive care or in the vascular laboratory for nonmonitored patients. The initial 10 patients and subsequent patients in whom there was a question of adequate deployment underwent completion abdominal roentgenography. Patient follow-up was difficult. Duplex ultrasonography was used to assess migration, thrombus adherent to the filter, and vena caval patency. Patients in whom filter placement was prophylactic were given anticoagulants at the discretion of the primary physician. Inadequate visualization or vena caval size greater than 28 mm prompted fluoroscopic placement of the vena caval filter, because only Greenfield titanium filters were used in the study. RESULTS: Twenty-nine patients were referred for vena caval interruption. Inadequate visualization occurred in four obese patients, and filters were placed by fluoroscopy. There were no vena caval measurements greater than 24 mm. Twenty-five filters were placed without technical difficulty. One filter tilted into the right renal vein, requiring a suprarenal filter placed by fluoroscopy. Patient retrieval for follow-up has been difficult, but by ultrasonography there has been one vena caval thrombosis and no major filter migration. There have been no reported pulmonary emboli other than the one patient with initial tilt of the filter. CONCLUSIONS: Placement of vena caval filters is feasible with duplex ultrasonography. Visualization is the only limiting condition to placement and occurs rarely. Reducing the need for fluoroscopy, lowering costs, and not needing to transport the critically ill patient support the use of this system. Intravascular ultrasonography in selected patients may eliminate the need for fluoroscopic placement of vena caval filters.