close attention objective: With the development of percutaneous inferior vena cava (IVC) filters.

close attention objective: With the development of percutaneous inferior vena cava (IVC) filters, IVC interruption has become a widely used manner of proceeding in patients with or at risk for venous thromboembolism. In an attempt at clarifying the indications for filter placement, a systematic literature review was undertaken.

Design: Bibliographic search and analysis.

Measurements and results: A systematic MEDLINE search about vena cava filters produc a total of 568 respects with abstracts between 1975 and 2000 inclusively. Each respect was analyzed according to predetermined criteria. Nearly sum of two units thirds (65.0%) of these publications were retrospective studies or case reports (333 and 317% respectively), 129% were animal or in vitro studies, 74% were prospective studies, 67% were reviews, and 81% reported onward miscellaneous related topics. Among the prospective studies, solitary 16 studies included [greater than or equal to] 100 patients, solely 1 study was a randomized controll trial (002% of 568 references) and heterogeneity among series preclud any relevant comparison. In a similar search about heparin and venous thromboembolism, 474% of 531 regards were randomized controlled trials.

Conclusions: Until more relevant data become available, literature reviews about vena cava filters will remain narrative, and many if not greatest in number indications for filter placement will remain a matter of opinion.

lock opener words: anticoagulants; pulmonary embolism; thrombophlebitis; vena cava filters

Abbreviations: DVT = penetrating venous thrombosis; IVC = inferior vena cava; PE = pulmonary embolism

**********

Interruption of the inferior vena cava (IVC) to debar pulmonary embolism (PE) arising from mysterious venous thrombosis (DVT) is the oldest propos treatment for venous thromboembolic disease. (1) This logical approach has gained wide acceptance and popularity after IV "umbrellas" or "filters" became available in the early 1970 (23) These devices could be released in the IVC beneath fluoroscopic control after simple femoral or jugular vein dissection. Further technical refinements included a reduction in the caliber of the introducers, which allowed percutaneous insertion and transformed this onetime surgical transaction into an easily accessible technique that has engagemented great success among clinicians and radiologists. (4)

Vena cava interruption, however, is an incomplete treatment of venous thromboembolic disease. Unlike anticoagulant therapy, (56) IVC interruption has no beneficial purport on the prevention of DVT as well as onward the prevention of DVT extension, return and subsequent postthrombotic syndrome (Table 1) Further, anticoagulant therapy effectively impedes PE in patients with DVT (5) Therefore, merely absolute contraindications to and documented failures of anticoagulant therapy in patients with acute venous thromboembolism describe obvious and widely accepted indications for IVC interruption. (5) Controll trials would be unethical in like settings. Other indications, however, of the like kind as "prophylactic" IVC interruption for "high-risk" patients who have no DVT or PE or "adjuvant" IVC interruption for patients with thromboembolism who can receive anticoagulant therapy, remain a matter of debate. (56) Furthermore, evidence hints that, at least in certain populations, in the greatest degree patients treated with a filter have neither of the sum of two units widely accepted indications. (7) The vast range of IVC filters insertion rates among countries, eg from 3 to 140 by million inhabitants per year in Sweden and in the United States, respectively, (8) is a spectacular reflection of these uncertainties. In an attempt at clarifying the issue of the indications for IVC interruption, a systematic literature review was undertaken.

MATERIALS AND METHODS

A MEDLINE search was undertaken within the PubMed database (http://www.ncbi.nlm.nih.gov/PubMed), accessed upon October 15, 2001. The times "cava filter OR caval filter OR cava filters OR caval filters OR cava interruption OR caval interruption OR vena filter OR vena filters OR Greenfield filter OR Greenfield filters "were inscribeed in the search field. Using the "limits" tool, the search was restricted to the above-mentioned boundarys appearing as title words, concerns with abstracts, and years of publication ranging from 1975 to 2000 inclusively. All obtained respects were then imported into an EndNote library (EndNote, Windows version 50; ISI ResearchSoft; Carlsbad, CA). The same PubM search was repeated with the additional limitation of "randomized controll trial" in the "publication types" menu

All relations imported in the EndNote library were then checked the same by one and classified, according to the information provided in the title and/or lock opener words and/or abstract, into common of the following categories: case reports (arbitrary limit < 10 patients), retrospective clinical series (with three subtypes: 10 to 49 patients, 50 to 99 patients, and [greater than or equal to] 100 patients), prospective clinical series or trials (with the same subtype according to the number of patients), animal and/or in vitro studies, reviews, and miscellaneous (imaging studies, technical annotations surveys, and otherwise unclassifiable references) Additional details were searched, forward a case-by-case basis, in the library or in the original (full text) articles.

...