Thrombus-trapping efficiency of the Greenfield filter in vivo

A total of 100 radiopaque blood clots were injected into seven adult sheep under fluoroscopic observation to assess the clot-trapping efficiency of the Greenfield vena caval filter. Eleven percent of the clots passed through the filter, with all failures limited to the two smallest clot sizes: 4 x 10 mm (five of 25) and 4 x 30 mm (six of 25). The Greenfield filter was 100% efficient in trapping larger clots (8 x 10 mm and 8 x 30 mm). There was a tendency toward poorer clot-trapping performance when the Greenfield filter was tilted within the vena cava, but this was not significant. There were no significant differences in hemodynamic measurements obtained immediately before and after injection of each clot, and no significant changes were noted in the animal's hemodynamic status or in blood gas measurements in those cases in which the clots passed through the filter. Since the caval diameter in sheep approximates that in humans, the authors conclude that the Greenfield filter in the clinical setting is capable of stopping large, likely fatal, emboli and a large number of smaller, possibly less clinically significant, emboli.