What Happens To Removable Vena Cava Filters?

Abstract & Commentary

By David J. Pierson, MD, This article originally appeared in the May 15, 2007 issue of Critical Care Alert. It was peer reviewed by William Thompson, MD.

Synopsis: Among 413 patients who underwent placement of a removable inferior vena cava filter following trauma for prophylaxis or treatment of pulmonary thromboembolism and survived to hospital discharge, subsequent removal of the filter was attempted in 116 of them and successful in only 91 (22%). Fewer patients were lost to follow-up at hospitals with policies requiring the service that placed the filters to follow them.

Source: Karmy-Jones R, et al. Practice patterns and outcomes of retrievable vena cava filters in trauma patients: An AAST multicenter study. J Trauma. 2007;62:17-24.

Trauma surgeons at 21 institutions participated in this retrospective study of inferior vena cava filter (IVCF) placement and follow-up. A total of 599 IVCFs were placed at 21 hospitals during the study year, 226 (0.8% of all admissions) at 7 high-volume hospitals (> 2000 trauma cases admitted) and 373 (2% of all admissions, P = 0.009) at 14 low-volume institutions. Of all IVCFs in the study (both permanent and removable), 74% were placed prophylactically — that is, in patients without evidence for pulmonary thromboembolism or deep-venous thrombosis.

Of the 599 IVCFs, 446 (79%) were removable filters. Ten of the 21 participating centers had policies whereby the trauma service followed up with the patients following IVCF placement, as well as coordinated removal. In 3 centers, these aspects were the responsibility of the service placing the filter, and in the remaining 8 centers, no policies were in place. Of the 413 patients who had removable filters placed and survived to hospital discharge, filter removal was attempted in only 116 patients (28%) 50 ± 61 (mean ± SD) days after placement. Filter removal was not actually carried out in 25 of these patients because of technical difficulties or the presence of residual thrombus. Inability to remove the filter was more common with the Cordis Endovascular Optease filter (8/11 attempts) than with the Bard Recovery (9/50) and Cook Gunter-Tulip (8/54) filters (P = 0.01).

Losing the patient to follow-up was the most common reason for not attempting removal of the filter, and this was significantly more frequent in hospitals where the placing service was not required to follow-up with the patient (122 of 273 patients, 45%) than in the 3 institutions having this requirement (4 of 65 patients, 6%; P = 0.001). Among the other patients in whom filter retrieval was not attempted, the most common reasons for this were immobility (124 patients), other ongoing risk for recurrent thrombosis (12 patients), and persistence of deep venous thrombosis (11 patients).


This was not a prospective study, and the descriptions provided for case identification, patient follow-up, complication detection, and other aspects of the methods are fairly cursory. For this reason, the reported absolute complication rates and the differences between these among the various indications, institutions, and devices should be interpreted cautiously. However, despite the study's limitations, it conveys several important messages. The use of "removable" IVCFs has become widespread. Many (if not most) such devices are being placed prophylactically rather than because of the inability to adequately anticoagulate patients with known venous thromboembolism. Despite the attractiveness of the concept of removing the filter once the acute threat of thrombosis has passed, it is clear that this is not actually taking place in many (if not most) instances.

Although it may be argued that the risk of subsequent deep-venous thrombosis and other long-term vascular complications of IVCFs may be different with current devices as compared to those of a decade or two ago, the final word on this issue is certainly not in at present. The large number of patients lost to follow-up in this study is worrisome, in that these lost patients may also not be receiving therapy with anticoagulants and may be at increased risk for complications related to the filters.

Following up patients seen at major trauma centers is a difficult and complicated problem, for numerous reasons. However, this study's finding of much greater follow-up success at institutions with protocols for follow-up of patients with IVCFs by the service that placed them indicates that such continuity of care is possible, even for these challenging patients.