- Email: [email protected]
National trends in utilization of inferior vena cava filters in the United States, 2000-2009
From the Society for Vascular Surgery
National trends in utilization of inferior vena cava filters in the United States, 2000-2009 SreyRam Kuy, MD, MHS,a Anahita Dua, MD,a Cheong J. Lee, MD,a Bhavin Patel, BS,b Sapan S. Desai, MD, PhD, MBA,c,d Arshish Dua, BS,e Aniko Szabo, PhD,f and Parag J. Patel, MD,b Milwaukee, Wisc; Houston, Tex; and Durham, NC Objective: To characterize national trends over a decade in utilization of inferior vena cava (IVC) filters in the U.S. by year, indication, hospital, and patient demographics. Methods: Retrospective cross-sectional study utilizing the Nationwide Inpatient Sample Database, 2000 to 2009. IVC filter placement was identified with International Classification of Disease, Ninth Edition codes. Survey weighting, bivariate, and multivariate analysis was performed. Results: The number of IVC filters placed in the U.S. increased by 234% over a decade, from 56,380 in 2000 to 132,049 in 2009. A total of 84.7% of patients had a pulmonary embolism or deep venous thrombosis. A total of 94.6% of IVC filters were placed in urban hospitals. The largest number of IVC filters was placed in the South, followed by
the Northeast, Midwest, and Western regions (38.7%, 25.8%, 22.4%, and 13%, respectively). Adjusting for other patient and hospital factors, independent predictors of IVC filter placement were year, hospital size, location, teaching status, patient age group 50 to 79 years, insurance status, and urgency of admission. Conclusions: The use of IVC filters has dramatically increased over the last decade in the U.S., with variation in utilization based on patient and hospital characteristics. The largest utilization of IVC filters was among patients aged 50 to 79 years, Medicare recipients, and the Southern region of the U.S. The majority of patients receiving IVC filters have a diagnosis of pulmonary embolism or deep venous thrombosis. (J Vasc Surg: Venous and Lym Dis 2014;2:15-20.)
Pulmonary embolism (PE) is the leading cause of preventable in-hospital mortality in the U.S., accounting for 300,000 deaths per year in the U.S.1 Inferior vena caval (IVC) interruption with a filter for the prevention of PE was first described in an animal model in 1968,2 and the results in human trials with the removable Eichelter intracaval filter were described in 1970.3 Lazar Greenfield then introduced a permanent IVC filter in 1973, which was inserted percutaneously in 1984.4 In 1979, only 2000 IVC filters were inserted nationally in the U.S.5 By 1999, there were 49,000 IVC filters placed that year in the U.S. In 2003, the U.S. Food and Drug Administration (FDA) approved the first retrievable filter.4,6 With the increased ease in IVC filter insertion and improvement in filter technology, it is unknown how much impact this has had on the utilization of IVC filters in the U.S. Since the landmark FDA approval of the retrievable IVC filter
in 2003, there have been few studies examining whether the national utilization of IVC filters in the U.S. has increased over this time period. Currently, there are several societal guidelines regarding the indications for insertion of IVC filters. The American College of Chest Physicians (ACCP) guidelines are the most stringent of these societal guidelines, and recommend IVC filter use in patients with known deep venous thrombosis (DVT) or PE and a contraindication to anticoagulation.7 Our aims were to (1) characterize national trends in number of IVC filters placed annually in the U.S. over a decade, from 2000 to 2009 and (2) determine what percentage of patients undergoing IVC filter placement have a PE or DVT.
From the Division of Vascular Surgery,a Division of Interventional Radiology,b Department of Surgery,e and Department of Biostatistics,f Medical College of Wisconsin, Milwaukee; the Department of Cardiothoracic and Vascular Surgery, University of Texas Houston, Houstonc; and the Department of Surgery, Duke University, Durham.d Author conflict of interest: none. Presented at the 2013 Vascular Annual Meeting of the Society for Vascular Surgery, San Francisco, Calif, May 30-June 1, 2013. Reprint requests: SreyRam Kuy, MD, MHS, Medical College of Wisconsin, Division of Vascular Surgery, 9200 W. Wisconsin Ave, Milwaukee, WI 53226 (e-mail: [email protected]). The editors and reviewers of this article have no relevant financial relationships to disclose per the Journal policy that requires reviewers to decline review of any manuscript for which they may have a conflict of interest. 2213-333X/$36.00 Copyright Ó 2014 by the Society for Vascular Surgery. http://dx.doi.org/10.1016/j.jvsv.2013.08.007
METHODS This was a retrospective cross-sectional analysis of hospital discharge data for 2000 to 2009 from the Health Care Utilization Project-Nationwide Inpatient Sample (HCUP-NIS) database, which is a stratified 20% sample of all inpatient admissions to nonfederal, acute care hospitals maintained by the Agency for Healthcare Research and Quality. It is the largest all-payer inpatient database in the U.S., with records from approximately 8 million hospital stays each year. This study received exemption from the Institutional Review Board at our institution because data were de-identified. Records were limited to hospitalized patients with placement of an IVC filter during that admission, as identified by International Classification of Diseases, Ninth Revision (ICD-9) procedure codes. ICD-9 diagnosis codes were also used to identify patients with a concurrent diagnosis of DVT or PE. 15
16 Kuy et al
JOURNAL OF VASCULAR SURGERY: VENOUS AND LYMPHATIC DISORDERS January 2014
Variables. Year was the primary independent variable of interest. Patient-level covariates included age, gender, and race/ethnicity (white, black, Hispanic, other, as coded in HCUP-NIS). Clinical covariates included admission urgency (elective vs nonelective) and insurance status (private, Medicaid, Medicare, self-pay). Hospital level variables included hospital location, region, and teaching status. The outcome of interest was IVC filter placement. Statistical analysis. Bivariate analysis of independent variables by outcomes was performed using c2 tests for categorical variables and analysis of variance for continuous variables. Multivariable logistic regression was used to examine placement of IVC filters. Data analysis and management were performed using SAS version 9.1 (Cary, NC). Statistical significance was set at a probability value of P # .05.
Table. Demographics of patients undergoing inferior vena cava (IVC) filter placement in the U.S., 2000-2009
Gender Age group, years
Race
Primary payer
RESULTS Demographics of patients undergoing IVC filter placement. Among patients who underwent placement of an IVC filter, only 35.8% had a diagnosis of PE, 70.9% had a diagnosis of DVT, and 84.7% of patients had a diagnosis of either PE or DVT (Table). The largest age group to undergo IVC filter placement were patients aged 50 to 79 years, comprising 57.1% of all IVC filters placed in the U.S. The majority of patients were Medicare patients (58.7%), with only 2.8% of patients being uninsured. Most IVC filters were placed in patients who had an emergent/urgent admission (77.1%) and were admitted through the emergency department (62.2%). Most IVC filters were placed in large (70.3%), urban (94.6%) hospitals. The largest number of IVC filters was placed in the South, followed by the Northeast, Midwest, and Western regions (38.7%, 25.8%, 22.4%, and 13%, respectively). National trends in IVC filter usage. There was a dramatic increase in IVC filter utilization over the study period (Fig 1). In the year 2000, there were 56,380 IVC filters placed in the U.S. In 2003, there were 83,657 IVC filters placed, an increase of 148% over the 3-year period. In the 3-year period after the FDA approval of IVC filters in 2003, there was a 139% increase in the number of IVC filters placed in the U.S., with 116,949 IVC filters inserted in 2006. By 2009, the number of IVC filters placed that year was 132,049 (P < .05). Overall, the number of IVC filters placed in the U.S. increased by 234% over this decade. When examined by year and indication, the number of IVC filters placed in patients with a PE or DVT increased yearly (Fig 2). However, among patients with an IVC filter, the percentage of patients who had a PE or DVT actually decreased over the decade period, with the lowest nadir in 2006. In the year 2000, 88% of patients with an IVC filter carried a diagnosis of PE or DVT. In 2003, this decreased to 86% of patients. By 2006, only 82% of patients with an IVC filter had a diagnosis of PE or DVT. By 2009, this had increased to 86% of patients. When examined by gender, the number of IVC filters has increased in women, but the percentage of IVC filters placed in women compared with men has actually
Weighted frequency
Percentage
Male Female <18
464,300 512,940 6696
47.5 52.5 0.7
18-49 50-79 80 White Black Hispanic Asian Other Medicare Medicaid Private Uninsured Other/unknown Emergent/urgent
166,430 558,117 246,049 556,573 114,741 55,633 8959 241,468 573,448 76,541 266,431 27,133 33,821 753,510
17.0 57.1 25.2 56.9 11.7 5.7 0.9 24.7 58.7 7.8 27.3 2.8 3.5 77.1
Elective Emergency room
223,864 476,067
22.9 62.2
Another facility Home No Yes No Yes Northeast
74,234 214,829 627,028 350,345 149,906 827,467 252,548
9.7 28.1 64.2 35.8 15.3 84.7 25.8
Midwest South West Small Medium Large Rural
219,234 378,188 127,403 71,597 217,332 685,142 52,340
22.4 38.7 13.0 7.4 22.3 70.3 5.4
Urban Nonteaching
921,732 434,944
94.6 44.7
Teaching
539,128
55.3
Characteristics
Urgency of admission Admission source PE PE or DVT Region of hospital
Hospital size Rural/urban hospital Hospital teaching status
DVT, Deep venous thrombosis; PE, pulmonary embolism.
decreased (P < .05; Fig 3). When examined by age group and insurance status, the number of IVC filters has increased across all age and insurance groups, but the percentage of IVC filters across the age and insurance groups has stayed stable (Figs 4 and 5). When examined by region, the number of IVC filters increased the greatest in the South, and the percentage of IVC filters placed in the South and West increased over the decade studied (Fig 6). Multivariable regression analysis. After adjusting for patient and hospital factors, independent predictors of IVC filter placement were year, hospital size, hospital location, hospital teaching status, patient age group 50 to 79 years, insurance status, and urgency of admission.
JOURNAL OF VASCULAR SURGERY: VENOUS AND LYMPHATIC DISORDERS Volume 2, Number 1
Kuy et al 17
Fig 1. National trends in inferior vena cava (IVC) filter utilization in the United States, 2000 to 2009.
DISCUSSION In this study, we found that the use of IVC filters has dramatically increased over the last decade in the U.S. (234%), with the majority of patients having a PE or DVT (84.7%). However, this study is limited by the inability to determine from this database whether these patients had a contraindication to anticoagulation or whether those patients without a diagnosis of PE or DVT had an acceptable indication for IVC filter placement. Currently, there is no consensus on the acceptable indications for IVC filter insertion, with guidelines varying in the U.S. by specialty society. The ACCP 2012 guidelines recommends IVC filter use in patients with a known
Fig 2. Trends in inferior vena cava (IVC) filter utilization by indication. DVT, Deep venous thrombosis; PE, pulmonary embolism.
Fig 3. Variations in inferior vena cava (IVC) filters utilization by gender.
DVT or PE and a contraindication to anticoagulation.7 In contrast, the Society of Interventional Radiology Clinical Practice Guidelines are broader, recommending IVC filter insertion for patients at risk for a clinically significant PE and a contraindication to anticoagulation, but they do stress that the sole function of an IVC filter is to prevent a clinically significant PE since IVC filters do not prevent or treat venous thrombosis.8 In contrast, the 2002 Eastern Association for the Surgery of Trauma (EAST) guidelines recommended consideration of prophylactic IVC filter use in high-risk trauma patients even in the absence of a documented DVT or PE.9 High-risk trauma patients were defined as patients with an ongoing bleed or recent brain, spinal cord or ocular injury, or multiple extremity injuries that preclude the use of pneumatic compression devices and who cannot be anticoagulated. In our study, 15.3% of patients did not have a diagnosis of DVT or PE. However, recent Markov modeling to examine costeffectiveness of prophylactic vs therapeutic use of IVC filters in high-risk trauma patients has shown that prophylactic IVC filter use was more costly and less effective as measured by quality-adjusted life years in high-risk trauma patients, with the study authors concluding that prophylactic placement of IVC filters in high-risk trauma patients is not cost-effective.10 In addition, a Cochrane review of IVC filter use for prevention of PE found that although IVC filters prevented PE, there was no reduction in mortality.11 These different societal guidelines regarding the placement of IVC filters prophylactically in patients at risk for DVT, in addition to the ease of IVC filter placement with improved technology and the presence of
18 Kuy et al
JOURNAL OF VASCULAR SURGERY: VENOUS AND LYMPHATIC DISORDERS January 2014
Fig 4. Variations in inferior vena cava (IVC) filter utilization by age group. Yrs, years.
Fig 5. Variations in inferior vena cava (IVC) filter utilization by insurance status.
retrievable filters, may all be factors contributing to the increased utilization of IVC filters over this decade period. In our study, as the number of IVC filters placed increased every year across the decade period studied, the percentage of patients who had an IVC filter and a diagnosis of PE or DVT decreased, reaching a nadir of 82% in 2006. Patients without a diagnosis of PE or DVT could possibly be from the trauma population where prophylactic filters were placed. Though these patients do not have an indication for an IVC filter as defined by the ACCP guidelines, it cannot be determined from this database whether these patients had an indication for an IVC filter based on the EAST guidelines (high-risk trauma patients in the absence of a PE or DVT). In addition to the variability in guidelines regarding indications of IVC filter placement, there are potential socioeconomic factors contributing to variability in IVC filter usage. We found variability in IVC filter usage by patient and hospital demographic, with most IVC filters placed in large, urban hospitals. More IVC filters were placed in the South than other regions and among patients aged 50 to 79 years old, compared with other age groups. Findings similar to ours have also been seen in other smaller population-based studies. A study utilizing the California Patient Discharge Database found a large variability between California hospitals in the use of IVC filters among patients with DVT.12 A study utilizing the State Inpatient Databases from 2006 to 2008 found that the average incidence of IVC filter placement varied widely across states and did not correlate with the incidence of
DVTs in those states.13 They found that states with the highest incidence of IVC filters per DVT had a greater rate of paid malpractice claims and higher general surgeon liability insurance premiums. There is also significant variability across trauma centers in the prophylactic use of IVC filters in high-risk trauma patients without DVT or PE (as defined by the EAST guidelines), ranging from as low as zero to as high as 206 IVC filters per one highrisk trauma patient, suggesting some trauma centers do not place prophylactic IVC filters in this high-risk trauma population, and others place filters in trauma patients who do not meet the EAST guidelines.14 Previous work has shown that compliance with any of the society guidelines on indications for IVC filter placement are poor, regardless of whether placement was done by interventional radiology, vascular surgery, or interventional cardiology.15 With the increase in IVC filter usage over this decade period, there has also been increasing recognition of the potential complications of IVC filters, which include an increased risk of deep venous thrombosis,16,17 caval thrombosis,18 filter migration, filter erosion through the caval wall, and filter fracture. Between 2005 and 2010, the FDA received 921 device adverse event reports involving retrievable IVC filters (328 events of filter migration, 146 filter embolizations, 70 perforations of the IVC, and 56 filter fractures).19 As a result, in 2010, the FDA issued an advisory about retrievable IVC filters, urging physicians to consider removing retrievable filters as soon as protection from PE is no longer needed. Retrievable filters are
JOURNAL OF VASCULAR SURGERY: VENOUS AND LYMPHATIC DISORDERS Volume 2, Number 1
Kuy et al 19
demographics. There has also been a decrease over this decade period in the percentage of patients with a diagnosis of DVT or PE among those who underwent IVC filter placement. Further investigation is warranted to examining the indications of IVC filter placement and factors in variabilty in IVC filter usage. Better consensus between professional society guidelines regarding indications for IVC filter placement will help providers in standardizing IVC filter usage. The Society for Vascular Surgery and Society of Interventional Radiology Predicting the Safety and Effectiveness of Inferior Vena Cava Filters (PRESERVE) registries are good steps forward in improving the care and surveillance of patients with IVC filters. AUTHOR CONTRIBUTIONS Conception and design: SK, AnD, BP, PP, AS, ArD, SD Analysis and interpretation: SK, AnD, BP, PP, AS, ArD, CL, SD Data collection: SK, AnD, BP, PP, AS, ArD, SD Writing the article: SK, AnD, BP, PP, AS, ArD, SD, CL Critical revision of the article: SK, AnD, BP, PP, AS, ArD, SD, CL Final approval of the article: SK, AnD, BP, PP, AS, ArD, SD, CL Statistical analysis: SK, AnD, BP, AS, ArD, SD Obtained funding: SK Overall responsibility: SK Fig 6. Variations in inferior vena cava (IVC) filter utilization by region.
often not removed. A systematic review of retrievable IVC filters found that only 34% of retrievable filters are removed, although some institutions report removal rates as low as 3.7%.20,21 Limitations. There are limitations to this study. This study examined IVC filters placed during inpatient hospitalizations, utilizing the Nationwide Inpatient Sample Database, which is a database of only inpatient hospitalizations. Unfortunately, outpatient data are not captured in this database. Therefore, we are not able to look at the rate of IVC filter placement on an outpatient setting. In this study, the placement of IVC filters was identified using ICD-9 codes. There is currently no differentiation between permanent and retrievable filters in ICD-9 coding. In addition, utilizing this database, we are unable to identify whether the IVC filter was placed by vascular surgeons, interventional radiologists, cardiologists, or other interventionalists. Unfortunately, we are not able to determine from this database whether patients had a contraindication to anticoagulation and whether or not they were on systemic anticoagulation during the IVC filter placement. CONCLUSIONS In summary, there has been a significant increase nationally in the utilization of IVC filters, with the most significant variation based on hospital and patient
REFERENCES 1. Tapson V. Acute pulmonary embolism. N Engl J Med 2008;358: 1037-52. 2. Eichelter P, Schenk WG Jr. Prophylaxis of pulmonary embolism: a new experimental approach with initial results. Arch Surg 1968;97:348-56. 3. Williams R, Schenk WG Jr. A removable intracaval filter for prevention of pulmonary embolism: early experience with the use of the Eichelter catheter in patients. Surgery 1970;68:999-1008. 4. Mauro M, Murphy K, Thomson K, Venbrux A, Morgan R, Zollikofer C. Image-guided interventions: expert radiology series, volume 2. Saunders: New York; 2008. 5. Stein P, Kayali F, Olson R. Twenty-one-year trends in the use of inferior vena cava filters. Arch Intern Med 2004;164:1541-5. 6. Günther Tulip vena cava filter product overview. Available at: http:// www.cookmedical.com/di/content/mmedia/TULOV1103.pdf. Accessed May 28, 2013. 7. Kearon C, Akl E, Comerota A, Prandoni P, Bounameaux H, Goldhaber S, et al. Antithrombotic therapy for VTE disease: antithrombotic therapy and prevention of thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines. Chest 2012;141(2 Suppl):e419S-94S. 8. Kaufman JA, Kinney TB, Streiff MB, Sing R, Proctor M, 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. 9. Rogers F, Cipolle M, Velmahos G, Rozycki G, Luchette F. 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. 10. Spangler E, Dillavou E, Smith K. Cost-effectiveness of guidelines for insertion of inferior vena cava filters in high-risk trauma patients. J Vasc Surg 2010;52:1537-45. 11. Young T, Tang H, Aukes J, Hughes R. Vena caval filters for the prevention of pulmonary embolism. Cochrane Database Syst Rev 2010; CD006212.
20 Kuy et al
JOURNAL OF VASCULAR SURGERY: VENOUS AND LYMPHATIC DISORDERS January 2014
12. White R, Geraghty E, Brunson A, Murin S, Wun T, Spencer F, et al. High variation between hospitals in vena cava filter use for venous thromboembolism. JAMA Intern Med 2013;173:506-12. 13. Meltzer A, Graham A, Kim J, Connolly P, Karwowski J, Bush H, et al. Clinical, demographic and medicolegal factors associated with geographic variation in inferior vena cava filter utilization: an interstate analysis. Surgery 2013;153:683-8. 14. Dossett L, Adams R, Cotton B. Unwarranted national variation in the use of prophylactic inferior vena cava filters after trauma: an analysis of the National Trauma Databank. J Trauma 2011;70:1066-70. 15. Baadh A, Zikria J, Rivoli S, Graham R, Javit D, Ansell J. Indications for inferior vena cava filter placement: do physicians comply with guidelines? J Vasc Interv Radiol 2012;23:989-95. 16. The PREPIC Study Group. Eight-year follow-up of patients with permanent vena cava filters in the prevention of pulmonary embolism. Circulation 2005;112:416-22. 17. Decousus H, Leizorovicz A, Parent F, Page Y, Tardy B, Girard P, et al. A clinical trial of vena caval filters in the prevention of pulmonary
18.
19.
20.
21.
embolism in patients with proximal deep-vein thrombosis. N Engl J Med 1998;338:409-15. Aziz F, Spate K, Wong J, Aruny J, Sumpio B. Changing patterns in the use of inferior vena cava filters: review of a single center experience. J Am Coll Surg 2007;205:564-9. Inferior Vena Cava (IVC) filters: initial communication: risk of adverse events with long term use. U.S. Department of Health & Human Services Food and Drug Administration. Available at: http://www.fda. gov/Safety/MedWatch/SafetyInformation/SafetyAlertsforHumanMe dicalProducts/ucm221707.htm. Accessed May 28, 2013. Angel LF, Tapson V, Galgon RE, Restrepo MI, Kaufman J. Systematic review of the use of retrievable inferior vena cava filters. J Vasc Interv Radiol 2011;22:1522-30.e3. Gaspard S, Gaspard D. Retrievable inferior vena cava filters are rarely removed. Am Surg 2009;75:426-8.
Submitted Jun 10, 2013; accepted Aug 23, 2013.
National trends in utilization of inferior vena cava filters in the United States, 2000-2009 SreyRam Kuy, MD, MHS,a Anahita Dua, MD,a Cheong J. Lee, MD,a Bhavin Patel, BS,b Sapan S. Desai, MD, PhD, MBA,c,d Arshish Dua, BS,e Aniko Szabo, PhD,f and Parag J. Patel, MD,b Milwaukee, Wisc; Houston, Tex; and Durham, NC Objective: To characterize national trends over a decade in utilization of inferior vena cava (IVC) filters in the U.S. by year, indication, hospital, and patient demographics. Methods: Retrospective cross-sectional study utilizing the Nationwide Inpatient Sample Database, 2000 to 2009. IVC filter placement was identified with International Classification of Disease, Ninth Edition codes. Survey weighting, bivariate, and multivariate analysis was performed. Results: The number of IVC filters placed in the U.S. increased by 234% over a decade, from 56,380 in 2000 to 132,049 in 2009. A total of 84.7% of patients had a pulmonary embolism or deep venous thrombosis. A total of 94.6% of IVC filters were placed in urban hospitals. The largest number of IVC filters was placed in the South, followed by
the Northeast, Midwest, and Western regions (38.7%, 25.8%, 22.4%, and 13%, respectively). Adjusting for other patient and hospital factors, independent predictors of IVC filter placement were year, hospital size, location, teaching status, patient age group 50 to 79 years, insurance status, and urgency of admission. Conclusions: The use of IVC filters has dramatically increased over the last decade in the U.S., with variation in utilization based on patient and hospital characteristics. The largest utilization of IVC filters was among patients aged 50 to 79 years, Medicare recipients, and the Southern region of the U.S. The majority of patients receiving IVC filters have a diagnosis of pulmonary embolism or deep venous thrombosis. (J Vasc Surg: Venous and Lym Dis 2014;2:15-20.)
Pulmonary embolism (PE) is the leading cause of preventable in-hospital mortality in the U.S., accounting for 300,000 deaths per year in the U.S.1 Inferior vena caval (IVC) interruption with a filter for the prevention of PE was first described in an animal model in 1968,2 and the results in human trials with the removable Eichelter intracaval filter were described in 1970.3 Lazar Greenfield then introduced a permanent IVC filter in 1973, which was inserted percutaneously in 1984.4 In 1979, only 2000 IVC filters were inserted nationally in the U.S.5 By 1999, there were 49,000 IVC filters placed that year in the U.S. In 2003, the U.S. Food and Drug Administration (FDA) approved the first retrievable filter.4,6 With the increased ease in IVC filter insertion and improvement in filter technology, it is unknown how much impact this has had on the utilization of IVC filters in the U.S. Since the landmark FDA approval of the retrievable IVC filter
in 2003, there have been few studies examining whether the national utilization of IVC filters in the U.S. has increased over this time period. Currently, there are several societal guidelines regarding the indications for insertion of IVC filters. The American College of Chest Physicians (ACCP) guidelines are the most stringent of these societal guidelines, and recommend IVC filter use in patients with known deep venous thrombosis (DVT) or PE and a contraindication to anticoagulation.7 Our aims were to (1) characterize national trends in number of IVC filters placed annually in the U.S. over a decade, from 2000 to 2009 and (2) determine what percentage of patients undergoing IVC filter placement have a PE or DVT.
From the Division of Vascular Surgery,a Division of Interventional Radiology,b Department of Surgery,e and Department of Biostatistics,f Medical College of Wisconsin, Milwaukee; the Department of Cardiothoracic and Vascular Surgery, University of Texas Houston, Houstonc; and the Department of Surgery, Duke University, Durham.d Author conflict of interest: none. Presented at the 2013 Vascular Annual Meeting of the Society for Vascular Surgery, San Francisco, Calif, May 30-June 1, 2013. Reprint requests: SreyRam Kuy, MD, MHS, Medical College of Wisconsin, Division of Vascular Surgery, 9200 W. Wisconsin Ave, Milwaukee, WI 53226 (e-mail: [email protected]). The editors and reviewers of this article have no relevant financial relationships to disclose per the Journal policy that requires reviewers to decline review of any manuscript for which they may have a conflict of interest. 2213-333X/$36.00 Copyright Ó 2014 by the Society for Vascular Surgery. http://dx.doi.org/10.1016/j.jvsv.2013.08.007
METHODS This was a retrospective cross-sectional analysis of hospital discharge data for 2000 to 2009 from the Health Care Utilization Project-Nationwide Inpatient Sample (HCUP-NIS) database, which is a stratified 20% sample of all inpatient admissions to nonfederal, acute care hospitals maintained by the Agency for Healthcare Research and Quality. It is the largest all-payer inpatient database in the U.S., with records from approximately 8 million hospital stays each year. This study received exemption from the Institutional Review Board at our institution because data were de-identified. Records were limited to hospitalized patients with placement of an IVC filter during that admission, as identified by International Classification of Diseases, Ninth Revision (ICD-9) procedure codes. ICD-9 diagnosis codes were also used to identify patients with a concurrent diagnosis of DVT or PE. 15
16 Kuy et al
JOURNAL OF VASCULAR SURGERY: VENOUS AND LYMPHATIC DISORDERS January 2014
Variables. Year was the primary independent variable of interest. Patient-level covariates included age, gender, and race/ethnicity (white, black, Hispanic, other, as coded in HCUP-NIS). Clinical covariates included admission urgency (elective vs nonelective) and insurance status (private, Medicaid, Medicare, self-pay). Hospital level variables included hospital location, region, and teaching status. The outcome of interest was IVC filter placement. Statistical analysis. Bivariate analysis of independent variables by outcomes was performed using c2 tests for categorical variables and analysis of variance for continuous variables. Multivariable logistic regression was used to examine placement of IVC filters. Data analysis and management were performed using SAS version 9.1 (Cary, NC). Statistical significance was set at a probability value of P # .05.
Table. Demographics of patients undergoing inferior vena cava (IVC) filter placement in the U.S., 2000-2009
Gender Age group, years
Race
Primary payer
RESULTS Demographics of patients undergoing IVC filter placement. Among patients who underwent placement of an IVC filter, only 35.8% had a diagnosis of PE, 70.9% had a diagnosis of DVT, and 84.7% of patients had a diagnosis of either PE or DVT (Table). The largest age group to undergo IVC filter placement were patients aged 50 to 79 years, comprising 57.1% of all IVC filters placed in the U.S. The majority of patients were Medicare patients (58.7%), with only 2.8% of patients being uninsured. Most IVC filters were placed in patients who had an emergent/urgent admission (77.1%) and were admitted through the emergency department (62.2%). Most IVC filters were placed in large (70.3%), urban (94.6%) hospitals. The largest number of IVC filters was placed in the South, followed by the Northeast, Midwest, and Western regions (38.7%, 25.8%, 22.4%, and 13%, respectively). National trends in IVC filter usage. There was a dramatic increase in IVC filter utilization over the study period (Fig 1). In the year 2000, there were 56,380 IVC filters placed in the U.S. In 2003, there were 83,657 IVC filters placed, an increase of 148% over the 3-year period. In the 3-year period after the FDA approval of IVC filters in 2003, there was a 139% increase in the number of IVC filters placed in the U.S., with 116,949 IVC filters inserted in 2006. By 2009, the number of IVC filters placed that year was 132,049 (P < .05). Overall, the number of IVC filters placed in the U.S. increased by 234% over this decade. When examined by year and indication, the number of IVC filters placed in patients with a PE or DVT increased yearly (Fig 2). However, among patients with an IVC filter, the percentage of patients who had a PE or DVT actually decreased over the decade period, with the lowest nadir in 2006. In the year 2000, 88% of patients with an IVC filter carried a diagnosis of PE or DVT. In 2003, this decreased to 86% of patients. By 2006, only 82% of patients with an IVC filter had a diagnosis of PE or DVT. By 2009, this had increased to 86% of patients. When examined by gender, the number of IVC filters has increased in women, but the percentage of IVC filters placed in women compared with men has actually
Weighted frequency
Percentage
Male Female <18
464,300 512,940 6696
47.5 52.5 0.7
18-49 50-79 80 White Black Hispanic Asian Other Medicare Medicaid Private Uninsured Other/unknown Emergent/urgent
166,430 558,117 246,049 556,573 114,741 55,633 8959 241,468 573,448 76,541 266,431 27,133 33,821 753,510
17.0 57.1 25.2 56.9 11.7 5.7 0.9 24.7 58.7 7.8 27.3 2.8 3.5 77.1
Elective Emergency room
223,864 476,067
22.9 62.2
Another facility Home No Yes No Yes Northeast
74,234 214,829 627,028 350,345 149,906 827,467 252,548
9.7 28.1 64.2 35.8 15.3 84.7 25.8
Midwest South West Small Medium Large Rural
219,234 378,188 127,403 71,597 217,332 685,142 52,340
22.4 38.7 13.0 7.4 22.3 70.3 5.4
Urban Nonteaching
921,732 434,944
94.6 44.7
Teaching
539,128
55.3
Characteristics
Urgency of admission Admission source PE PE or DVT Region of hospital
Hospital size Rural/urban hospital Hospital teaching status
DVT, Deep venous thrombosis; PE, pulmonary embolism.
decreased (P < .05; Fig 3). When examined by age group and insurance status, the number of IVC filters has increased across all age and insurance groups, but the percentage of IVC filters across the age and insurance groups has stayed stable (Figs 4 and 5). When examined by region, the number of IVC filters increased the greatest in the South, and the percentage of IVC filters placed in the South and West increased over the decade studied (Fig 6). Multivariable regression analysis. After adjusting for patient and hospital factors, independent predictors of IVC filter placement were year, hospital size, hospital location, hospital teaching status, patient age group 50 to 79 years, insurance status, and urgency of admission.
JOURNAL OF VASCULAR SURGERY: VENOUS AND LYMPHATIC DISORDERS Volume 2, Number 1
Kuy et al 17
Fig 1. National trends in inferior vena cava (IVC) filter utilization in the United States, 2000 to 2009.
DISCUSSION In this study, we found that the use of IVC filters has dramatically increased over the last decade in the U.S. (234%), with the majority of patients having a PE or DVT (84.7%). However, this study is limited by the inability to determine from this database whether these patients had a contraindication to anticoagulation or whether those patients without a diagnosis of PE or DVT had an acceptable indication for IVC filter placement. Currently, there is no consensus on the acceptable indications for IVC filter insertion, with guidelines varying in the U.S. by specialty society. The ACCP 2012 guidelines recommends IVC filter use in patients with a known
Fig 2. Trends in inferior vena cava (IVC) filter utilization by indication. DVT, Deep venous thrombosis; PE, pulmonary embolism.
Fig 3. Variations in inferior vena cava (IVC) filters utilization by gender.
DVT or PE and a contraindication to anticoagulation.7 In contrast, the Society of Interventional Radiology Clinical Practice Guidelines are broader, recommending IVC filter insertion for patients at risk for a clinically significant PE and a contraindication to anticoagulation, but they do stress that the sole function of an IVC filter is to prevent a clinically significant PE since IVC filters do not prevent or treat venous thrombosis.8 In contrast, the 2002 Eastern Association for the Surgery of Trauma (EAST) guidelines recommended consideration of prophylactic IVC filter use in high-risk trauma patients even in the absence of a documented DVT or PE.9 High-risk trauma patients were defined as patients with an ongoing bleed or recent brain, spinal cord or ocular injury, or multiple extremity injuries that preclude the use of pneumatic compression devices and who cannot be anticoagulated. In our study, 15.3% of patients did not have a diagnosis of DVT or PE. However, recent Markov modeling to examine costeffectiveness of prophylactic vs therapeutic use of IVC filters in high-risk trauma patients has shown that prophylactic IVC filter use was more costly and less effective as measured by quality-adjusted life years in high-risk trauma patients, with the study authors concluding that prophylactic placement of IVC filters in high-risk trauma patients is not cost-effective.10 In addition, a Cochrane review of IVC filter use for prevention of PE found that although IVC filters prevented PE, there was no reduction in mortality.11 These different societal guidelines regarding the placement of IVC filters prophylactically in patients at risk for DVT, in addition to the ease of IVC filter placement with improved technology and the presence of
18 Kuy et al
JOURNAL OF VASCULAR SURGERY: VENOUS AND LYMPHATIC DISORDERS January 2014
Fig 4. Variations in inferior vena cava (IVC) filter utilization by age group. Yrs, years.
Fig 5. Variations in inferior vena cava (IVC) filter utilization by insurance status.
retrievable filters, may all be factors contributing to the increased utilization of IVC filters over this decade period. In our study, as the number of IVC filters placed increased every year across the decade period studied, the percentage of patients who had an IVC filter and a diagnosis of PE or DVT decreased, reaching a nadir of 82% in 2006. Patients without a diagnosis of PE or DVT could possibly be from the trauma population where prophylactic filters were placed. Though these patients do not have an indication for an IVC filter as defined by the ACCP guidelines, it cannot be determined from this database whether these patients had an indication for an IVC filter based on the EAST guidelines (high-risk trauma patients in the absence of a PE or DVT). In addition to the variability in guidelines regarding indications of IVC filter placement, there are potential socioeconomic factors contributing to variability in IVC filter usage. We found variability in IVC filter usage by patient and hospital demographic, with most IVC filters placed in large, urban hospitals. More IVC filters were placed in the South than other regions and among patients aged 50 to 79 years old, compared with other age groups. Findings similar to ours have also been seen in other smaller population-based studies. A study utilizing the California Patient Discharge Database found a large variability between California hospitals in the use of IVC filters among patients with DVT.12 A study utilizing the State Inpatient Databases from 2006 to 2008 found that the average incidence of IVC filter placement varied widely across states and did not correlate with the incidence of
DVTs in those states.13 They found that states with the highest incidence of IVC filters per DVT had a greater rate of paid malpractice claims and higher general surgeon liability insurance premiums. There is also significant variability across trauma centers in the prophylactic use of IVC filters in high-risk trauma patients without DVT or PE (as defined by the EAST guidelines), ranging from as low as zero to as high as 206 IVC filters per one highrisk trauma patient, suggesting some trauma centers do not place prophylactic IVC filters in this high-risk trauma population, and others place filters in trauma patients who do not meet the EAST guidelines.14 Previous work has shown that compliance with any of the society guidelines on indications for IVC filter placement are poor, regardless of whether placement was done by interventional radiology, vascular surgery, or interventional cardiology.15 With the increase in IVC filter usage over this decade period, there has also been increasing recognition of the potential complications of IVC filters, which include an increased risk of deep venous thrombosis,16,17 caval thrombosis,18 filter migration, filter erosion through the caval wall, and filter fracture. Between 2005 and 2010, the FDA received 921 device adverse event reports involving retrievable IVC filters (328 events of filter migration, 146 filter embolizations, 70 perforations of the IVC, and 56 filter fractures).19 As a result, in 2010, the FDA issued an advisory about retrievable IVC filters, urging physicians to consider removing retrievable filters as soon as protection from PE is no longer needed. Retrievable filters are
JOURNAL OF VASCULAR SURGERY: VENOUS AND LYMPHATIC DISORDERS Volume 2, Number 1
Kuy et al 19
demographics. There has also been a decrease over this decade period in the percentage of patients with a diagnosis of DVT or PE among those who underwent IVC filter placement. Further investigation is warranted to examining the indications of IVC filter placement and factors in variabilty in IVC filter usage. Better consensus between professional society guidelines regarding indications for IVC filter placement will help providers in standardizing IVC filter usage. The Society for Vascular Surgery and Society of Interventional Radiology Predicting the Safety and Effectiveness of Inferior Vena Cava Filters (PRESERVE) registries are good steps forward in improving the care and surveillance of patients with IVC filters. AUTHOR CONTRIBUTIONS Conception and design: SK, AnD, BP, PP, AS, ArD, SD Analysis and interpretation: SK, AnD, BP, PP, AS, ArD, CL, SD Data collection: SK, AnD, BP, PP, AS, ArD, SD Writing the article: SK, AnD, BP, PP, AS, ArD, SD, CL Critical revision of the article: SK, AnD, BP, PP, AS, ArD, SD, CL Final approval of the article: SK, AnD, BP, PP, AS, ArD, SD, CL Statistical analysis: SK, AnD, BP, AS, ArD, SD Obtained funding: SK Overall responsibility: SK Fig 6. Variations in inferior vena cava (IVC) filter utilization by region.
often not removed. A systematic review of retrievable IVC filters found that only 34% of retrievable filters are removed, although some institutions report removal rates as low as 3.7%.20,21 Limitations. There are limitations to this study. This study examined IVC filters placed during inpatient hospitalizations, utilizing the Nationwide Inpatient Sample Database, which is a database of only inpatient hospitalizations. Unfortunately, outpatient data are not captured in this database. Therefore, we are not able to look at the rate of IVC filter placement on an outpatient setting. In this study, the placement of IVC filters was identified using ICD-9 codes. There is currently no differentiation between permanent and retrievable filters in ICD-9 coding. In addition, utilizing this database, we are unable to identify whether the IVC filter was placed by vascular surgeons, interventional radiologists, cardiologists, or other interventionalists. Unfortunately, we are not able to determine from this database whether patients had a contraindication to anticoagulation and whether or not they were on systemic anticoagulation during the IVC filter placement. CONCLUSIONS In summary, there has been a significant increase nationally in the utilization of IVC filters, with the most significant variation based on hospital and patient
REFERENCES 1. Tapson V. Acute pulmonary embolism. N Engl J Med 2008;358: 1037-52. 2. Eichelter P, Schenk WG Jr. Prophylaxis of pulmonary embolism: a new experimental approach with initial results. Arch Surg 1968;97:348-56. 3. Williams R, Schenk WG Jr. A removable intracaval filter for prevention of pulmonary embolism: early experience with the use of the Eichelter catheter in patients. Surgery 1970;68:999-1008. 4. Mauro M, Murphy K, Thomson K, Venbrux A, Morgan R, Zollikofer C. Image-guided interventions: expert radiology series, volume 2. Saunders: New York; 2008. 5. Stein P, Kayali F, Olson R. Twenty-one-year trends in the use of inferior vena cava filters. Arch Intern Med 2004;164:1541-5. 6. Günther Tulip vena cava filter product overview. Available at: http:// www.cookmedical.com/di/content/mmedia/TULOV1103.pdf. Accessed May 28, 2013. 7. Kearon C, Akl E, Comerota A, Prandoni P, Bounameaux H, Goldhaber S, et al. Antithrombotic therapy for VTE disease: antithrombotic therapy and prevention of thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines. Chest 2012;141(2 Suppl):e419S-94S. 8. Kaufman JA, Kinney TB, Streiff MB, Sing R, Proctor M, 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. 9. Rogers F, Cipolle M, Velmahos G, Rozycki G, Luchette F. 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. 10. Spangler E, Dillavou E, Smith K. Cost-effectiveness of guidelines for insertion of inferior vena cava filters in high-risk trauma patients. J Vasc Surg 2010;52:1537-45. 11. Young T, Tang H, Aukes J, Hughes R. Vena caval filters for the prevention of pulmonary embolism. Cochrane Database Syst Rev 2010; CD006212.
20 Kuy et al
JOURNAL OF VASCULAR SURGERY: VENOUS AND LYMPHATIC DISORDERS January 2014
12. White R, Geraghty E, Brunson A, Murin S, Wun T, Spencer F, et al. High variation between hospitals in vena cava filter use for venous thromboembolism. JAMA Intern Med 2013;173:506-12. 13. Meltzer A, Graham A, Kim J, Connolly P, Karwowski J, Bush H, et al. Clinical, demographic and medicolegal factors associated with geographic variation in inferior vena cava filter utilization: an interstate analysis. Surgery 2013;153:683-8. 14. Dossett L, Adams R, Cotton B. Unwarranted national variation in the use of prophylactic inferior vena cava filters after trauma: an analysis of the National Trauma Databank. J Trauma 2011;70:1066-70. 15. Baadh A, Zikria J, Rivoli S, Graham R, Javit D, Ansell J. Indications for inferior vena cava filter placement: do physicians comply with guidelines? J Vasc Interv Radiol 2012;23:989-95. 16. The PREPIC Study Group. Eight-year follow-up of patients with permanent vena cava filters in the prevention of pulmonary embolism. Circulation 2005;112:416-22. 17. Decousus H, Leizorovicz A, Parent F, Page Y, Tardy B, Girard P, et al. A clinical trial of vena caval filters in the prevention of pulmonary
18.
19.
20.
21.
embolism in patients with proximal deep-vein thrombosis. N Engl J Med 1998;338:409-15. Aziz F, Spate K, Wong J, Aruny J, Sumpio B. Changing patterns in the use of inferior vena cava filters: review of a single center experience. J Am Coll Surg 2007;205:564-9. Inferior Vena Cava (IVC) filters: initial communication: risk of adverse events with long term use. U.S. Department of Health & Human Services Food and Drug Administration. Available at: http://www.fda. gov/Safety/MedWatch/SafetyInformation/SafetyAlertsforHumanMe dicalProducts/ucm221707.htm. Accessed May 28, 2013. Angel LF, Tapson V, Galgon RE, Restrepo MI, Kaufman J. Systematic review of the use of retrievable inferior vena cava filters. J Vasc Interv Radiol 2011;22:1522-30.e3. Gaspard S, Gaspard D. Retrievable inferior vena cava filters are rarely removed. Am Surg 2009;75:426-8.
Submitted Jun 10, 2013; accepted Aug 23, 2013.