March 26th, 2013

Should the Use of IVC Filters be Filtered Until We Have More Data?

Preventing pulmonary embolism (PE) by interrupting the inferior vena cava (IVC), first introduced by Trousseau during the mid-1800s, has received growing attention lately with recent studies showing a remarkable increase in their use (e.g., Stein et al and Duszak et al). But what is the evidence base behind this mechanistically appealing approach? The largest randomized trial for the use of IVC filters was the PREPIC (Prevention du Risque d’Embolie Pulmonaire par Interruption Cave) trial. PREPIC enrolled 400 (of the anticipated 800) patients with proximal deep vein thrombosis (with or without PE), 94% of whom received anticoagulation for at least three months. Placement of IVC filters was associated with reduced risk of PE,  but increased risk of deep vein thrombosis, with no mortality difference at 2-year or 8-year follow-up.

The available expert statements and guidelines, including those by the AHA and the ACCP, however, recommend IVC filters mostly for cases with contraindications to anticoagulation or those with recurrent PE despite receiving anticoagulation;  two reasonable recommendations  for which we lack data from high-quality, comparative effectiveness studies.

Lack of convincing evidence, costs and expertise considerations, as well as local culture effects may lead into variations for use of IVC filters in the spectrum of patients with venous thromboembolism (VTE). Using the administrative data from 263 hospitals in California, White and colleagues recently showed a remarkable variation in use of IVC filters for acute VTE hospitalization across hospitals, from 0% to 38.96% (interquartile range: 6.23%-18.14%). Even among geographically close hospitals, such variations were remarkable. Of note, the hospital effect persisted even after extensive adjustment for clinical factors and comorbidities.

Despite the potential benefits, placing filters is costly and is not risk-free, making it a challenging decision for patients, physicians, and society, and the lack of high-quality evidence hampers the selection of the best possible decisions. There is a real need for definitive trials to answer the questions about the utility of IVC filters for the existing indications, as stated recently. I have heard that a group of investigators are about to launch a large, multicentric randomized trial in that regard. Until then, perhaps anticoagulation should remain the cornerstone of VTE treatment. While use of IVC filters is most plausible in patients with contraindications for anticoagulation, we simply lack high-quality comparative effectiveness data to support this appealing option.

What are your reflections on this issue? Where do you see the future of IVC filters?

4 Responses to “Should the Use of IVC Filters be Filtered Until We Have More Data?”

  1. Benjamin Peter Geisler, MD, MPH says:

    In a new article by Sarosiek et al., 8% of participants had a venous thrombo-embolic event (one-third of these a pulmonary embolism); however, only half of them had a known VTE before placement, and 25% were discharged on anticoagulants. Accompanying these interesting findings are a study on the variation of IVF use (as cited above), a short editorial, and a narrative review. The authors rightly point out the lacking evidence of beneficial health outcomes after IVFs, the infrequent use of retrievable filters, and the sometimes wrong initial assessments that prevent anticoagulant use in the first place. They demand a randomized controlled trial from the manufacturers or a federally-funded one. In my opinion, those would possibly lead to more subgroup-related questions. After all, patients receiving IVFs are a very heterogeneous group, and further retrospective or post-hoc analysis of prospective studies might elucidate certain populations with very different risk/benefit profiles. (cross-posted from

  2. Joel Wolkowicz, MDCM says:

    I concur. It seems an invasive and expensive procedure should be justified by evidence or lack or alternative therapy.

  3. William DeMedio, MD says:

    Initial studies should focus on finding out why there has been an increase in the use of filters and what is driving it. Filters do have their place, but the scientific evidence behind them is lacking. I will use this post as a “think twice” flag when a patient of mine has a descision to make regarding the use of an IVC filter. Invasive procedures should only be used where there is clear patient benefit.

  4. Judith Andersen, AB, MD says:

    There are several issues that have contributed to the increased use of IVC filters, the most important of which is the sense that the use of retrievable filters may supplant the use of periprocedural pharmacologic VTE prophylaxis. And there are orthopedic surgeons who arrange insertion and removal of these filters as if they are dietary supplements, with no consequences for their placement and removal.
    It’s a cinch — they provide temporary protection against potentially lethal pulmonary embolism — but no protection against clot progession through the filter and subsequent embolization. And their “retrievable” characteristics make them seem innocuous, when most are either neither retrieved or retrieved and then replaced for a second surgical procedure, insuring repetitive damage to the inferior vena cava.

    I have no quarrel with good mechanics, but question the wisdom of a procedural strategy that does not address the periprocedural pathophysiology. VTE prevention pharmacology is now well-honed, straightforward and, in most circumstances safe and effective. There are certainly patients whose situations mandate mechanical protection, but far fewer than the current trend suggests. And the glib insertion, removal, and reinsertion for a subsequent procedure is a prescription for disaster, particularly in older patients — who constitute the major population for which this strategy is employed.