Selective venography versus nonselective venography before vena cava filter placement: evidence for more, not less☆

Article Outline

• Abstract
• Methods
• Results
• Discussion
• Acknowledgements
• References
• Copyright

Abstract 

Objective

We undertook this study to determine whether additional use of selective venography, compared with nonselective venography alone, reveals more abnormal anatomic venous findings that lead to changes in vena cava filter (VCF) position.

Methods

From January 1998 to June 2002, 94 patients underwent VCF placement by vascular surgeons at a university tertiary care center. Indications, techniques, decision analysis, and complications were reviewed. Nonselective venography and selective venography of the inferior vena cava (IVC) were evaluated for image quality, abnormal findings, aberrant anatomy, and the anatomic relationship of vertebral bodies to major venous tributaries.

Results

Absolute and relative indications for VCF placement were 44% and 56%, respectively. Jugular, femoral, and subclavian vein approach was used in 47%, 47%, and 6% of patients, respectively. Seventy-three percent of VCFs were placed in the catheterization laboratory, 21% in the operating room, and 5% at the bedside. Nonselective venography was performed in 80 patients (85%), of whom 44% had undergone selective venography. At nonselective venography plus selective venography 7.5% of patients had an abnormal finding (IVC compression, n = 3; IVC thrombus, n = 2; tortuosity, n = 1). Similarly, 17.5% of patients had aberrant anatomy (accessory renal vein, n = 8; IVC duplication, n = 3; large low right gonadal vein, n = 2; megacava, n = 2). Nonselective venography plus selective venography demonstrated that 16% of VCFs required a major change in position, 10% of which were placed above the renal veins. Compared with nonselective venography alone, selective venography enabled detection of significantly more abnormal and aberrant findings (9% vs 49%; P < .001). Changes in VCF placement were necessary significantly more often in patients undergoing additional selective venography compared with nonselective venography alone (31% vs 4%; P = .003). In one patient in the series, a VCF was malpositioned in the iliac vein with intravascular ultrasound visualization.

Conclusion

When nonselective venography plus selective venography were performed, 23% of patients had either an abnormal finding or aberrant anatomy, and most of these required a major change in VCF position. Nonselective venography plus selective venography redefines the criterion standard and, because of limitations of other methods of vena cava visualization for VCF deployment, should be performed in most patients.

Vena cava filters (VCFs) are frequently deployed during treatment and prophylaxis of venous thromboembolism. Since the initial description by Greenfield et al1 in 1973, VCFs have been routinely used to prevent pulmonary embolism. Absolute indications for VCF most commonly observed in the setting of venous thromboembolism include contraindications to anticoagulation therapy, and pulmonary embolism despite therapeutic anticoagulation. In contrast, indications for prophylaxis against pulmonary embolism in the absence of known venous thromboembolism are more controversial, and include multitrauma, cancer, severe cardiopulmonary disease, hypercoagulable states, and prolonged immobilization.2, 3, 4

Depending on institutional trends and physician training, different methods to optimally visualize the vena cava are used in deployment of VCFs. Not uncommonly, VCFs are placed with fluoroscopic guidance to identify the bony landmarks of the vertebral bodies.5, 6 More recently, technologic advances have led to placement of these devices with visualization with either transabdominal ultrasound7, 8 or intravascular ultrasound (IVUS).9, 10, 11 Proponents of ultrasound cite the advantages of better patient safety with insertion at the bedside, portability, ease of placement with reliable results, and cost containment.7, 8, 9, 10, 11 Potential shortcomings of ultrasound include difficulty in optimal venous visualization, steep learning curve, limitations of access site, and dependence on technologist skill.

Although contrast venography is used most often before VCF placement, little is known about how the discovery of major venous anomalies leads to a decision to change the position of the VCF.12, 13, 14, 15 More important, there is considerable debate as to whether the addition of selective venography after nonselective venography reveals more venous anomalies and thus redefines the criterion standard. We examined our experience with nonselective venography and selective venography for placement of VCFs.

Back to Article Outline

Methods 

A retrospective review was conducted of all consecutive patients in whom a VCF was deployed by six vascular surgeons between January 1998 and June 2002. All vascular surgeons were or are currently full-time faculty in the Division of Vascular Surgery at Southern Illinois University, Springfield, Illinois. The procedures were performed at two tertiary care centers, Memorial Medical Center and St John's Hospital. All patients in whom the VCF was deployed in the inferior vena cava (IVC) were identified with a prospectively maintained vascular registry (Patient Analysis and Tracking System; Axis Software, Portland, Ore) and cross-referenced with the university practice billing database (Signature Billing System; Siemens, Malvern, Pa), to assure that no patients were overlooked. Information obtained from the medical records included patient demographic data, risk factors, indications for VCF insertion, periprocedural data, complications, procedure setting, and changes in serum creatinine concentration in patients who received intravenous contrast medium. Routine follow-up consisted of patient evaluation at 4 to 6 hours after the procedure and the following day. The study was approved by the institutional review board.

VCF placement was performed after administration of local anesthesia and sedation or with the patient under general anesthesia if VCF placement was performed in conjunction with a major operation. Generally the right internal jugular vein was the preferred site for VCF insertion. If this site could not be used because of problems such as need for a cervical collar or previous right internal jugular thrombosis, the right common femoral vein was the preferred insertion site. The method of VCF placement and whether nonselective venography or selective venography was used varied among vascular surgeons. Empiric placement was performed with fluoroscopy to identify the third lumbar vertebral body for deployment. IVC visualization with IVUS was with either a 6F, 12.5 MHz or 9F, 9 MHz IVUS catheter (Boston Scientific, Maple Grove, Minn). Nonselective venography was performed with a 5F pigtail catheter (Cordis, Miami, Fla). Nonselective plus selective venography was performed with digital subtraction angiography, with a 15-inch image intensifier. For nonselective venography the pigtail catheter was positioned at the third lumbar vertebral body, and 25 mL of nonionic contrast agent was injected per second, for a total of 2 seconds, with a power injector. Injection was at 600 psi while the patient held respiration. Selective venography was performed by hand injection of contrast material with a 10 mL syringe attached to a four-way stopcock or manifold. Usually contrast material for hand injection was diluted with normal saline solution to 50% concentration. First-order catheterization for selective venography was in accordance with that described in Current Procedural Terminology (Code 36011).16 Several techniques were used to perform selective venography. These included using curved catheters (eg, 5F multipurpose catheter or 5F Cobra 2 catheter; Cordis) to directly engage unknown IVC side branches, and guide wire–directed catheter placement. Vena cava diameter was measured selectively at the discretion of the attending vascular surgeon. Cavamegaly was defined as diameter 28 mm or greater.

VCF procedure privileges in the catheterization laboratory and whether nonselective venography or selective venography was performed varied among the six vascular surgeons. One vascular surgeon (D.E.R.) placed VCFs only in the operating room, using either lumbar bony landmarks or nonselective venography. The remaining five vascular surgeons (R.B.M., L.A.G., M.M.S., M.A.M., K.J.H.) also had VCF procedure privileges in the catheterization laboratory. For these five vascular surgeons, use of selective venography in addition to nonselective venography varied. Generally, one of the vascular surgeons (K.J.H.) routinely performed selective venography of the renal veins, iliac veins, or other venous anomalies, and the other four (R.B.M., L.A.G., M.M.S., M.A.M.) performed selective venography, depending on the quality of visualization of venous tributaries at nonselective venography.

Venograms were reviewed in conjunction with the information provided by the procedural report by an attending vascular surgeon (R.B.M.) and a vascular fellow (J.S.D.). Nonselective venograms were graded with the criteria outlined in Table I. In brief, the quality of nonselective venograms was based on whether the renal and iliac veins could be reliably identified for safe VCF placement. In addition, the lowest renal vein and common iliac vein confluence was aligned with corresponding vertebral bodies and disk spaces.

Table I. Criteria for grading quality of nonselective venograms

Venograms were also evaluated for abnormal findings and aberrant anatomy. Abnormal findings included compression, significant tortuosity, and presence of thrombus. Aberrant anatomic findings included accessory renal veins greater than one third the diameter of each renal vein proper, IVC duplication, low right gonadal vein, and cavamegaly. Procedure reports and venograms were reviewed to determine whether abnormal and aberrant findings led to a decision for a major change in VCF position. A major change in VCF position was defined as placing the VCF in the IVC above the renal veins or inability to place the VCF just inferior (within 2 cm) to the lowest proper renal vein. Deploying the VCF in the lower aspect of the IVC, just above the iliac bifurcation, was considered a major change in position if the anatomy just below the renal veins was compromised due to an aberrant anatomic or abnormal finding.

A χ² test was used to compare abnormal and aberrant venous findings between patients undergoing nonselective and selective venography, and to determine whether these findings led to a change in VCF position. An unpaired t test was used to compare contrast volume between nonselective and selective groups, and preprocedural and postprocedural creatinine concentrations. P < .05 was accepted as statistically significant. Data are expressed as mean ± SD.

Back to Article Outline

Results 

Ninety-five VCFs were deployed in 94 patients (53% male; mean age, 53 years [range, 15-90 years]). Risk factors for venous thromboembolism are shown in Table II. Before VCF insertion, 57 patients (61%) underwent venous duplex ultrasound scanning. Thirty-seven patients (39%) had deep venous thrombosis, and 29 patients (31%) had pulmonary embolism. Forty patients (44%) had an absolute indication for VCF placement, and 54 (56%) had a relative indication for VCF placement (Table III). The VCF was deployed in the catheterization laboratory in 69 patients (73.4%), the operating room in 20 patients (21.3%), and the intensive care unit in 4 patients (4.3%). In one patient (1.1%) a VCF was deployed in the left common iliac vein with IVUS in the intensive care unit. The next day a second VCF was correctly placed in the IVC before a major orthopedic operation. Catheter access to the vena cava included the right internal jugular vein in 44 patients (46.8%), right common femoral vein in 41 patients (43.6%), right subclavian vein in 6 patients (6.4%), and left common femoral vein in 2 patients (2.1%). Access for the one patient (1.1%) in whom two VCFs were required were the left common femoral vein and right internal jugular vein.

Table II. Risk factors for thromboembolic events in patients with vena cava filter

Table III. Absolute and relative indications in patients with vena cava filter

VTE, Venous thromboembolism; PE, pulmonary embolism; DVT, deep venous thrombosis.

VCFs were deployed empirically over the third lumbar vertebrae in 2 patients (2.1%), and IVUS was used exclusively before deployment in 10 patients (10.6%). Three patients (3.2%) underwent both nonselective venography and IVUS before VCF deployment. No aberrant or abnormal findings were noted with either imaging method. Venography was performed in 80 patients (85%), nonselective venography alone in 45 patients(56%) and additional selective venography of major venous tributaries in 35 patients (44%). Frequency of selective venography used in addition to nonselective venography by each vascular surgeon is shown in Table IV.

Table IV. Analysis by vascular surgeon of number of vena cava filter deployments with selective venography and where the procedure was performed

Selective catheterization was of both renal veins in 26 patients (32.5%), unilateral or accessory renal veins in 4 patients (5.0%), both common iliac veins in 7 patients (8.8%), and unilateral common iliac vein in 10 patients (12.5%). In 15 patients (18.8%) renal and iliac veins were selectively catheterized. With the criteria defined in Table I, 16% of nonselective venograms were graded as excellent, 16% as good, 45.3% as marginal, 6.7% as suboptimal, and 16% as poor.

Six patients (7.5%) had abnormal findings at venography. These included IVC or iliac vein thrombus in 2 patients (2.5%), significant IVC tortuosity in 1 patient (1.3%), and IVC compression in 3 patients (3.8%). Fourteen patients (17.5%) had aberrant anatomic findings. These included accessory renal vein in 8 patients (10%), left-sided vena cava or IVC duplication in 3 patients (3.8%), large low-lying right gonadal vein in 2 patients (2.5%), and cavamegaly in 2 patients (2.5.%). Three patients (3.8%) had two aberrant or abnormal findings. These included cavamegaly and accessory renal vein, IVC tortuosity and cavamegaly, and IVC thrombus and low-lying right gonadal vein in one patient each. Of the 80 patients undergoing venography before VCF placement, 23% had abnormal or aberrant anatomic findings. Abnormal or aberrant anatomic findings were found significantly more often (P < .001) at selective venography (17 of 35, 49%) compared with nonselective venography (4 of 45, 9%).

Seven aberrant or abnormal findings required a change in VCF position below the renal veins. These included five accessory renal veins, one IVC compression, and one large right gonadal vein. Six aberrant or abnormal findings required VCF placement above the renal veins. These included two instances of cavamegaly, two IVC duplications, one IVC compression, and one large right gonadal vein. Eight aberrant or abnormal findings required no change in VCF position. These included three accessory renal veins, two IVC thrombi, one left-sided IVC, one IVC compression, and one tortuous IVC.

When examining the relationship of major venous tributaries visualized on venograms to vertebral bony landmarks, the lowest renal vein corresponded to the L1 vertebral body in 17.3%, L1-2 disk space in 36.5%, L2 vertebral body in 34.6%, L2-3 disk space in 7.7%, and L3 vertebral body in 3.8%. Similarly, the common iliac vein confluence corresponded to the L5-S1 disk space in 3.8%, L5 vertebral body in 80.8%, L4-5 disk space in 11.5%, and L4 vertebral body in 3.8%. The relationship of venous tributaries to the vertebral column is shown in Fig 1.

• 
  • View Large Image
  • Download to PowerPoint

• Fig 1. 

  Diagram illustrates relationship of lowest renal veins and common iliac vein confluences to lumbar vertebral bodies and disk spaces.

Of the entire cohort, the stainless steel Greenfield filter (Meditech, Watertown, MA) was deployed in 21 patients (22.3%), the titanium Greenfield filter (Meditech) in 37 patients (39.4%), and the TrapEase filter (Cordis) in 35 patients (37.2%). One patient (1.1%) received both a TrapEase filter and a titanium Greenfield filter. The TrapEase filter was used more frequently toward the latter half of the series, because of its smaller delivery profile (6F) and bidirectional deployment capability. The decision to deploy a VCF in a position other than just inferior to the lowest proper renal vein was made significantly more often (P = .003) in patients undergoing selective venography (11 of 35, 31.4%) compared with those undergoing nonselective venography (2 of 45, 4.4%).

A mean of 79.9 ± 62.5 mL (range, 30-380 mL) of intravenous contrast medium was used during venography. Patients undergoing additional selective venography received significantly more intravenous contrast medium (P < .001) than those undergoing only nonselective venography (111.7 ± 73.9 mL vs 51.1 ± 28.9 mL, respectively). In 56 of 80 patients (70%) undergoing venography serum creatinine concentration was measured before VCF and at 4 to 7 days of follow-up. Two patients undergoing hemodialysis because of chronic renal failure were excluded from the analysis. Follow-up serum creatinine concentration in patients undergoing only nonselective venography (0.78 ± 0.26 mg/dL) was significantly lower (P = .02) compared with serum creatinine concentration before VCF (0.92 ± 0.35 mg/dL). For patients undergoing additional selective venography (n = 21), serum creatinine concentration before and after VCF placement (0.84 ± 0.20 mg/dL vs 0.77 ± 0.17 mg/dL, respectively) did not significantly change (P = .22). As a whole, serum creatinine concentration did not increase by more than 1 mg/dL in any patients, nor did any patient require hemodialysis as a result of contrast agent–induced nephropathy.

Seven patients (8.8%) had complications due to VCF deployment. Three patients had small hematomas at the access site, none of which required surgery or transfusion. One patient had a pneumothorax, which did not require tube thorocostomy, and subsequently resolved. In one patient who underwent concomitant aortography, a common femoral artery pseudoaneurysm developed, which was successfully managed with duplex ultrasound compression. In one patient, initial placement of the access sheath was in the common femoral artery, without untoward effect. In one patient, a VCF was wrongly placed in the left common iliac vein with IVUS.

Five patients (5.3%) died within 30 days of VCF placement. Although no autopsies were performed, none of the deaths was attributed to complications or failure of VCF. Causes of death included brain death, multiple system organ failure, and advanced malignancy. At mean follow-up after VCF placement of 19 months (range, 3-56 months), recurrent venous thromboembolism occurred in four patients (4.3%). Three patients had deep venous thrombosis, and one patient had a pulmonary embolism. This patient had extensive upper extremity and tricuspid valve thrombus, and subsequently required pulmonary embolectomy.

Back to Article Outline

Discussion 

The use of VCFs has dramatically reduced the occurrence of pulmonary embolism in patients at high risk.17 Nevertheless, recurrent pulmonary embolism after VCF placement is as high as 4.4%.18, 19, 20 Considering the variation of venous anatomy and the potential for vena cava disease, proper position of a VCF is paramount in prevention of pulmonary embolism. Obtaining correct VCF position is, in this modern era, far more dependent on adequate vena cava visualization than on device design or physician training. Our results demonstrate two important and related findings. In using additional selective catheterization to obtain venograms of major venous tributaries to the IVC, significantly more aberrant anatomic and abnormal venous findings were discovered. More important, most of these discoveries led to a decision to place the VCF in an alternate position. Twenty-three percent of our patients undergoing venography had aberrant anatomy or abnormal venous findings. Anomalies were discovered in 49% of patients undergoing selective venography, compared with 9% of patients undergoing only nonselective venography. Moreover, 31.4% of patients with selective venography had a change in VCF position, compared with 4.4% of patients with only nonselective venography.

These findings may have important implications when considering the use of other methods of vena cava localization or visualization for proper positioning of a VCF. Not uncommonly, empiric placement of a VCF in the IVC at the approximate level of the L3 vertebral body with fluoroscopy is standard practice. Given the results of this study, in which the lowest renal vein visualized at venography was aligned with precise vertebral levels, caution should be used in continuing this method of VCF deployment. More than 1 in 10 patients undergoing venography in our study had a renal vein between the top of the L2-3 disk space and the bottom of the L3 vertebral body. These findings lead us to recommend abandoning the technique of empiric placement with bony landmarks, despite safe placement of a small number of VCFs (n = 2) in our patients with this method.

More recently, transabdominal duplex ultrasound7, 8 or IVUS9, 10, 11 is being used to visualize the vena cava for VCF placement. The advantages of these methods include ability to perform the procedure at the bedside, and cost-reduction.7, 8, 9, 10, 11 From a practical standpoint, the preciseness of vena cava visualization is highly dependent on the technologist's skill and the interpreter's abilities. Thus the learning curve is steep with these methods, as exemplified by a recent study that reported VCF maldeployment in 6 patients.8 In contrast, no VCFs were maldeployed with venography in our patient cohort, while we experienced the steep learning curve of using ultrasound as exemplified by one VCF incorrectly placed in the common iliac vein. The discovery from this report that 23% of patients had either aberrant venous anatomy or an abnormal finding provides useful information for those centers that use ultrasound as the primary method of vena cava visualization. Although we agree that use of ultrasound at the bedside may be necessary in a small number of clinically unstable patients in the intensive care unit, most patients can be safely transported to the catheterization laboratory, as demonstrated with our patient cohort.

Relying completely on nonselective venography for optimal VCF position may also have limitations when desiring optimal VCF position. Even with the use of a power injector, optimal contrast load, and a 15-inch image intensifier, we observed 68% of nonselective venography to be marginal, suboptimal, or poor (Table I). It is interesting that of the 24 nonselective venograms graded of excellent or good quality, only 1 showed a large accessory renal vein that led to placing the VCF in a much lower position. Others have had similar findings. Hicks et al12 discovered that an additional 18% of patients had significant venous anomalies at subsequent selective renal venography. Similar to our findings, they noted that findings at venography led to change in VCF position in 30% of patients. Mejia et al14 and Martin et al,13 using only nonselective venography, found change in VCF position necessary in 11% and 15% of patients, respectively. Of patients undergoing only nonselective venography in our study (n = 45), 4 had aberrant anatomic or abnormal venous findings. Two of these patients required a change in VCF position. Of interest, it could be postulated that in the 41 patients in whom aberrant anatomic or abnormal venous findings were not detected, anomalies were missed in an additional 15 patients because selective catheterization was not used. Fig 2 shows how a major aberrant anatomic finding can be missed when only nonselective venography is used.

• 
  • View Large Image
  • Download to PowerPoint

• Fig 2. 

  Digital subtraction angiograms show nonselective venogram of inferior vena cava (A) and subsequent selective venogram of left common iliac vein (B). Duplicated inferior vena cava (arrow) leading to left renal vein is shown only with selective venography. Vena cava filter was deployed in inferior vena cava above renal veins.

The retrospective nature of patient follow-up did not provide reliable information as to whether the additional use of selective venography for VCF placement made a difference in prevention of pulmonary embolism during long-term follow-up. It should be emphasized that the hypothesis of this study was not to address this important issue, but to determine whether selective catheterization led to a change in deployment of the VCF away from the usual position just below the renal veins. From a monetary perspective, the relatively small difference in hospital charges to patients did not provide a compelling argument for not using selective venography. Sample data from four patients showed an approximate charge of $6500 for VCF placement in the operating room (empiric placement without venography), $7300 with IVUS at the bedside, $5900 with nonselective venography, and $7700 with selective venography.

Selective catheterization in our patient cohort was performed at the discretion of the attending vascular surgeon. Lack of a strict protocol as to when selective venography was to be used and the retrospective nature of the study represent significant bias. Whereas one attending vascular surgeon used selective venography almost routinely to further delineate the major venous tributaries, others used it based on the initial findings of nonselective venography (Table IV). Nevertheless, a significantly higher proportion of patients were discovered to have aberrant anatomic or abnormal venous findings when selective venography was used. This provides more evidence for physicians, who routinely use venography to place VCFs, to have a low threshold to selectively catheterize the renal, iliac, and other large veins draining into the vena cava. Alternatively stated, if nonselective venography does not enable easy identification of both renal veins and most of the left iliac vein (primarily for accessory vena cava identification), selective venography should be used to define these and any other abnormality to optimize VCF position.

Back to Article Outline

Acknowledgements 

We acknowledge the contributions of two previous faculty members of Southern Illinois University. Those cited in Methods and Table IV, not previous listed as authors, include Mark A. Mattos, MD (M.A.M.) and Maurice S. Solis, MD (M.S.S.).

Back to Article Outline

References 

1. Greenfield LJ, McCurdy JR, Brown PP, Elkins RC. A new intracaval filter permitting continued flow and resolution of emboli. Surgery. 1973;73:599–606
   • View In Article
   • MEDLINE
2. Quirke TE, Ritota PC, Swan KG. Inferior vena caval filter use in US trauma centers (a practitioner's survey). J Trauma. 1997;43:333–337
   • View In Article
   • MEDLINE
3. American College of Chest Physicians. Consensus Committee on Pulmonary Embolism . Opinions regarding the diagnosis and management of venous thromboembolic disease. Chest. 1998;113:499–504
   • View In Article
   • MEDLINE
   • CrossRef
4. Sharafuddin MJ, Corson JD. Percutaneous devices for vena cava filtration. In:  Rutherford RB editors. Vascular surgery. Philadelphia, Pa: Saunders; 2000;
   • View In Article
5. Greenfield LJ. Technical considerations for insertion of vena caval filters. Surg Gynecol Obstet. 1979;148:422–426
   • View In Article
   • MEDLINE
6. Otchy DP, Elliott BM. The malpositioned Greenfield filter (lessons learned). Am Surg. 1987;53:580–583
   • View In Article
   • MEDLINE
7. Neuzil DF, Garrard CL, Berkman RA, Pierce R, Nasland TC. Duplex-directed vena caval filter placement (report of initial experience). Surgery. 1998;123:470–474
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (109 KB)
   • CrossRef
8. Conners MS, Becker S, Guzman RJ, Passman MA, Pierce R, Kelly T, et al.  Duplex scan-directed placement of inferior vena cava filters (a five-year institutional experience). J Vasc Surg. 2002;35:286–291
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (91 KB)
   • CrossRef
9. Ebaugh JL, Chiou AC, Morasch MD, Matsumura JS, Pearch WH. Bedside vena cava filter placement guided with intravascular ultrasound. J Vasc Surg. 2001;34:21–26
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (78 KB)
   • CrossRef
10. Matsuura JH, White RA, Kopchok G, Nishinian G, Woody JD, Rosenthal D, et al.  Vena caval filter placement by intravascular ultrasound. Cardiovasc Surg. 2001;9:571–574
    • View In Article
    • MEDLINE
    • CrossRef
11. Bonn J, Kiu JB, Eschelman DJ, Sullivan KL, Pinheiro LW, Gardiner GA. Intravascular ultrasound as an alternative to positive-contrast vena cavography prior to filter placement. J Vasc Interv Radiol. 1999;10:843–849
    • View In Article
    • Abstract
    • Full-Text PDF (2285 KB)
    • CrossRef
12. Hicks ME, Maiden ES, Vesely TM, Picus D, Darcy MD. Prospective anatomic study of the inferior vena cava and renal veins (comparison of selective renal venography with cavography and relevance in filter placement). J Vasc Interv Radiol. 1995;6:721–729
    • View In Article
    • Abstract
    • Full-Text PDF (6352 KB)
    • CrossRef
13. Martin KD, Kempczinski RF, Fowl RJ. Are routine vena cavograms necessary before Greenfield filter placement?. Surgery. 1989;106:647–651
    • View In Article
    • MEDLINE
14. Mejia EA, Saroyan RM, Balkin PW, Kerstein MD. Analysis of inferior venacavography before Greenfield filter placement. Ann Vasc Surg. 1989;3:232–235
    • View In Article
    • Abstract
    • Full-Text PDF (740 KB)
    • CrossRef
15. Brown DB, Labuski MR, Cardella JF, Singh H, Waybill PN. Determination of inferior vena cava diameter in the angiography suite (comparison of three common methods). J Vasc Interv Radiol. 1999;10:143–147
    • View In Article
    • Abstract
    • Full-Text PDF (1977 KB)
    • CrossRef
16. Current Procedural Terminology, CPT 2003. Chicago, Ill: AMA Press; 2003. p 126
    • View In Article
17. Carlin AM, Tyburski JG, Wilson RF, Steffes C. Prophylactic and therapeutic inferior vena cava filters to prevent pulmonary emboli in trauma patient. Arch Surg. 2002;137:521–525
    • View In Article
    • MEDLINE
    • CrossRef
18. Greenfield LJ, Proctor MC, Cho KJ, Cutler BS, Ferris EJ, McFarland D, et al.  Extended evaluation of the titanium Greenfield vena caval filter. J Vasc Surg. 1994;20:458–465
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (1431 KB)
    • CrossRef
19. Mohan CR, Hoballah JJ, Sharp WJ, Kresowik TF, Lu C, Corson JD. Comparitive efficacy and complications of vena caval filters. J Vasc Surg. 1995;21:235–246
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (2611 KB)
    • CrossRef
20. Greenfield LJ, Zocco J, Wilk J, Schroeder TM, Elkins RC. Clinical experience with the Kim-Ray Greenfield vena caval filter. Ann Surg. 1977;185:692–698
    • View In Article
    • MEDLINE
    • CrossRef