- Email: [email protected]
Prospective implementation of an algorithm for bedside intravascular ultrasound-guided filter placement in critically ill patients
From the American Venous Forum
Prospective implementation of an algorithm for bedside intravascular ultrasound-guided filter placement in critically ill patients Christopher D. Killingsworth, MD,a Steven M. Taylor, MD,b Mark A. Patterson, MD,b Jordan A. Weinberg, MD,c Gerald McGwin Jr, MS, PhD,c Sherry M. Melton, MD,c Donald A. Reiff, MD,c Jeffrey D. Kerby, MD,c Loring W. Rue, MD,c William D. Jordan Jr, MD,b and Marc A. Passman, MD,a Birmingham, Ala Background: Although contrast venography is the standard imaging method for inferior vena cava (IVC) filter insertion, intravascular ultrasound (IVUS) imaging is a safe and effective option that allows for bedside filter placement and is especially advantageous for immobilized critically ill patients by limiting resource use, risk of transportation, and cost. This study reviewed the effectiveness of a prospectively implemented algorithm for IVUS-guided IVC filter placement in this high-risk population. Methods: Current evidence-based guidelines were used to create a clinical decision algorithm for IVUS-guided IVC filter placement in critically ill patients. After a defined lead-in phase to allow dissemination of techniques, the algorithm was prospectively implemented on January 1, 2008. Data were collected for 1 year using accepted reporting standards and a quality assurance review performed based on intent-to-treat at 6, 12, and 18 months. Results: As defined in the prospectively implemented algorithm, 109 patients met criteria for IVUS-directed bedside IVC filter placement. Technical feasibility was 98.1%. Only 2 patients had inadequate IVUS visualization for bedside filter placement and required subsequent placement in the endovascular suite. Technical success, defined as proper deployment in an infrarenal position, was achieved in 104 of the remaining 107 patients (97.2%). The filter was permanent in 21 (19.6%) and retrievable in 86 (80.3%). The single-puncture technique was used in 101 (94.4%), with additional dual access required in 6 (5.6%). Periprocedural complications were rare but included malpositioning requiring retrieval and repositioning in three patients, filter tilt >15° in two, and arteriovenous fistula in one. The 30-day mortality rate for the bedside group was 5.5%, with no filter-related deaths. Conclusions: Successful placement of IVC filters using IVUS-guided imaging at the bedside in critically ill patients can be established through an evidence-based prospectively implemented algorithm, thereby limiting the need for transport in this high-risk population. ( J Vasc Surg 2010;51:1215-21.)
The effectiveness of inferior vena cava (IVC) filters as a treatment for the prevention of fatal venous thromboembolism (VTE) is well documented.1-4 Although original transvenous filters were placed in the operating room through open venous access, advances in technology and low-profile delivery systems allowed the evolution of percutaneous placement in the endovascular suite to become the standard technique. Venography has the advantage of providing accurate assessment of venous anatomic landmarks and real-time positioning of the filter under fluoroscopic guidance and remains the most common approach From the Division of General Surgery,a Sections Vascular Surgery and Endovascular Therapy,b and Trauma, Burns and Surgical Critical Care,c University of Alabama at Birmingham.University of Alabama at Birmingham. Competition of interest: none. Presented at the Twenty-first Annual Meeting of the American Venous Forum, Phoenix, Ariz, Feb 10-14, 2009. Correspondence: Marc A. Passman, MD, Section of Vascular Surgery and Endovascular Therapy, University of Alabama at Birmingham, BDB 503, 1530 3rd Ave South, Birmingham, AL 35294-0012 (e-mail: marc. [email protected]). The editors and reviewers of this article have no relevant financial relationships to disclose per the JVS policy that requires reviewers to decline review of any manuscript for which they may have a competition of interest. 0741-5214/$36.00 Copyright © 2010 by the Society for Vascular Surgery. doi:10.1016/j.jvs.2009.12.041
for the placement of IVC filters at most institutions. However, transporting critically ill and immobilized patients to the endovascular suite for filter placement is not without complication and requires extra support staff, cost, and time. Conventional filter placement also bears the additional risk of radiation and contrast exposure.5,6 In recent years, bedside placement of IVC filters guided by transabdominal duplex ultrasound (DUS) imaging7-10 or intravascular ultrasound (IVUS) imaging11-19 has been shown to be safe, effective, and reliable. It has the advantage of avoiding transportation of critically ill patients and has become a preferred approach for IVC filter placement in some high-volume centers.20,21 Yet, wider dissemination of bedside techniques elsewhere has been limited by concerns for adequate imaging, the potential of missed venous anomalies, learning curve issues, and comfort level using bedside techniques, concerns that have previously limited introduction of these bedside techniques at our institution. This study reviewed the effectiveness of a new prospectively implemented algorithm for IVUS-guided filter placement in critically ill patients. METHODS Filter algorithm. A clinical decision algorithm for IVUS-guided IVC filter placement in critically ill patients 1215
JOURNAL OF VASCULAR SURGERY May 2010
1216 Killingsworth et al
Fig. Prospectively implemented decision algorithm for intravascular ultrasound– guided bedside filter placement in critically ill patients. DVT, Deep venous thrombosis; ESRD, end-stage renal disease; GCS, Glasgow coma scale; ICU, intensive care unit; IVC, inferior vena cava; MR, magnetic resonance; PE, pulmonary embolism; PUD, peptic ulcer disease; VTE, venous thromboembolism.
was created from current evidence-based guidelines (Fig).22-27 ●
●
Decision point 1. Indications for placement included standard indications, consisting of documented deep venous thrombosis (DVT) or pulmonary embolism (PE) with contraindication or complication of anticoagulation, and prophylactic indications consisting of high risk for VTE and an increased bleeding risk precluding anticoagulation-based prophylaxis. Decision point 2. Choice of filter type was according to predetermination of permanent or optional needs using the following criteria: A permanent filter (Stainless Steel Greenfield Filter. Boston Scientific, Natick, Mass) was used when the anticipated risk of VTE extended beyond the acceptable retrieval window because of injury type or severity. An optional filter (Günther-Tulip Filter, Cook Inc, Bloomington, Ind) was used if there was a defined retrievable end point within 2 to 6 months, or if the end point was uncertain
●
but permanent need was possible. Patients with retrievable filters were evaluated for removal at 2 to 6 months after placement, and a decision to proceed with removal was determined by current medical status applied to accepted indications for retrieval.28 Decision point 3. Preprocedural lower extremity venous DUS imaging was obtained to evaluate the presence of lower extremity DVT. If available, pre-existing computed tomography scans were reviewed for vena cava and renal vein anatomic detail, thrombosis, or anomalies. The decision to place the filter at the bedside was based on critical illness in the intensive care unit setting, injury type and severity, and potential risk of transportation. Placement of the filter in the endovascular suite was reserved for noncritically ill patients for whom transportation risk was low, if femoral access was not possible (bilateral lower extremity DVT), or the presence of IVC thrombosis or vena caval anomalies precluded safe placement. Additional venographic guid-
JOURNAL OF VASCULAR SURGERY Volume 51, Number 5
ance was also required for failure of bedside techniques due to inadequate imaging or access, or if a bedside placed filter was malpositioned. Technique for IVUS-guided filter placement. The IVUS system used was the Volcano s5 System, with a Visions PV8.2F (8F, 8.3-MHz) IVUS catheter (Rancho Cordova, Calif). After sterile preparation and adequate local anesthesia, percutaneous femoral venous access was obtained by using anatomic landmarks, and an 8F, 10-cm sheath was placed for IVUS imaging. Access was preferably from the right side if there was no evidence of DVT at the access site, thereby providing a straighter course for filterdelivery catheters. If dual access was needed or DVT was noted in the right femoral position, then left-sided access was obtained. A 0.035-inch guidewire was directed into the vena cava, followed by the IVUS probe, which was advanced to the level of the right atrium of the heart. A pullback technique was used to sequentially identify the venous anatomic landmarks, including the right atrium, hepatic veins, renal veins, and the confluence of the iliac veins.29 The IVUS probe was directed just below the level of the lowest most renal vein, and IVC diameter measurements were confirmed to be ⬍28 mm before proceeding with filter delivery. With appropriate IVUS imaging of the vena cava confirmed, the technical decision to proceed with bedside placement of filter was based on preference for single-puncture technique first, reserving dual-access options if additional directed imaging was needed. For the single-access technique, the IVUS probe was used to precisely direct the end of the sheath to the level of the lowest most renal vein. Selected filters used in this study were based on need for 8F access for the IVUS probe to allow single-access technique through at least the samesized sheath or greater required for filter delivery. For the Günther-Tulip filter with predetermined marks, the filter-delivery catheter was advanced to the mark aligning the tip of the filter with the end of the 8F sheath, and then the sheath was withdrawn in a “pin-pull” fashion to allow deployment of the filter at the infrarenal level.30 For Greenfield filter without predetermined marks, the 8F sheath was up-sized to the 15F introducer sheath over a wire, and the IVUS probe was reloaded to direct the end of the sheath to the level of the renal veins. The 15F sheath was then pulled back over the IVUS probe a distance equivalent to the length that the filter-delivery catheter extended beyond the sheath (approximately 7 cm), so when the filter-delivery catheter was loaded into the sheath, the tip of the filter would precisely align with the lowest most renal vein upon deployment.31 For dual-access needs, if the iliac confluence required better localization, contralateral access was used as an adjunct to the single-puncture technique with placement of a 5F sheath, followed by passage of a 0.035-inch guidewire. With IVUS imaging of the vena cava through the initial ipsilateral access, the point at which the contralateral guidewire was visualized corresponded to the iliac confluence. An extended dual-access technique was used if imaging
Killingsworth et al 1217
alone with the single puncture was inadequate or if direct visualization of the filter-delivery catheter tip at the renal vein was required, or both. With the IVUS probe left at the renal vein level through the ipsilateral access, the corresponding sheath and filter-delivery catheter were directed to the level of the renal veins under direct vision through the contralateral access. Once the position of the filterdelivery catheter tip was confirmed at the renal veins, the IVUS probe was pulled back and the filter deployed. For completion imaging, the IVUS probe was carefully advanced to confirm apposition of the filter legs to the IVC wall and, in the absence of resistance, was directed through the filter legs to confirm position of the tip at the renal vein level. After removal of the filter-delivery catheter and sheath(s), gentle manual pressure over the puncture site(s) was applied for hemostasis. Postprocedural plain abdominal radiographs were obtained to verify filter position and alignment when possible relative to lumbar vertebral anatomy. Data analysis. There was a defined lead-in phase before the protocol was initiated to allow modification of the algorithm and dissemination of techniques among the vascular surgeons responsible for filter placement (M.A.P., S.M.T., M.A.P., W.D.J.). The algorithm was prospectively implemented on January 1, 2008, with a 1-year enrollment phase. Accepted reporting standards were used to collect the data,32 and a quality assurance data review was performed based on intent-to-treat at 6, 12, and 18 months. Technical feasibility was defined as intent to place the filter with IVUS guidance and determined by ability to obtain appropriate access and adequate visualization. Technical success for placement was defined as a properly aligned filter in the infrarenal position with the tip at or near the renal vein. Additional 30-day end points included procedural-related complications, filter-related problems, and death. Filter retrievability end points based on the data review intervals were also followed up at a minimum of 6 months after filter placement. Mean data were compared by t test, and values of P ⬍ .05 were considered significant. RESULTS Between January 1 and December 31, 2008, 109 patients met the criteria for IVUS-directed bedside IVC filter placement based on the prospectively implemented decision algorithm and constitute the study population. Of the 109 patients, 26 (24%) were women and 83 (76%) were men, with a mean age of 45 ⫾ 18.4 (SD) years (range, 16-79 years). Indications for filter placement included known DVT and contraindication to anticoagulation in 34 (31.2%), hemorrhagic complications from anticoagulation in setting of known DVT in 5 (4.6%), failure to prevent pulmonary thromboembolism despite proper anticoagulation in 3 (2.8%), and poor compliance with anticoagulation with known DVT in 1 (0.9%). A prophylactic filter was indicated in 66 patients (60.5%) due to multisystem trauma that increased the risk for DVT with concurrent contraindications to anticoagulation such as intracranial hemor-
JOURNAL OF VASCULAR SURGERY May 2010
1218 Killingsworth et al
Table. Demographic, risk factors, and indications for bedside vena cava filter placement using the prospectively implemented algorithm Patient characteristics Demographic data Male/female ratio Age, years Median Mean ⫾ standard deviation Range Risk factor characteristics, No. (%) Immobilization (exclude paraplegia) Paraplegia from spinal cord injury Limb trauma Closed head injury Spinal fracture Indications for filter placement, No. (%) Prophylactic placement with contraindication to anticoagulation DVT with contraindication to anticoagulation Hemorrhagic complications from anticoagulation with known DVT Failure of anticoagulation to prevent pulmonary embolism Poor compliance with anticoagulation and known DVT
N ⫽ 109 3.2:1 46 45 ⫾ 18.4 16-79 77 (70.6) 27 (24.8) 80 (73.4) 56 (51.4) 52 (47.7)
no instances of IVC occlusion, venous ulceration, or filter migration. Of the 86 patients with retrievable filters, 14 died at some interval after placement. This left 72 patients with potentially retrievable filters, and 32 (44.4%) returned for follow-up evaluation. The decision was made to attempt removal of the filter in 20 patients (62.5%) and leave the filter in 12 (37.5%). Of the 20 attempted filter retrievals, 17 (85%) were successful. The average time from placement to retrieval was 70.6 ⫾ 43 days (range 15-180 days) in the 17 successful removals and 95 ⫾ 17.6 days (range 82-115 days) in the 3 unsuccessful removals. These differences did not reach statistical significance on comparison of means (t test, P ⫽ .12). In the three unsuccessful attempts, filter incorporation into the IVC prevented retrieval. DISCUSSION
66 (60.5) 34 (31.2) 5 (4.6) 3 (2.8) 1 (0.9)
DVT, Deep venous thrombosis.
rhage or in patients with spinal cord injury with extended immobilization (Table). By intent-to-treat, bedside filter placement with IVUS was technically feasible in 107 of these 109 patients (98.1%). Bedside placement was not possible in two patients due to inadequate visualization on IVUS imaging. Of the 107 placed at bedside, technical success was 97.2%. Three retrievable filters were malpositioned— two in relation to the renal veins and required retrieval/ replacement, and one in the iliac vein that could not be retrieved and required an additional filter. A single-puncture venous access was used in 101 of the 107 placements (94.4%) and dual access was required in 6 (5.6%). A Günther-Tulip retrievable filter was used 86 patients (80.3%) and a permanent Stainless Steel Greenfield filter in the remaining 21 (19.6%). Complications occurred in 6 of 109 patients (5.5%), including three malpositioned filters as detailed above, one common femoral arteriovenous fistula that spontaneously resolved on ultrasound imaging at 3 months, and two filters with minor axial tilt. No instances of IVC occlusion or access site thrombosis were documented from clinical symptoms or ultrasound imaging ⱕ30 days. There were no procedurally related deaths, but six unrelated deaths (5.5%) occurred ⱕ30 days as a result of sepsis or complications from intracranial hemorrhage. Follow-up data at the 18-month interval showed no long-term filter complications. Only 4 of the 34 patients who returned for clinical follow up ⬎4 months were found to have postphlebitic chronic mild ankle or calf edema, which was treated with compression stockings. There were
Although evidence-based guidelines for treatment of documented VTE support therapeutic anticoagulation, anticoagulation may be contraindicated in the setting of critical illness, traumatic injury, or other situations associated with increased bleeding risk. Given the increased prevalence of VTE, associated risk of fatal PE, and difficulties with anticoagulation in the critically ill, vena caval interruption with filter devices is an important consideration. The positive effect of IVC insertion on survival remains unproven, but there is sufficient nonrandomized evidence proving effective protection from PE. The only randomized study in which patients with documented extremity DVT underwent permanent IVC filter placement showed a lower incidence of PE ⱕ12 days.1 A long-term follow-up of 8 years showed significant reduction in PE compared with controls (6.2% vs 15.1%), but an increased rate of recurrent DVT and no difference in survival.2 The American College of Chest Physicians (ACCP) current guidelines (ACCP 2008) strongly recommend placement of an IVC filter in patients with documented VTE and a contraindication to anticoagulation, complication from anticoagulation, or VTE despite therapeutic anticoagulation.26 Use of IVC filters for VTE prophylaxis is still debated, however. Conventional measures of thromboprophylaxis with anticoagulation or sequential mechanical compression devices, even when used properly, may be relatively ineffective in the setting of critical illness and multisystem trauma.33 Although ACCP 2004 suggested that evidence supporting the use of IVC filters for prophylaxis was equivocal,24 and the updated ACCP 2008 recommends against IVC filters for VTE prophylaxis in trauma patients,27 other evidence-based guidelines, including those from the 2002 Eastern Association for the Surgery of Trauma (EAST)22 suggest a role for filters in selected high-risk patients, and most series report a decreased incidence of all PE and fatal PE with the use of IVC filters in appropriately selected high-risk patients.18,19,34,35 Given the frequency of VTE and the complexity of multiple decision points, it would follow that an evidencebased algorithmic approach for management is necessary. The validity of these and other published evidence-based
JOURNAL OF VASCULAR SURGERY Volume 51, Number 5
guidelines was not tested in this study, but all decision points used in the prospective algorithm were derived from available evidence in the literature. Published recommendations were integrated into the prospective algorithm for filter placement in critically ill patients not only to guide clinical decision making for filter placement (decision point 1) but also to direct filter selection type (decision point 2) and bedside filter placement options (decision point 3). During a 1-year enrollment period, 109 critically ill patients met the criteria for IVUS-directed bedside filter placement based on the prospectively implemented decision algorithm and constituted the study population. Patient demographics, critical illness, and injury pattern in this study were also similar to other recently published reports mainly consisting of multisystem trauma patients at high risk for VTE.31 Regarding decision point 1, the assumption in this study in using current evidence-based guidelines is that the decision to place a filter in a critically ill patient depends on the strength of the data composing these recommendations. Although the recommendations for use of filters in patients with documented VTE who cannot receive anticoagulation is strong (39.5% of this study population), use for VTE prophylaxis remains equivocal (60.5% of this study population). Furthermore, although this prospective algorithm was based on ACCP 2004,23,24 the recommendations from ACCP 2008,26,27 which were published midway through study recruitment, were not significantly different for patients with documented VTE and contraindications to anticoagulation favoring IVC filter placement. The strength of the recommendations for VTE prophylaxis against filters was different between ACCP 200424 and ACCP 2008,27 but given the complexity of the evidence, other published guidelines such as EAST22 were also factored into the decision tree, thereby supporting filter use in the prospective algorithm in selected critically ill patients with a concomitant contraindication to anticoagulation. Recent trends have shown increased use of IVC filters in various clinical scenarios despite the discrepancies in evidence-based guidelines.36,37 This trend has been partly driven by several factors, including the equivocal evidencebased guidelines for VTE prophylaxis, advances in filter technology that allow smaller-profile percutaneous delivery, and filter designs that allow retrievability. Retrievable filters seemingly have the advantage of protection from PE in the critically ill during the time in which they are most at risk but allow removal when the risk is lessened, thereby avoiding long-term potential complications such as vena cava occlusion. Current guidelines support use of retrievable filters if indications for a permanent filter are not present, risk of PE is acceptably low, return to high risk for VTE is not anticipated, life expectancy is reasonable, and the filter can be safely removed.28 On the basis of these guidelines, decision point 2 in the prospective algorithm directed filter type selection; thus, a retrievable filter (Günther-Tulip) was used in 80.3% of the study population, and a permanent filter (Stainless Steel Greenfield filter) in the remaining 19.6%. Additional analysis of
Killingsworth et al 1219
patients with a potentially retrievable filter showed that only 44% returned for potential removal, of whom 62.5% met criteria for retrieval, with an 85% technical success rate for retrieval. These findings for retrievability are similar to those reported in other series21,38-41 showing variability in patients returning for filter removal that likely reflects multifactorial issues, including ongoing disabilities, patient compliance, and lost follow-up. This reinforces the need for optional filters to provide adequate protection with a longer duration for potential filter retrieval. Although the current United States Food and Drug Administration (FDA) indications for use for the Günther-Tulip filter used in this study recommend removal ⱕ20 days, optimal timing for optional filters is still being evaluated. The data in this study would support other reports that removal of the Günther-Tulip filter can be performed well after 20 days.42 Although the indwell time for successful retrieval (average, 70 days) in this study was shorter than that in the unsuccessful group (average, 95 days), this did not reach statistical significance. Other newer-generation filter designs may allow for more extended filter retrieval windows, and the filter device selection used in the prospective algorithm is currently being re-evaluated but remains applicable to all current FDA-approved optional filter devices. Increasing use of IVC filters has also led to an expansion of bedside placement options, with transabdominal DUS or IVUS imaging, which have been shown in several reports to be safe and effective, with additional advantages including decreased risk of transport, no radiation or contrast exposure, decreased operating room and staff use, and cost-effectiveness.7-19 Because of increasing demands for filter placement at our institution, bedside techniques using IVUS guidance were disseminated according to prior experience levels and a proctored process. Once technical expertise was in place, the purpose of this study was then to evaluate the use of the prospectively implemented algorithm for IVUS-guided filter placement in critically ill patients, with decision point 3 determining bedside options. Of the 109 patients who met the criteria for IVUS-directed bedside IVC filter placement as defined in the prospectively implemented algorithm, technical feasibility was 98.1%, technical success was 97.2%, complications were few, and no patients died of filter- or PE-related causes. The outcomes in this study compare favorably with other published bedside filter experiences,7-19 but what is different in this study is the prospective algorithm-driven decision making that supported bedside placement. Furthermore, although other published reports advocate single- or dual-access techniques, the prospective algorithm in this study included a pathway preferring a single-access approach (94.4%) first, reserving dual access (5.6%) for inadequate imaging with single access or if direct visualization of the filter-delivery catheter tip at the renal vein was required. A preference for single access can avoid potential complications of additional access, but the need for a larger sheath for the IVUS probe does technically limit filter
1220 Killingsworth et al
choices for single access to filter-delivery systems of at least an 8F sheath. Regardless, with only minor modifications in the single-access techniques, application of bedside options for both a permanent (Greenfield) and optional filter system (Günther-Tulip or Celect) were possible with the prospective algorithm. Although this study highlights how an algorithm for bedside IVUS-guided filter placement can be implemented and is supported by prospectively defined end points and data acquisition showing its effectiveness, there are some limitations. First, data analysis was based on intent-to-treat for bedside IVUS-guided filter placement, but there was no control group, such as venographic-directed filters, for comparison of effectiveness. This single-arm design is based in part on the purpose of showing safe introduction and dissemination of IVUS-directed filter techniques at our institution within a group of dedicated practitioners, while at the same time acknowledging that standard venographic filter placement techniques are performed across several specialties at our institution, making wider comparison difficult. Furthermore, this study was designed to show effectiveness of prospective implementation of the IVUSguided filter algorithm only and was not powered for comparison with another group. CONCLUSIONS IVC filters remain an integral component in the management of VTE. Although evidence-based guidelines supporting the use of IVC filters in selected situations are still equivocal and the role of retrievable filters is still evolving, use of a prospective algorithm for bedside filter placement in critically ill patients can help standardize an approach for VTE management. Our data suggest that a prospectively implemented algorithmic approach based on the best available evidence allows appropriate and safe placement of bedside filters in the critically ill. AUTHOR CONTRIBUTIONS Conception and design: MAP Analysis and interpretation: CK, ST, MP, JW, GM, SM, DR, JK, LR, WJ, MAP Data collection: CK, MAP Writing the article: CK, MAP Critical revision of the article: CK, ST, MP, JW, GM, SM, DR, JK, LR, WJ, MAP Final approval of the article: CK, ST, MP, JW, GM, SM, DR, JK, LR, WJ, MAP Statistical analysis: CK, MAP Obtained funding: N/A Overall responsibility: MAP MAP, Marc A. Passman; MP, Mark A. Patterson. REFERENCES 1. Decousus H, Leizorovicz A, Parent F, Page Y, Tardy B, Girard P, et al. A clinical trial of vena cava filters in the prevention of pulmonary embolism in patients with proximal deep-vein thrombosis. New Engl J Med 1998;338:409-15.
JOURNAL OF VASCULAR SURGERY May 2010
2. PREPIC Study Group. Eight year follow-up of patients with permanent vena cava filters in the prevention of PE. The PREPIC (Prevention du Risque d’Embolie Pulmonaire par Interruption Cave Study) randomized study. Circulation 2005;112:416-22. 3. Greenfield LJ, Proctor MC. Twenty-year clinical experience with the Greenfield filter. Cardiovasc Surg 1995;3:199-205. 4. Becker DM, Philbrick JT, Selby JB. Inferior vena cava filters. Indications, safety, effectiveness. Arch Intern Med 1992;152:1985-94. 5. Stearley HE. Patients’ outcomes: intrahospital transportation and monitoring of critically ill patients by a specially trained ICU nursing staff. Am J Crit Care 1998;7:282-7. 6. Szem JW, Hydo LJ, Fischer E, Kapur S, Klemperer J, Barie PS. Highrisk intrahospital transport of critically ill patients: safety and outcome of the necessary “road trip.” Crit Care Med 1995;23:1660-6. 7. Neuzil DF, Garrard CL, Berkman RA, Pierce R, Naslund TC. Duplexdirected vena cava filter placement: report of initial experience. Surgery 1998;123:470-4. 8. Nunn C, Neuzil D, Naslund T, Bass JG, Jenkins JM, Pierce R, et al. Cost-effective method for bedside insertion of vena cava filters in trauma patients. J Trauma 1997;43:753-58. 9. Van Natta T, Morris JA Jr, Eddy VA, Nunn CR, Rutherford EJ, Neuzil D, et al. Elective bedside surgery in critically injured patients is safe and cost effective. Ann Surg 1998;227:618-26. 10. Conners M, Becker S, Guzman R, Passman M, Pierce R, Kelly T, et al. Duplex scan-directed placement of inferior vena cava filters: a five-year institutional experience. J Vasc Surg 2002;35:286-91. 11. Garrett J, Passman MA, Guzman RJ, Dattilo JB, Naslund TC. Expanding options for bedside placement of inferior vena cava filters with intravascular ultrasound when trans-abdominal Duplex ultrasound imaging is inadequate. Ann Vasc Surg 2004;18:329-34 12. Wellons ED, Rosenthal D, Shuler FW, Levitt AB, Matsumura J, Henderson VJ. Real-time intravascular ultrasound-guided placement of a removable inferior vena cava filter. J Trauma 2004;57:20-5. 13. Bonn J, Liu JB, Eschelman DJ, Sullivan KL, Pinheiro LW, Gardiner GA Jr. Intravascular ultrasound as an alternative to positive contrast vena cavography prior to filter placement. J Vasc Interv Radiol 1999;10: 843-9. 14. Oppat W, Chiou A, Matsumura J. Intravascular ultrasound guided vena cava filter placement. J Endovasc Surg 1999;6:285-7. 15. Matsuura JH, White RA, Kopchok G, Nishinian G, Woody JD, Rosenthal D, et al. Vena cava filter placement by intravascular ultrasound. Cardiovasc Surg 2001;9:571-4. 16. Ebaugh J, Chiou A, Morasch M, Matsumura J, Pearce W. Bedside vena cava filter placement guided with intravascular ultrasound. J Vasc Surg 2001;34:21-6. 17. Ashley D, Gamblin C, Burch S, Slois M. Accurate deployment of vena cava filters: comparison of intravascular ultrasound and contrast venography. J Trauma 2001;50:975-81. 18. Sing RF, Cicci CK, Smith CH, Messick WJ. Bedside insertion of inferior vena cava filters in the intensive care unit. J Trauma 1999;47:1104-7. 19. Tola JC, Holtzmann R, Lottenberg L. Bedside placement of inferior vena cava filters in the intensive care unit. Am Surg 1999;65:833-7. 20. Corriere MA, Passman MA, Dattilo JB, Guzman RJ, Naslund TC. Comparison of bedside trans-abdominal Duplex ultrasound versus contrast venography for inferior vena cava filter placement: What is the best imaging modality? Ann Vasc Surg 2005;19:229-34. 21. Rosenthal D, Wellons ED, Levitt AB, Shuler FW, O’Conner RE, Henderson VJ. Role of prophylactic temporary inferior vena cava filters placed at the ICU: bedside under intravascular ultrasound guidance in patients with multiple trauma. J Vasc Surg 2004;40:958-64. 22. Rogers FB, Cipolle MD, Velmahos G, Rozycki G, Luchette FA. Practice management guidelines for the prevention of venous thromboembolism in trauma patients: the EAST Practice Management Guidelines Work Group. J Trauma 2002;53:142-64. 23. Büller HR, Agnelli G, Hull RD, Hyers TM, Prins MH, Raskob GE. Antithrombotic therapy for venous thromboembolic disease: the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest 2004;126(suppl 3):401-28S.
JOURNAL OF VASCULAR SURGERY Volume 51, Number 5
24. Geerts WH, Pineo GF, Heit JA, Bergqvist D, Lassen MR, Colwell CW, et al. Prevention of venous thromboembolism. Seventh ACCP Conference on Thrombotic and Thrombolytic Therapy. Chest 2004;126:338-400S. 25. Young T, Tang H, Aukes J, Hughes R. Vena caval filters for the prevention of pulmonary embolism (Review). Cochrane Database Syst Rev 2007;CD006212. 26. Kearon C, Kahn SR, Agnelli G, Goldhaber S, Raskob GE, Comerota AJ, et al. Antithrombotic therapy for venous thromboembolic disease. Chest 2008;133:454-545S. 27. Geerts WH, Bergqvist D, Pineo GF, Heit JA, Samama CM, Lassen MR, et al. Prevention of venous thromboembolism. Chest 2008;133:381453S. 28. Kaufman JA, Kinney TB, Streiff MB, Sing RF, Proctor MC, Becker D, et al. Guidelines for the use of retrievable and convertible vena cava filters: report from the Society of Interventional Radiology Multidisciplinary Consensus Conference. J Vasc Interv Radiol 2006;17:449-59. 29. Passman MA, Dattilo JB, Guzman RJ, Naslund TC. Bedside placement of inferior vena cava filters by using transabdominal duplex ultrasonography and intravascular ultrasound imaging. J Vasc Surg 2005;42:1027-32. 30. Jacobs DL, Motaganahalli RL, Peterson BG. Bedside vena cava filter placement with intravascular ultrasound: a simple, accurate, single venous access method. J Vasc Surg 2007;46:1284-6. 31. Passman MA. Regarding “Bedside vena cava filter placement with intravascular ultrasound: a simple, accurate, single venous access method.” J Vasc Surg 2008;48:257. 32. Greenfield LJ, Rutherford RB, and participants in the vena caval filter consensus conference. Recommended reporting standards for vena caval filter placement and patient followup. J Vasc Surg 1999;30:573-9. 33. Geerts WH, Jay RM, Code KI, Chen E, Szalai JP, Saibil EA, Hamilton PA. A comparison of low-dose heparin with low-molecular-weight
Killingsworth et al 1221
34.
35.
36. 37. 38.
39.
40.
41.
42.
heparin as prophylaxis against venous thromboembolism after major trauma. N Engl J Med 1996;335:701-7. Khansarinia S, Dennis JW, Veldenz HC, Butcher JL, Hartland L. Prophylactic Greenfield filter placement in selected high-risk trauma patients. J Vasc Surg 1995;22:231-6. Carlin AM, Tyburski JG, Wilson RF, Steffes C. Prophylactic and therapeutic inferior vena cava filters to prevent pulmonary emboli in trauma patients. Arch Surg 2002;137:521-7. Stein PD, Kayali F, Olson RE. Twenty-one year trends in the use of inferior vena cava filters. Arch Intern Med 2004;164:1541-5. Jaff MR, Goldhaber SZ, Tapson VF. High utilization rate of vena cava filters in deep vein thrombosis. Thromb Haemost 2005;93:1117-9. Sing RF, Camp SM, Heniford BT, Rutherford EJ, Dix S, Reilly PM, et al. Timing of pulmonary emboli after trauma: implications for retrievable vena cava filters. J Trauma 2006;60:732-5. Karmy-Jones R, Jurkovich GJ, Velmahos GC, Burdick T, Spaniolas K, Todd SR, et al. Practice patterns and outcomes of retrievable vena cava filters in trauma patients: an AAST multicenter study. J Trauma 2007; 62:17-25. Meier C, Keller IS, Pfiffner R, Labler L, Trentz O, Pfammatter T. Early experience with the retrievable OptEase vena cava filter in high-risk trauma patients. Eur J Vasc Endovasc Surg 2006;32:589-95. Rosenthal D, Wellons ED, Lai KM, Bikk A, Henderson VJ. Retrievable inferior vena cava filters: early clinical experience. J Cardiovasc Surg 2005;46:163-9. Rosenthal D, Wellons ED, Hancock SM, Burkett AB. Retrievability of the Günther Tulip vena cava filter after dwell times longer than 180 days in patients with multiple trauma. J Endovasc Ther 2007;14:406-10.
Submitted Aug 12, 2009; accepted Dec 13, 2009.
Prospective implementation of an algorithm for bedside intravascular ultrasound-guided filter placement in critically ill patients Christopher D. Killingsworth, MD,a Steven M. Taylor, MD,b Mark A. Patterson, MD,b Jordan A. Weinberg, MD,c Gerald McGwin Jr, MS, PhD,c Sherry M. Melton, MD,c Donald A. Reiff, MD,c Jeffrey D. Kerby, MD,c Loring W. Rue, MD,c William D. Jordan Jr, MD,b and Marc A. Passman, MD,a Birmingham, Ala Background: Although contrast venography is the standard imaging method for inferior vena cava (IVC) filter insertion, intravascular ultrasound (IVUS) imaging is a safe and effective option that allows for bedside filter placement and is especially advantageous for immobilized critically ill patients by limiting resource use, risk of transportation, and cost. This study reviewed the effectiveness of a prospectively implemented algorithm for IVUS-guided IVC filter placement in this high-risk population. Methods: Current evidence-based guidelines were used to create a clinical decision algorithm for IVUS-guided IVC filter placement in critically ill patients. After a defined lead-in phase to allow dissemination of techniques, the algorithm was prospectively implemented on January 1, 2008. Data were collected for 1 year using accepted reporting standards and a quality assurance review performed based on intent-to-treat at 6, 12, and 18 months. Results: As defined in the prospectively implemented algorithm, 109 patients met criteria for IVUS-directed bedside IVC filter placement. Technical feasibility was 98.1%. Only 2 patients had inadequate IVUS visualization for bedside filter placement and required subsequent placement in the endovascular suite. Technical success, defined as proper deployment in an infrarenal position, was achieved in 104 of the remaining 107 patients (97.2%). The filter was permanent in 21 (19.6%) and retrievable in 86 (80.3%). The single-puncture technique was used in 101 (94.4%), with additional dual access required in 6 (5.6%). Periprocedural complications were rare but included malpositioning requiring retrieval and repositioning in three patients, filter tilt >15° in two, and arteriovenous fistula in one. The 30-day mortality rate for the bedside group was 5.5%, with no filter-related deaths. Conclusions: Successful placement of IVC filters using IVUS-guided imaging at the bedside in critically ill patients can be established through an evidence-based prospectively implemented algorithm, thereby limiting the need for transport in this high-risk population. ( J Vasc Surg 2010;51:1215-21.)
The effectiveness of inferior vena cava (IVC) filters as a treatment for the prevention of fatal venous thromboembolism (VTE) is well documented.1-4 Although original transvenous filters were placed in the operating room through open venous access, advances in technology and low-profile delivery systems allowed the evolution of percutaneous placement in the endovascular suite to become the standard technique. Venography has the advantage of providing accurate assessment of venous anatomic landmarks and real-time positioning of the filter under fluoroscopic guidance and remains the most common approach From the Division of General Surgery,a Sections Vascular Surgery and Endovascular Therapy,b and Trauma, Burns and Surgical Critical Care,c University of Alabama at Birmingham.University of Alabama at Birmingham. Competition of interest: none. Presented at the Twenty-first Annual Meeting of the American Venous Forum, Phoenix, Ariz, Feb 10-14, 2009. Correspondence: Marc A. Passman, MD, Section of Vascular Surgery and Endovascular Therapy, University of Alabama at Birmingham, BDB 503, 1530 3rd Ave South, Birmingham, AL 35294-0012 (e-mail: marc. [email protected]). The editors and reviewers of this article have no relevant financial relationships to disclose per the JVS policy that requires reviewers to decline review of any manuscript for which they may have a competition of interest. 0741-5214/$36.00 Copyright © 2010 by the Society for Vascular Surgery. doi:10.1016/j.jvs.2009.12.041
for the placement of IVC filters at most institutions. However, transporting critically ill and immobilized patients to the endovascular suite for filter placement is not without complication and requires extra support staff, cost, and time. Conventional filter placement also bears the additional risk of radiation and contrast exposure.5,6 In recent years, bedside placement of IVC filters guided by transabdominal duplex ultrasound (DUS) imaging7-10 or intravascular ultrasound (IVUS) imaging11-19 has been shown to be safe, effective, and reliable. It has the advantage of avoiding transportation of critically ill patients and has become a preferred approach for IVC filter placement in some high-volume centers.20,21 Yet, wider dissemination of bedside techniques elsewhere has been limited by concerns for adequate imaging, the potential of missed venous anomalies, learning curve issues, and comfort level using bedside techniques, concerns that have previously limited introduction of these bedside techniques at our institution. This study reviewed the effectiveness of a new prospectively implemented algorithm for IVUS-guided filter placement in critically ill patients. METHODS Filter algorithm. A clinical decision algorithm for IVUS-guided IVC filter placement in critically ill patients 1215
JOURNAL OF VASCULAR SURGERY May 2010
1216 Killingsworth et al
Fig. Prospectively implemented decision algorithm for intravascular ultrasound– guided bedside filter placement in critically ill patients. DVT, Deep venous thrombosis; ESRD, end-stage renal disease; GCS, Glasgow coma scale; ICU, intensive care unit; IVC, inferior vena cava; MR, magnetic resonance; PE, pulmonary embolism; PUD, peptic ulcer disease; VTE, venous thromboembolism.
was created from current evidence-based guidelines (Fig).22-27 ●
●
Decision point 1. Indications for placement included standard indications, consisting of documented deep venous thrombosis (DVT) or pulmonary embolism (PE) with contraindication or complication of anticoagulation, and prophylactic indications consisting of high risk for VTE and an increased bleeding risk precluding anticoagulation-based prophylaxis. Decision point 2. Choice of filter type was according to predetermination of permanent or optional needs using the following criteria: A permanent filter (Stainless Steel Greenfield Filter. Boston Scientific, Natick, Mass) was used when the anticipated risk of VTE extended beyond the acceptable retrieval window because of injury type or severity. An optional filter (Günther-Tulip Filter, Cook Inc, Bloomington, Ind) was used if there was a defined retrievable end point within 2 to 6 months, or if the end point was uncertain
●
but permanent need was possible. Patients with retrievable filters were evaluated for removal at 2 to 6 months after placement, and a decision to proceed with removal was determined by current medical status applied to accepted indications for retrieval.28 Decision point 3. Preprocedural lower extremity venous DUS imaging was obtained to evaluate the presence of lower extremity DVT. If available, pre-existing computed tomography scans were reviewed for vena cava and renal vein anatomic detail, thrombosis, or anomalies. The decision to place the filter at the bedside was based on critical illness in the intensive care unit setting, injury type and severity, and potential risk of transportation. Placement of the filter in the endovascular suite was reserved for noncritically ill patients for whom transportation risk was low, if femoral access was not possible (bilateral lower extremity DVT), or the presence of IVC thrombosis or vena caval anomalies precluded safe placement. Additional venographic guid-
JOURNAL OF VASCULAR SURGERY Volume 51, Number 5
ance was also required for failure of bedside techniques due to inadequate imaging or access, or if a bedside placed filter was malpositioned. Technique for IVUS-guided filter placement. The IVUS system used was the Volcano s5 System, with a Visions PV8.2F (8F, 8.3-MHz) IVUS catheter (Rancho Cordova, Calif). After sterile preparation and adequate local anesthesia, percutaneous femoral venous access was obtained by using anatomic landmarks, and an 8F, 10-cm sheath was placed for IVUS imaging. Access was preferably from the right side if there was no evidence of DVT at the access site, thereby providing a straighter course for filterdelivery catheters. If dual access was needed or DVT was noted in the right femoral position, then left-sided access was obtained. A 0.035-inch guidewire was directed into the vena cava, followed by the IVUS probe, which was advanced to the level of the right atrium of the heart. A pullback technique was used to sequentially identify the venous anatomic landmarks, including the right atrium, hepatic veins, renal veins, and the confluence of the iliac veins.29 The IVUS probe was directed just below the level of the lowest most renal vein, and IVC diameter measurements were confirmed to be ⬍28 mm before proceeding with filter delivery. With appropriate IVUS imaging of the vena cava confirmed, the technical decision to proceed with bedside placement of filter was based on preference for single-puncture technique first, reserving dual-access options if additional directed imaging was needed. For the single-access technique, the IVUS probe was used to precisely direct the end of the sheath to the level of the lowest most renal vein. Selected filters used in this study were based on need for 8F access for the IVUS probe to allow single-access technique through at least the samesized sheath or greater required for filter delivery. For the Günther-Tulip filter with predetermined marks, the filter-delivery catheter was advanced to the mark aligning the tip of the filter with the end of the 8F sheath, and then the sheath was withdrawn in a “pin-pull” fashion to allow deployment of the filter at the infrarenal level.30 For Greenfield filter without predetermined marks, the 8F sheath was up-sized to the 15F introducer sheath over a wire, and the IVUS probe was reloaded to direct the end of the sheath to the level of the renal veins. The 15F sheath was then pulled back over the IVUS probe a distance equivalent to the length that the filter-delivery catheter extended beyond the sheath (approximately 7 cm), so when the filter-delivery catheter was loaded into the sheath, the tip of the filter would precisely align with the lowest most renal vein upon deployment.31 For dual-access needs, if the iliac confluence required better localization, contralateral access was used as an adjunct to the single-puncture technique with placement of a 5F sheath, followed by passage of a 0.035-inch guidewire. With IVUS imaging of the vena cava through the initial ipsilateral access, the point at which the contralateral guidewire was visualized corresponded to the iliac confluence. An extended dual-access technique was used if imaging
Killingsworth et al 1217
alone with the single puncture was inadequate or if direct visualization of the filter-delivery catheter tip at the renal vein was required, or both. With the IVUS probe left at the renal vein level through the ipsilateral access, the corresponding sheath and filter-delivery catheter were directed to the level of the renal veins under direct vision through the contralateral access. Once the position of the filterdelivery catheter tip was confirmed at the renal veins, the IVUS probe was pulled back and the filter deployed. For completion imaging, the IVUS probe was carefully advanced to confirm apposition of the filter legs to the IVC wall and, in the absence of resistance, was directed through the filter legs to confirm position of the tip at the renal vein level. After removal of the filter-delivery catheter and sheath(s), gentle manual pressure over the puncture site(s) was applied for hemostasis. Postprocedural plain abdominal radiographs were obtained to verify filter position and alignment when possible relative to lumbar vertebral anatomy. Data analysis. There was a defined lead-in phase before the protocol was initiated to allow modification of the algorithm and dissemination of techniques among the vascular surgeons responsible for filter placement (M.A.P., S.M.T., M.A.P., W.D.J.). The algorithm was prospectively implemented on January 1, 2008, with a 1-year enrollment phase. Accepted reporting standards were used to collect the data,32 and a quality assurance data review was performed based on intent-to-treat at 6, 12, and 18 months. Technical feasibility was defined as intent to place the filter with IVUS guidance and determined by ability to obtain appropriate access and adequate visualization. Technical success for placement was defined as a properly aligned filter in the infrarenal position with the tip at or near the renal vein. Additional 30-day end points included procedural-related complications, filter-related problems, and death. Filter retrievability end points based on the data review intervals were also followed up at a minimum of 6 months after filter placement. Mean data were compared by t test, and values of P ⬍ .05 were considered significant. RESULTS Between January 1 and December 31, 2008, 109 patients met the criteria for IVUS-directed bedside IVC filter placement based on the prospectively implemented decision algorithm and constitute the study population. Of the 109 patients, 26 (24%) were women and 83 (76%) were men, with a mean age of 45 ⫾ 18.4 (SD) years (range, 16-79 years). Indications for filter placement included known DVT and contraindication to anticoagulation in 34 (31.2%), hemorrhagic complications from anticoagulation in setting of known DVT in 5 (4.6%), failure to prevent pulmonary thromboembolism despite proper anticoagulation in 3 (2.8%), and poor compliance with anticoagulation with known DVT in 1 (0.9%). A prophylactic filter was indicated in 66 patients (60.5%) due to multisystem trauma that increased the risk for DVT with concurrent contraindications to anticoagulation such as intracranial hemor-
JOURNAL OF VASCULAR SURGERY May 2010
1218 Killingsworth et al
Table. Demographic, risk factors, and indications for bedside vena cava filter placement using the prospectively implemented algorithm Patient characteristics Demographic data Male/female ratio Age, years Median Mean ⫾ standard deviation Range Risk factor characteristics, No. (%) Immobilization (exclude paraplegia) Paraplegia from spinal cord injury Limb trauma Closed head injury Spinal fracture Indications for filter placement, No. (%) Prophylactic placement with contraindication to anticoagulation DVT with contraindication to anticoagulation Hemorrhagic complications from anticoagulation with known DVT Failure of anticoagulation to prevent pulmonary embolism Poor compliance with anticoagulation and known DVT
N ⫽ 109 3.2:1 46 45 ⫾ 18.4 16-79 77 (70.6) 27 (24.8) 80 (73.4) 56 (51.4) 52 (47.7)
no instances of IVC occlusion, venous ulceration, or filter migration. Of the 86 patients with retrievable filters, 14 died at some interval after placement. This left 72 patients with potentially retrievable filters, and 32 (44.4%) returned for follow-up evaluation. The decision was made to attempt removal of the filter in 20 patients (62.5%) and leave the filter in 12 (37.5%). Of the 20 attempted filter retrievals, 17 (85%) were successful. The average time from placement to retrieval was 70.6 ⫾ 43 days (range 15-180 days) in the 17 successful removals and 95 ⫾ 17.6 days (range 82-115 days) in the 3 unsuccessful removals. These differences did not reach statistical significance on comparison of means (t test, P ⫽ .12). In the three unsuccessful attempts, filter incorporation into the IVC prevented retrieval. DISCUSSION
66 (60.5) 34 (31.2) 5 (4.6) 3 (2.8) 1 (0.9)
DVT, Deep venous thrombosis.
rhage or in patients with spinal cord injury with extended immobilization (Table). By intent-to-treat, bedside filter placement with IVUS was technically feasible in 107 of these 109 patients (98.1%). Bedside placement was not possible in two patients due to inadequate visualization on IVUS imaging. Of the 107 placed at bedside, technical success was 97.2%. Three retrievable filters were malpositioned— two in relation to the renal veins and required retrieval/ replacement, and one in the iliac vein that could not be retrieved and required an additional filter. A single-puncture venous access was used in 101 of the 107 placements (94.4%) and dual access was required in 6 (5.6%). A Günther-Tulip retrievable filter was used 86 patients (80.3%) and a permanent Stainless Steel Greenfield filter in the remaining 21 (19.6%). Complications occurred in 6 of 109 patients (5.5%), including three malpositioned filters as detailed above, one common femoral arteriovenous fistula that spontaneously resolved on ultrasound imaging at 3 months, and two filters with minor axial tilt. No instances of IVC occlusion or access site thrombosis were documented from clinical symptoms or ultrasound imaging ⱕ30 days. There were no procedurally related deaths, but six unrelated deaths (5.5%) occurred ⱕ30 days as a result of sepsis or complications from intracranial hemorrhage. Follow-up data at the 18-month interval showed no long-term filter complications. Only 4 of the 34 patients who returned for clinical follow up ⬎4 months were found to have postphlebitic chronic mild ankle or calf edema, which was treated with compression stockings. There were
Although evidence-based guidelines for treatment of documented VTE support therapeutic anticoagulation, anticoagulation may be contraindicated in the setting of critical illness, traumatic injury, or other situations associated with increased bleeding risk. Given the increased prevalence of VTE, associated risk of fatal PE, and difficulties with anticoagulation in the critically ill, vena caval interruption with filter devices is an important consideration. The positive effect of IVC insertion on survival remains unproven, but there is sufficient nonrandomized evidence proving effective protection from PE. The only randomized study in which patients with documented extremity DVT underwent permanent IVC filter placement showed a lower incidence of PE ⱕ12 days.1 A long-term follow-up of 8 years showed significant reduction in PE compared with controls (6.2% vs 15.1%), but an increased rate of recurrent DVT and no difference in survival.2 The American College of Chest Physicians (ACCP) current guidelines (ACCP 2008) strongly recommend placement of an IVC filter in patients with documented VTE and a contraindication to anticoagulation, complication from anticoagulation, or VTE despite therapeutic anticoagulation.26 Use of IVC filters for VTE prophylaxis is still debated, however. Conventional measures of thromboprophylaxis with anticoagulation or sequential mechanical compression devices, even when used properly, may be relatively ineffective in the setting of critical illness and multisystem trauma.33 Although ACCP 2004 suggested that evidence supporting the use of IVC filters for prophylaxis was equivocal,24 and the updated ACCP 2008 recommends against IVC filters for VTE prophylaxis in trauma patients,27 other evidence-based guidelines, including those from the 2002 Eastern Association for the Surgery of Trauma (EAST)22 suggest a role for filters in selected high-risk patients, and most series report a decreased incidence of all PE and fatal PE with the use of IVC filters in appropriately selected high-risk patients.18,19,34,35 Given the frequency of VTE and the complexity of multiple decision points, it would follow that an evidencebased algorithmic approach for management is necessary. The validity of these and other published evidence-based
JOURNAL OF VASCULAR SURGERY Volume 51, Number 5
guidelines was not tested in this study, but all decision points used in the prospective algorithm were derived from available evidence in the literature. Published recommendations were integrated into the prospective algorithm for filter placement in critically ill patients not only to guide clinical decision making for filter placement (decision point 1) but also to direct filter selection type (decision point 2) and bedside filter placement options (decision point 3). During a 1-year enrollment period, 109 critically ill patients met the criteria for IVUS-directed bedside filter placement based on the prospectively implemented decision algorithm and constituted the study population. Patient demographics, critical illness, and injury pattern in this study were also similar to other recently published reports mainly consisting of multisystem trauma patients at high risk for VTE.31 Regarding decision point 1, the assumption in this study in using current evidence-based guidelines is that the decision to place a filter in a critically ill patient depends on the strength of the data composing these recommendations. Although the recommendations for use of filters in patients with documented VTE who cannot receive anticoagulation is strong (39.5% of this study population), use for VTE prophylaxis remains equivocal (60.5% of this study population). Furthermore, although this prospective algorithm was based on ACCP 2004,23,24 the recommendations from ACCP 2008,26,27 which were published midway through study recruitment, were not significantly different for patients with documented VTE and contraindications to anticoagulation favoring IVC filter placement. The strength of the recommendations for VTE prophylaxis against filters was different between ACCP 200424 and ACCP 2008,27 but given the complexity of the evidence, other published guidelines such as EAST22 were also factored into the decision tree, thereby supporting filter use in the prospective algorithm in selected critically ill patients with a concomitant contraindication to anticoagulation. Recent trends have shown increased use of IVC filters in various clinical scenarios despite the discrepancies in evidence-based guidelines.36,37 This trend has been partly driven by several factors, including the equivocal evidencebased guidelines for VTE prophylaxis, advances in filter technology that allow smaller-profile percutaneous delivery, and filter designs that allow retrievability. Retrievable filters seemingly have the advantage of protection from PE in the critically ill during the time in which they are most at risk but allow removal when the risk is lessened, thereby avoiding long-term potential complications such as vena cava occlusion. Current guidelines support use of retrievable filters if indications for a permanent filter are not present, risk of PE is acceptably low, return to high risk for VTE is not anticipated, life expectancy is reasonable, and the filter can be safely removed.28 On the basis of these guidelines, decision point 2 in the prospective algorithm directed filter type selection; thus, a retrievable filter (Günther-Tulip) was used in 80.3% of the study population, and a permanent filter (Stainless Steel Greenfield filter) in the remaining 19.6%. Additional analysis of
Killingsworth et al 1219
patients with a potentially retrievable filter showed that only 44% returned for potential removal, of whom 62.5% met criteria for retrieval, with an 85% technical success rate for retrieval. These findings for retrievability are similar to those reported in other series21,38-41 showing variability in patients returning for filter removal that likely reflects multifactorial issues, including ongoing disabilities, patient compliance, and lost follow-up. This reinforces the need for optional filters to provide adequate protection with a longer duration for potential filter retrieval. Although the current United States Food and Drug Administration (FDA) indications for use for the Günther-Tulip filter used in this study recommend removal ⱕ20 days, optimal timing for optional filters is still being evaluated. The data in this study would support other reports that removal of the Günther-Tulip filter can be performed well after 20 days.42 Although the indwell time for successful retrieval (average, 70 days) in this study was shorter than that in the unsuccessful group (average, 95 days), this did not reach statistical significance. Other newer-generation filter designs may allow for more extended filter retrieval windows, and the filter device selection used in the prospective algorithm is currently being re-evaluated but remains applicable to all current FDA-approved optional filter devices. Increasing use of IVC filters has also led to an expansion of bedside placement options, with transabdominal DUS or IVUS imaging, which have been shown in several reports to be safe and effective, with additional advantages including decreased risk of transport, no radiation or contrast exposure, decreased operating room and staff use, and cost-effectiveness.7-19 Because of increasing demands for filter placement at our institution, bedside techniques using IVUS guidance were disseminated according to prior experience levels and a proctored process. Once technical expertise was in place, the purpose of this study was then to evaluate the use of the prospectively implemented algorithm for IVUS-guided filter placement in critically ill patients, with decision point 3 determining bedside options. Of the 109 patients who met the criteria for IVUS-directed bedside IVC filter placement as defined in the prospectively implemented algorithm, technical feasibility was 98.1%, technical success was 97.2%, complications were few, and no patients died of filter- or PE-related causes. The outcomes in this study compare favorably with other published bedside filter experiences,7-19 but what is different in this study is the prospective algorithm-driven decision making that supported bedside placement. Furthermore, although other published reports advocate single- or dual-access techniques, the prospective algorithm in this study included a pathway preferring a single-access approach (94.4%) first, reserving dual access (5.6%) for inadequate imaging with single access or if direct visualization of the filter-delivery catheter tip at the renal vein was required. A preference for single access can avoid potential complications of additional access, but the need for a larger sheath for the IVUS probe does technically limit filter
1220 Killingsworth et al
choices for single access to filter-delivery systems of at least an 8F sheath. Regardless, with only minor modifications in the single-access techniques, application of bedside options for both a permanent (Greenfield) and optional filter system (Günther-Tulip or Celect) were possible with the prospective algorithm. Although this study highlights how an algorithm for bedside IVUS-guided filter placement can be implemented and is supported by prospectively defined end points and data acquisition showing its effectiveness, there are some limitations. First, data analysis was based on intent-to-treat for bedside IVUS-guided filter placement, but there was no control group, such as venographic-directed filters, for comparison of effectiveness. This single-arm design is based in part on the purpose of showing safe introduction and dissemination of IVUS-directed filter techniques at our institution within a group of dedicated practitioners, while at the same time acknowledging that standard venographic filter placement techniques are performed across several specialties at our institution, making wider comparison difficult. Furthermore, this study was designed to show effectiveness of prospective implementation of the IVUSguided filter algorithm only and was not powered for comparison with another group. CONCLUSIONS IVC filters remain an integral component in the management of VTE. Although evidence-based guidelines supporting the use of IVC filters in selected situations are still equivocal and the role of retrievable filters is still evolving, use of a prospective algorithm for bedside filter placement in critically ill patients can help standardize an approach for VTE management. Our data suggest that a prospectively implemented algorithmic approach based on the best available evidence allows appropriate and safe placement of bedside filters in the critically ill. AUTHOR CONTRIBUTIONS Conception and design: MAP Analysis and interpretation: CK, ST, MP, JW, GM, SM, DR, JK, LR, WJ, MAP Data collection: CK, MAP Writing the article: CK, MAP Critical revision of the article: CK, ST, MP, JW, GM, SM, DR, JK, LR, WJ, MAP Final approval of the article: CK, ST, MP, JW, GM, SM, DR, JK, LR, WJ, MAP Statistical analysis: CK, MAP Obtained funding: N/A Overall responsibility: MAP MAP, Marc A. Passman; MP, Mark A. Patterson. REFERENCES 1. Decousus H, Leizorovicz A, Parent F, Page Y, Tardy B, Girard P, et al. A clinical trial of vena cava filters in the prevention of pulmonary embolism in patients with proximal deep-vein thrombosis. New Engl J Med 1998;338:409-15.
JOURNAL OF VASCULAR SURGERY May 2010
2. PREPIC Study Group. Eight year follow-up of patients with permanent vena cava filters in the prevention of PE. The PREPIC (Prevention du Risque d’Embolie Pulmonaire par Interruption Cave Study) randomized study. Circulation 2005;112:416-22. 3. Greenfield LJ, Proctor MC. Twenty-year clinical experience with the Greenfield filter. Cardiovasc Surg 1995;3:199-205. 4. Becker DM, Philbrick JT, Selby JB. Inferior vena cava filters. Indications, safety, effectiveness. Arch Intern Med 1992;152:1985-94. 5. Stearley HE. Patients’ outcomes: intrahospital transportation and monitoring of critically ill patients by a specially trained ICU nursing staff. Am J Crit Care 1998;7:282-7. 6. Szem JW, Hydo LJ, Fischer E, Kapur S, Klemperer J, Barie PS. Highrisk intrahospital transport of critically ill patients: safety and outcome of the necessary “road trip.” Crit Care Med 1995;23:1660-6. 7. Neuzil DF, Garrard CL, Berkman RA, Pierce R, Naslund TC. Duplexdirected vena cava filter placement: report of initial experience. Surgery 1998;123:470-4. 8. Nunn C, Neuzil D, Naslund T, Bass JG, Jenkins JM, Pierce R, et al. Cost-effective method for bedside insertion of vena cava filters in trauma patients. J Trauma 1997;43:753-58. 9. Van Natta T, Morris JA Jr, Eddy VA, Nunn CR, Rutherford EJ, Neuzil D, et al. Elective bedside surgery in critically injured patients is safe and cost effective. Ann Surg 1998;227:618-26. 10. Conners M, Becker S, Guzman R, Passman M, Pierce R, Kelly T, et al. Duplex scan-directed placement of inferior vena cava filters: a five-year institutional experience. J Vasc Surg 2002;35:286-91. 11. Garrett J, Passman MA, Guzman RJ, Dattilo JB, Naslund TC. Expanding options for bedside placement of inferior vena cava filters with intravascular ultrasound when trans-abdominal Duplex ultrasound imaging is inadequate. Ann Vasc Surg 2004;18:329-34 12. Wellons ED, Rosenthal D, Shuler FW, Levitt AB, Matsumura J, Henderson VJ. Real-time intravascular ultrasound-guided placement of a removable inferior vena cava filter. J Trauma 2004;57:20-5. 13. Bonn J, Liu JB, Eschelman DJ, Sullivan KL, Pinheiro LW, Gardiner GA Jr. Intravascular ultrasound as an alternative to positive contrast vena cavography prior to filter placement. J Vasc Interv Radiol 1999;10: 843-9. 14. Oppat W, Chiou A, Matsumura J. Intravascular ultrasound guided vena cava filter placement. J Endovasc Surg 1999;6:285-7. 15. Matsuura JH, White RA, Kopchok G, Nishinian G, Woody JD, Rosenthal D, et al. Vena cava filter placement by intravascular ultrasound. Cardiovasc Surg 2001;9:571-4. 16. Ebaugh J, Chiou A, Morasch M, Matsumura J, Pearce W. Bedside vena cava filter placement guided with intravascular ultrasound. J Vasc Surg 2001;34:21-6. 17. Ashley D, Gamblin C, Burch S, Slois M. Accurate deployment of vena cava filters: comparison of intravascular ultrasound and contrast venography. J Trauma 2001;50:975-81. 18. Sing RF, Cicci CK, Smith CH, Messick WJ. Bedside insertion of inferior vena cava filters in the intensive care unit. J Trauma 1999;47:1104-7. 19. Tola JC, Holtzmann R, Lottenberg L. Bedside placement of inferior vena cava filters in the intensive care unit. Am Surg 1999;65:833-7. 20. Corriere MA, Passman MA, Dattilo JB, Guzman RJ, Naslund TC. Comparison of bedside trans-abdominal Duplex ultrasound versus contrast venography for inferior vena cava filter placement: What is the best imaging modality? Ann Vasc Surg 2005;19:229-34. 21. Rosenthal D, Wellons ED, Levitt AB, Shuler FW, O’Conner RE, Henderson VJ. Role of prophylactic temporary inferior vena cava filters placed at the ICU: bedside under intravascular ultrasound guidance in patients with multiple trauma. J Vasc Surg 2004;40:958-64. 22. Rogers FB, Cipolle MD, Velmahos G, Rozycki G, Luchette FA. Practice management guidelines for the prevention of venous thromboembolism in trauma patients: the EAST Practice Management Guidelines Work Group. J Trauma 2002;53:142-64. 23. Büller HR, Agnelli G, Hull RD, Hyers TM, Prins MH, Raskob GE. Antithrombotic therapy for venous thromboembolic disease: the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest 2004;126(suppl 3):401-28S.
JOURNAL OF VASCULAR SURGERY Volume 51, Number 5
24. Geerts WH, Pineo GF, Heit JA, Bergqvist D, Lassen MR, Colwell CW, et al. Prevention of venous thromboembolism. Seventh ACCP Conference on Thrombotic and Thrombolytic Therapy. Chest 2004;126:338-400S. 25. Young T, Tang H, Aukes J, Hughes R. Vena caval filters for the prevention of pulmonary embolism (Review). Cochrane Database Syst Rev 2007;CD006212. 26. Kearon C, Kahn SR, Agnelli G, Goldhaber S, Raskob GE, Comerota AJ, et al. Antithrombotic therapy for venous thromboembolic disease. Chest 2008;133:454-545S. 27. Geerts WH, Bergqvist D, Pineo GF, Heit JA, Samama CM, Lassen MR, et al. Prevention of venous thromboembolism. Chest 2008;133:381453S. 28. Kaufman JA, Kinney TB, Streiff MB, Sing RF, Proctor MC, Becker D, et al. Guidelines for the use of retrievable and convertible vena cava filters: report from the Society of Interventional Radiology Multidisciplinary Consensus Conference. J Vasc Interv Radiol 2006;17:449-59. 29. Passman MA, Dattilo JB, Guzman RJ, Naslund TC. Bedside placement of inferior vena cava filters by using transabdominal duplex ultrasonography and intravascular ultrasound imaging. J Vasc Surg 2005;42:1027-32. 30. Jacobs DL, Motaganahalli RL, Peterson BG. Bedside vena cava filter placement with intravascular ultrasound: a simple, accurate, single venous access method. J Vasc Surg 2007;46:1284-6. 31. Passman MA. Regarding “Bedside vena cava filter placement with intravascular ultrasound: a simple, accurate, single venous access method.” J Vasc Surg 2008;48:257. 32. Greenfield LJ, Rutherford RB, and participants in the vena caval filter consensus conference. Recommended reporting standards for vena caval filter placement and patient followup. J Vasc Surg 1999;30:573-9. 33. Geerts WH, Jay RM, Code KI, Chen E, Szalai JP, Saibil EA, Hamilton PA. A comparison of low-dose heparin with low-molecular-weight
Killingsworth et al 1221
34.
35.
36. 37. 38.
39.
40.
41.
42.
heparin as prophylaxis against venous thromboembolism after major trauma. N Engl J Med 1996;335:701-7. Khansarinia S, Dennis JW, Veldenz HC, Butcher JL, Hartland L. Prophylactic Greenfield filter placement in selected high-risk trauma patients. J Vasc Surg 1995;22:231-6. Carlin AM, Tyburski JG, Wilson RF, Steffes C. Prophylactic and therapeutic inferior vena cava filters to prevent pulmonary emboli in trauma patients. Arch Surg 2002;137:521-7. Stein PD, Kayali F, Olson RE. Twenty-one year trends in the use of inferior vena cava filters. Arch Intern Med 2004;164:1541-5. Jaff MR, Goldhaber SZ, Tapson VF. High utilization rate of vena cava filters in deep vein thrombosis. Thromb Haemost 2005;93:1117-9. Sing RF, Camp SM, Heniford BT, Rutherford EJ, Dix S, Reilly PM, et al. Timing of pulmonary emboli after trauma: implications for retrievable vena cava filters. J Trauma 2006;60:732-5. Karmy-Jones R, Jurkovich GJ, Velmahos GC, Burdick T, Spaniolas K, Todd SR, et al. Practice patterns and outcomes of retrievable vena cava filters in trauma patients: an AAST multicenter study. J Trauma 2007; 62:17-25. Meier C, Keller IS, Pfiffner R, Labler L, Trentz O, Pfammatter T. Early experience with the retrievable OptEase vena cava filter in high-risk trauma patients. Eur J Vasc Endovasc Surg 2006;32:589-95. Rosenthal D, Wellons ED, Lai KM, Bikk A, Henderson VJ. Retrievable inferior vena cava filters: early clinical experience. J Cardiovasc Surg 2005;46:163-9. Rosenthal D, Wellons ED, Hancock SM, Burkett AB. Retrievability of the Günther Tulip vena cava filter after dwell times longer than 180 days in patients with multiple trauma. J Endovasc Ther 2007;14:406-10.
Submitted Aug 12, 2009; accepted Dec 13, 2009.