Increasing use of endovascular therapy in acute arterial injuries: Analysis of the National Trauma Data Bank

Article Outline

• Abstract
• Material and methods
• Results
• Discussion
• Conclusion
• Author contributions
• Appendix. 
• References
• Copyright

Objective

The application of endovascular technology for the emergency treatment of traumatic vascular injuries is a new frontier. This study examines recent nationwide use of endovascular therapy in acute arterial traumatic injuries.

Methods

This retrospective study used the National Trauma Data Bank (NTDB). Cases with a diagnosis of arterial vascular injury were identified according to the International Classification of Diseases, Ninth Revision, Clinical Modification, and procedure codes for endovascular therapy were selected. A descriptive analysis and multiple regressions were performed to identify variables predictive of outcomes.

Results

From 1994 to 2003, 12,732 arterial injuries were identified. Between 1997 (when the first endovascular repair was recorded in the NTDB) and 2003, 7286 open arterial repairs and 281 endovascular repairs were recorded for an overall utilization rate for endovascular procedures of 3.7%. The yearly number of endovascular procedures registered in the NTDB increased 27-fold, from four in 1997 to 107 in 2003. Use of stents substantially increased from 12 in 2000 to 30 in 2003; endograft use increased from one in 2000 to 37 in 2003. Nearly equal numbers of blunt (n = 134) and penetrating (n = 111) injuries were treated. The injury severity score (median, interquartile range [IRQ]) was significantly lower in patients who underwent an endovascular procedure at 13 (IRQ, 9 to 26) for trauma vs patients requiring an open procedure at 20 (IRQ, 10 to 34; P < .001), a finding corroborated by the lower number of associated injuries in patients undergoing endovascular repair (8.7 ± 7.2 vs 13.0 ± 16.1, P < .001). Using multivariable regression to control for differences in injury severity score and associated injuries, mortality was significantly lower for patients undergoing endovascular procedures (odds ratio, 0.18; P = .029) including those with an arterial injury of the torso or head and neck (odds ratio, 0.51, P = .007). Total length of hospital stay also tended to be lower for patients undergoing endovascular procedures by 18% (P = .064).

Conclusion

The use of endovascular therapy in the setting of acute trauma is increasing in a dramatic fashion and is being used to treat a wide variety of vessels injured by blunt and penetrating mechanisms. Endovascular therapy appears to be particularly suitable for patients who present with less severe injuries and greater hemodynamic stability. These preliminary data suggest that the use of endovascular therapy for acute traumatic arterial injuries yields shorter lengths of stay and improved survival.

Endovascular surgery has realized great success in the elective treatment of vascular disease and is increasingly used in emergency situations such as ruptured aortic aneurysms.¹ However, its usefulness in the realm of vascular trauma, while potentially exciting, has not been systematically studied. Vascular injury, including occlusion, dissection, pseudoaneurysm, and arteriovenous fistula, may result from blunt or penetrating trauma. Although conventional open surgical repair is the standard treatment for most of these injuries, it is often challenging because of its urgent nature, associated injuries, possible anatomic distortion, excessive bleeding, and frequent contaminated fields.²

The utilization of endovascular surgery for the management of arterial vascular trauma offers many potential benefits. Angiography has always had a role in the diagnosis of arterial trauma, but endovascular therapy now allows treatment to be rapidly coupled with diagnosis. Embolization is a well-established technique for hemostasis and is preferable to open surgery when intervention for bleeding pelvic fractures is necessary.³, ⁴ Significant tissue destruction and contamination in the vicinity of the injury may make remote vascular access advantageous. Finally, endovascular balloon occlusion for vascular control may be used as an adjunct to surgery when conventional exposure is technically challenging.⁵

This study examines nationwide trends in the application of endovascular surgery to treat acute arterial trauma from 1994 to 2003, a period of time when elective endovascular techniques were being widely disseminated and popularized. Our goal was to assess new techniques distinctive to the modern endovascular era and their application to arterial trauma for vascular reconstruction, not just embolization or occlusion. We systematically examined the incidence of usage, injury patterns, and patient characteristics to better define when endovascular management is most feasible.

Back to Article Outline

Material and methods 

The most recently available version of the National Trauma Data Bank (NTDB, version 4.3),⁶ which includes data on trauma admissions between the years 1994 to 2003, was evaluated for endovascular interventions. The NTDB is maintained by the Subcommittee on Trauma Registry Programs from the American College of Surgeons Committee on Trauma. It contains a non-population-based sample of hospitalized trauma patients in the United States, with a strong contribution by larger hospitals with younger, more severely injured patients. All hospitals that participate have been designated as level I to IV trauma centers. The NTDB is the most complete national database for injured patients currently available. All data are subject to continuous quality improvement and a system of logistic checks set up by the NTDB. For NTDB 4.3, 448 trauma centers participated, with a predominance of facilities designated as level I and II.

Patient identification is based on the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM)⁷ diagnosis and procedure codes for arterial injury (Table I, online only). With this approach, we excluded venous injuries commonly associated with pelvic injuries and sought to study only recent usage of advanced endovascular therapy for major arterial injuries. Within this selected group, the data were further stratified by ICD-9-CM procedure codes for open or endovascular repair (Table II. (online only), Table III. (online only)).

Additional data included patient age, sex, race, hospital teaching status, hospital type (public or private), geographic region (Midwest, Northeast, South, West), type of injury (blunt or penetrating), first set of emergency department (ED) vital signs, Glasgow Coma Score (GCS) in the ED, total injury severity score (ISS), revised trauma score (RTS), total number of associated injuries, and total hospital length of stay and intensive care unit (ICU) days.

Categoric variables for the treatment groups were compared using χ² analysis or the Fisher exact test, as appropriate. Continuous variables were compared using the two-tailed t test. Two-group comparisons for ordered categoric variables were performed using the Wilcoxon-Mann-Whitney test. We used linear and logistic regression models to assess factors associated with length of stay and in-hospital mortality. Because hospital length of stay lacks normality of the residual error, log transformation analysis were used for hospital length of stay, and back-transformations were used to evaluate percentage change from the median value.⁸

Significance for all tests was set at P < .05. All P values are two-tailed. All analyses were performed using Stata 9.0 statistical software (StataCorp LP, College Station, Tex).

Back to Article Outline

Results 

We identified 12,732 patients with vascular injuries from 1994 to 2003. An endovascular technique was used to treat 281 patients, yielding an overall utilization rate of endovascular procedures for diagnosed acute arterial trauma of 2.2%. The first arterial endovascular procedure registered in the NTDB was in 1997, and only four were performed. Between 1997 and 2003, 7286 patients underwent open repair of their arterial injury, and 281 endovascular procedures (3.7%) were performed. Between 2000 and 2003, the proportion of endovascular interventions vs open for arterial trauma increased from 2.4% to 8.1% (Fig 1). Stent use substantially increased from 12 in 2000 to 30 in 2003; endograft use increased from one in 2000 to 37 in 2003.

• 
  • View Large Image
  • Download to PowerPoint

• Fig 1. 

  Endovascular (Endo) procedures by year submitted to the National Trauma Data Bank (NTDB) from 1997 to 2003. No endovascular procedures were recorded by the NTDB from 1994 to 1997. The progressive increase in open arterial repair between 1997 and 2000 corresponds with an overall increase in the number of centers reporting data to the NTDB and probably does not represent a true increase in the incidence of arterial trauma.

Our study attempted to identify arterial injuries only and corresponding use of endovascular techniques. We found two hypogastric arterial injuries requiring endovascular procedures and one that underwent an open procedure. This supports our attempt to exclude many of the coil embolizations that are frequently performed for refractory venous pelvic bleeding.

There were no differences in age or sex between the two treatment groups (Table IV). Roughly equal numbers of blunt and penetrating injuries were treated by either technique. More endovascular procedures were performed at public than private institutions (163 vs 118) and university hospitals compared with community hospitals (157 vs 124), perhaps reflecting that this type of emerging treatment is more likely to be applied in institutions with an academic affiliation. There were no regional variations in use of endovascular procedures vs open procedures among trauma centers reporting to the NTDB among the Northeast, Midwest, South, or West.

Table IV. Results for endovascular vs open procedures for arterial trauma from the National Trauma Data Bank from 1997 to 2003

Variable⁎	Endovascular	Open (2000-2003)	P
Procedures, total	281	7286	…
Age, y	34.8±16.3	35.1±16.5	NS
Sex, No. (%)
Female	52(18.5)	1523(20.9)	NS
Male	229(81.5)	5763(79.1)
Type of injury, No. (%)
Blunt	154(55)	3934(54)	NS
Penetrating	127(45)	3352(46)
Arteries most commonly injured by blunt trauma	Iliacartery	Thoracicaorta
	Thoracicaorta	Poplitealartery	…
	Internalcarotidartery	Brachialartery
Arteries most commonly injured by penetrating trauma	Brachialartery	Brachialartery
	Superficialfemoralartery	Superficialfemoralartery	…
Institution, No. (%)
Public	163(58)	2842(39)
Private	118(42)	4444(61)	…
University	157(56)	5027(69)
Community	124(44)	2259(31)	…
Geographic region
Northeast	16(5.7)	510(7)
South	116(41.3)	2703(37.1)	…
Midwest	89(31.7)	2098(28.8)
West	60(21.3)	1975(27.1)
Injury severity score†	13(9-26)	20(10-34)	<.001
Revised trauma score‡	6.9±2.1	5.8±3.1	<.001
Associated injuries, No.	8.7±7.2	13.0±16.1	<.001
Patients with head injuries, %	2.65	3.27	NS
ED GSC	12.1±4.7	11.2±5.1	.006
Mortality
Overall	29/281(10.3)	919/7286(12.6)	NS
Thoracic aortic injury	1/29(3)	337/1759(19)	.03
Length of stay, d
Overall	16.2±17.4	19.9±22.8	.06§
ICU	8.5±12.2	10.6±15.2

ED GSC, Emergency department Glasgow Coma Score; ICU, intensive care unit.

The ISS and RTS values were significantly different in the endovascular cohort compared with those who received an open procedure; both scores indicate that the endovascular patients were more stable on presentation than patients treated with open vascular repair (Table IV, Fig 2). Correspondingly, the number of associated injuries for patients undergoing endovascular procedures was also lower than for patients undergoing open procedures. There were fewer head injuries and higher GCS scores recorded in the ED for endovascular patients relative to those undergoing an open procedure.

• 
  • View Large Image
  • Download to PowerPoint

• Fig 2. 

  Patients with arterial injury undergoing endovascular procedures have a lower injury severity score and fewer associated injuries compared with patients undergoing open procedures. Values are the mean, error bars show the standard error of the mean.

Overall, slightly more patients undergoing an endovascular procedure survived to hospital discharge (89.7%) compared with patients undergoing an open procedure (87.3%) for their arterial injury, although this difference in crude mortality was not statistically significant (P = .25). Patients with thoracic aortic injuries seemed to benefit more from endovascular repair. During the study period, 29 patients with thoracic aortic injuries (ICD-9 diagnostic codes 901.0, 441.01, 441.1, and 441.2) were treated by an endovascular technique with one in-hospital death (3% mortality), but 1759 thoracic aortic injuries were treated by open surgery with 337 deaths (19% mortality). This difference in mortality for thoracic aortic injuries was statistically significant (P = .03)

Logistic regression analysis performed on the entire patient cohort also revealed an overall survival advantage for endovascular repair after controlling for ISS, patient age, number of associated injuries, and total ICU days (odds ratio for death, 0.18; 95% confidence interval [CI], 0.040 to 0.84, P = .029). To further explore this apparent difference in mortality, we used the same variables to perform a regression subanalysis of the outcomes of patients who sustained injuries to the large vessels of the torso or head and neck because successful endovascular management may spare the patient a very large magnitude or risky open surgical procedure. Patients with a torso arterial injury treated by an endovascular procedure had a significant survival advantage compared with an open procedure (odds ratio for death, 0.51; 95% CI, 0.31 to 0.83, P = .007).

We also attempted to determine whether endovascular therapy provided protection from stroke for head and neck arterial injuries or amputation for extremity arterial injuries. However, the NTDB contained no record of stroke, cerebrovascular accident, or amputation in our study population.

Endovascular procedures were associated with a trend towards decreased total length of overall hospital stay by 18% (P = .064) after controlling for ISS, patient age, number of associated injuries, total ICU days, and death.

Back to Article Outline

Discussion 

In patients with a traumatic arterial injury, a standard open surgical approach for repair is no longer the only option. The use of endovascular therapy in the setting of acute trauma is increasing in dramatic fashion, as evidenced by the steep slope of procedure volume in recent years (Fig 1). There was no corresponding change in the number of open procedures between 2000 and 2003. Our objective was to document this dramatic increase in endovascular arterial procedures and attempt to describe the patient population benefiting from these less-invasive procedures. To the best of our knowledge, this is the first systematic evaluation of endovascular technology for arterial injury in trauma among the myriad case reports and case series that have been published.

To date, these procedures have been performed on patients with fewer associated injuries and who are more stable, as demonstrated by ISS and RTS values. Traumatologists may feel more comfortable allowing relatively stable patients to be transported to a special procedures suite for the time required to accomplish catheter-based diagnosis and treatment. In our experience, unstable trauma patients are transported to the angiography suite only when open surgical interventions are considered ineffectual, such as for transcatheter embolization for control of tertiary arterial or venous hemorrhage, or both, after a pelvic or visceral injury. We excluded this patient population from our analysis by focusing on treatment of traumatic arterial injuries of anatomically named vessels using endovascular therapies such as stenting, endografts, and angioplasty. Feasibility of endovascular therapy in this population of patients has now been clearly demonstrated. Extension of such treatment to less stable patients with arterial injury, who might otherwise receive open surgery, will probably require modifications of infrastructure that provide access to advanced imaging capability and endovascular inventory in the same areas where ongoing trauma evaluation and resuscitation can occur.

The potential benefits of an endovascular procedure are obvious and may have been responsible for the improved outcomes in this group compared with patients receiving an open repair. In thoracic aortic trauma, for example, the benefits of an endovascular procedure include avoidance of thoracotomy-associated complications, no need for aortic cross-clamping with related ischemic events, less blood loss and fluid shifts, and no extracorporeal bypass. Indeed, when we analyzed traumatic injuries to the torso, a distinct in-hospital survival advantage was demonstrated for endovascular vs open repair. This advantage is also reflected in the trend toward decreased total hospital length of stay, even when controlling for ISS and the RTS, which are indicators of the patient’s physiologic status. Others have made the same observation.⁹, ¹⁰

There are potential disadvantages of endovascular procedures for acute arterial trauma. Many implants, especially endografts, require ongoing surveillance for complications such as migration or the development of endoleaks. The average age of trauma victims in our study was 34.8 years. These patients are presumably going to have a nearly normal life expectancy. Is long-term surveillance necessary? Continued evaluation of outcomes after endovascular treatment will be important and will add a level of complexity to follow-up care. In patients undergoing elective abdominal aortic aneurysm repair, mid-term survival and quality of life measures have not shown a distinct advantage of endovascular procedures compared with open procedures.¹¹ However, assuming that subsequent studies also show improved short-term survival with endovascular treatment, its widespread use is probably justified even the incidence of late complications is increased.

Improved outcomes may also result from the development of endovascular devices more appropriately sized for this patient population. Young trauma victims have smaller aortas compared with the older aneurysm patient population. In one study, trauma victims were noted to have an average thoracic aorta diameter of 19.3 mm, which is smaller than conventional endograft sizes.¹² We have been challenged by the need to fit large-diameter devices into small aortas, often resulting in device buckling or crimping.

The potential use of endovascular therapy as a bridge to an open procedure while the patient is recovering from associated injuries is another avenue that could potentially be explored in the future.

Our study is subject to the errors generally associated with extracting information from a large administrative database. ICD-9-CM diagnostic and procedure codes are used, which at times, can be vague and lack specificity, especially for procedures. The database is subject to selection bias and information bias from nonsystematic sampling. The quality of the data is difficult to assess independently and under-reporting is a potential confounder. We are unable to report the types of associated injuries or complications (such as stroke) and the type of concurrent surgical procedures (such as amputations) with accuracy given the constraints of the NTDB.

During the time period analyzed, the number of trauma centers contributing data to the NTDB did not remain constant, and the progressive increase in open surgical repairs observed from 1997 through 2000 corresponds with an overall increase in participating centers (Fig 1). However, from 2000 to 2003, the number of open surgical repairs for arterial trauma did plateau and remain relatively stable while a threefold to fourfold increase in endovascular interventions occurred. We think it is highly unlikely that the disproportionate rise in endovascular vs open procedures is simply due to a change in the composition of the centers reporting data to the NTDB.

With regard to specific vascular outcomes, the NTDB does not permit us to evaluate the appropriateness of the endovascular procedure given the clinical situation. Nor can we identify complications of treatment that are specifically attributable to the type of vascular repair used, such as development of endoleaks, or device migration, or long-term durability issues such as endograft longevity or in-stent restenosis.

Back to Article Outline

Conclusion 

Increased familiarity with and refinements in endovascular techniques have allowed surgeons to adopt this approach in nonelective situations such as ruptured aneurysms and trauma. The potential benefits of this technology make it a very attractive alternative to the previous standard, especially when treating trauma that involves vessels of the torso. Additional studies are necessary to confirm our finding of improved survival and to characterize the long-term performance of endovascular repairs in this young patient population. However, our findings provide a platform and justification for even more widespread use of catheter-based procedures for the treatment of arterial trauma.

Back to Article Outline

Author contributions 

Back to Article Outline

Appendix. 

Additional material for this article may be found online at www.jvascsurg.org.

Table I. (online only). International Classification of Diseases, Ninth Revision, Clinical Modification diagnosis codes used to define vascular injury, aneurysm, and arterial dissection in the study population

Table II. (online only). International Classification of Diseases, Ninth Revision, Clinical Modification procedure codes used to define endovascular procedures in the study population

Table III. (online only). International Classification of Diseases, Ninth Revision, Clinical Modification procedure codes used to define open surgical procedures in the study population

Back to Article Outline

References 

1. Greco G, Egorova N, Anderson PL, Gelijns A, Moskowitz A, Nowygrod R, et al. Outcomes of endovascular treatment of ruptured abdominal aortic aneurysms. J Vasc Surg. 2006;43:453–459
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (244 KB)
   • CrossRef
2. McArthur CS, Marin ML. Endovascular therapy for the treatment of arterial trauma. Mt Sinai J Med. 2004;71:4–11
   • View In Article
   • MEDLINE
3. Scalea TM, Sclafani S. Interventional techniques in vascular trauma. Surg Clin North Am. 2001;81:1281–1297xii
   • View In Article
   • Abstract
   • Full-Text PDF (3057 KB)
   • CrossRef
4. Mirza A, Ellis T. Initial management of pelvic and femoral fractures in the multiply injured patient. Crit Care Clin. 2004;20:159–170
   • View In Article
   • Full Text
   • Full-Text PDF (352 KB)
   • CrossRef
5. Marin ML, Veith FJ, Panetta TF, Cynamon J, Sanchez LA, Schwartz ML, et al. Transluminally placed endovascular stented graft repair for arterial trauma. J Vasc Surg. 1994;20:466–472discussion 472-3
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (4393 KB)
   • CrossRef
6. National Trauma Data Bank (NTDB). American College of Surgeons. http://www.facs.org/trauma/ntdb.html2006;Accessed Jan
   • View In Article
7. National Center for Health Statistics, United States Department of Health & Human Services. International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM). http://www.cdc.gov/nchs/about/otheract/icd9/abticd9.htm2006;Accessed Jan
   • View In Article
8. Vittinghoff E, Glidden D, Shiboski SC, McCulloch C. Statistics for biology and health. In:  Gail M editors. Regression methods in biostatistics (Linear, logistic, survival and repeated measures models). New York, NY: Springer Science; 2005;
   • View In Article
9. Lebl DR, Dicker RA, Spain DA, Brundage SI. Dramatic shift in the primary management of traumatic thoracic aortic rupture. Arch Surg. 2006;141:177–180
   • View In Article
   • MEDLINE
   • CrossRef
10. Hoornweg LL, Dinkelman MK, Goslings JC, Reekers JA, Verhagen HJ, Verhoeven EL, et al. Endovascular management of traumatic ruptures of the thoracic aorta: a retrospective multicenter analysis of 28 cases in The Netherlands. J Vasc Surg. 2006;43:1096–1102
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (102 KB)
    • CrossRef
11. Prinssen M, Buskens E, Blankensteijn JD. Quality of life endovascular and open AAA repair (Results of a randomised trial). Eur J Vasc Endovasc Surg. 2004;27:121–127
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (438 KB)
    • CrossRef
12. Borsa JJ, Hoffer EK, Karmy-Jones R, Fontaine AB, Bloch RD, Yoon JK, et al. Angiographic description of blunt traumatic injuries to the thoracic aorta with specific relevance to endograft repair. J Endovasc Ther. 2002;9(suppl 2):II84–II91
    • View In Article