Follow-up after insertion of Bird's Nest inferior vena caval filters

To determine the long-term clinical and radiographic outcome of patients who undergo insertion of a Gianturco-Roehm Bird's Nest vena caval filter (Cook Inc., Bloomington, Ind.).

The medical records of 40 patients who, over a 34-month period, underwent insertion of a Bird's Nest filter were reviewed and the reasons for filter insertion determined. The causes of any subsequent deaths were noted, and the autopsy findings, when available, were reviewed. Surviving patients were contacted, and 12 were willing to return for follow-up imaging, which consisted of plain radiography, real-time ultrasonography, colour Doppler imaging and contrast-enhanced computed tomography. Changes in filter position, the presence of thrombus and perforation of the vessel wall by the filter struts were documented.

All 10 patients who underwent filter insertion while in an intensive care unit (ICU) died, an average of 22 days after the procedure. Eleven other patients also died. Recurrent pulmonary embolism was not suspected in any of these patients, and five autopsies revealed no caval thrombosis. Imaging studies in 12 of the surviving patients revealed no occlusion of the inferior vena cava and no filter migration; however, the vena caval wall was perforated in all of these patients. Nonocclusive intrafilter thrombus was detected by colour Doppler imaging in three patients.

In this small group of patients the Bird's Nest filter was effective in preventing recurrent pulmonary embolism and caused less caval thrombosis than has previously been suspected, although intrafilter clot was found in 25% of the patients who underwent follow-up imaging. Colour Doppler imaging is the method of choice for detecting nonocclusive thrombus with this type of filter. Perforation of the caval wall was universal but not clinically symptomatic. Finally, guidelines should be established to ensure the cost-effective use of inferior vena caval filters in ICU patients.