existing trauma and critical care scoring systems

8/12/2019 Existing Trauma and Critical Care Scoring Systems

1/8

From the Society for Clinical Vascular Surgery

Existing trauma and critical care scoring systemsunderestimate mortality among vascular trauma

patientsShang A. Loh, MD, Caron B. Rockman, MD, Christine Chung, BS, Thomas S. Maldonado, MD,

Mark A. Adelman, MD, Neal S. Cayne, MD, H. Leon Pachter, MD,andFiras F. Mussa, MD, New York, NY

Background:The impact of vascular injuries on patient mortality has not been well evaluated in multi-trauma patients.This study seeks to determine (1) whether the presence of vascular trauma negatively affects outcome compared withnonvascular trauma (NVT) and (2) the utility of existing severity scoring systems in predicting mortality among vasculartrauma (VT) patients.

Methods: A retrospective review of our trauma database from January 2005 to December 2007 was conducted.Demographics, Injury Severity Scores (ISS), Revised Trauma Scores (RTS), Trauma ScoreInjury Severity Scores(TRISS), Acute Physiology and Chronic Health Evaluation II (APACHE II) scores, and mortality rates were compared.

Control patients were selected from a matching cohort based on ISS. Comparisons were made between groups based onthe above scoring systems. Statistical analysis used 2 analysis and Studentt-tests.

Results:Fifty VT and 50 NVT patients were identified with no significant differences in age, gender, mechanism of injury,ISS, RTS, or TRISS. The mean APACHE II score was higher in VT compared with NVT (12.3 vs 8.8, P< .05). Overallmortality was higher in VT compared with NVT but did not reach statistical significance (24% vs 11.8%, P .108). VTpatients with RTS score>5 had a higher mortality rate (26% vs 2.2%,P .007). VT patients with an ISS score >24 hada higher mortalitycompared with NVT patients (61% vs 28.6%,P .04). VT patients with an APACHE II score80%had a higher mortality rate (13.9% vs 0%, P .05).

Conclusions:In multi-trauma patients, the presence of vascular injury was associated with increased mortality in less

severely injured patients based on the RTS, TRISS, and APACHE II scores. These scoring systems underestimatedmortality in patients with vascular trauma. Level of care and future trauma algorithms should be adjusted in the presenceof vascular trauma. ( J Vasc Surg 2011;53:359-66.)

Trauma is the leading cause of mortality in the first four

decades of life.1

While the epidemiology and outcomes ofnonvascular trauma have been well studied, vascular traumahas not been well characterized outside of the militaryexperience.

1, 2

More than 10 years ago, Caps reported thatthe incidence of vascular trauma had significantly increasedover the previous 5 decades, contributing to overall traumamorbidity and mortality.2 Furthermore, studies haveshown that abdominal vascular trauma has been associated

with higher overall mortality and poorer functional out-comes than seen in the general trauma population.3-7

However, none of these studies matched patients based onseverity of injury. In addition, current trauma scoring sys-tems, which utilize either physiologic or anatomic parame-

ters to determine injury severity,8-10 have not been vali-

dated specifically for vascular trauma. The aim of this studywas to examine mortality in patients with vascular trauma atan urban level I trauma center compared with patients withnonvascular trauma and matching Injury Severity Scores(ISS). Furthermore, we sought to evaluate the usefulness ofcommon injury severity scoring systems in predicting mor-tality for patients with vascular injuries.

METHODS

Study design. A retrospective analysis of the traumadatabase from January 2005 to December 2007 at an urbanlevel I trauma center was conducted. Vascular trauma (VT)patients were identified based on International Classifica-

tion of Disease9 (ICD-9) diagnoses codes and repre-sented all vascular trauma for this time frame. A contempo-rary cohort of multi-trauma patients with nonvasculartrauma (NVT) and matching ISS was selected as a controlgroup. The study was approved by the Institutional ReviewBoard. Parameters collected included: age, sex, vital signs,Glasgow Coma Scale (GCS), mechanism of injury, locationand severity of injury, mode of diagnosis, need for a vascularconsult, need for vascular intervention, and mortality. Vas-cular injury severity was graded on a 5-point scale which

was a slightly modified version of the original AbbreviatedInjury Score (AIS) described by Moore et al.11-13 Injuryseverity ranged from 1 (minor), 2 (moderate), 3 (severe), 4

Fromthe New YorkUniversity School of Medicine,Department of Surgery,

Division of Vascular and Endovascular Surgery.

Competition of interest: none.

Presented at the Thirty-eighth Annual Symposiumof the Society for Clinical

Vascular Surgery, April 7-10, 2010, Scottsdale, Ariz.

Reprint requests: Firas F. Mussa, MD, Division of Vascular and Endovascu-

lar Surgery, New York University Langone Medical Center, 530 First

Avenue, Suite 6F, New York, NY 10016 (e-mail: [email protected]).

The editors and reviewers of this article have no relevant financial relation-

ships 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 2011 by the Society for Vascular Surgery.

doi:10.1016/j.jvs.2010.08.074

359
mailto:[email protected]:[email protected]:[email protected]


2/8

(critical), to 5 (unsurvivable).Outcome was also graded ona 5-point scale from 1 (return to baseline function), 2 (mildfunctional deficit), 3 (moderate functional deficit), 4 (se-

vere functional deficit), to 5 (death). In addition, theRevised Trauma Score (RTS), Trauma ScoreInjury Sever-

ity Score (TRISS), and Acute Physiology and ChronicHealth Evaluation II (APACHE II) score were calculatedbased on admission parameters.

The primary end point was overall mortality. The pre-dicted mortality based on ISS, RTS, TRISS, and APACHEII scores was also compared with actual mortality.

Diagnosis of vascular injuries. Upon admission tothe trauma bay, Advanced Trauma Life Support (ATLS)protocols were observed. All vascular injuries were diag-nosed within the initial hours in the trauma bay or operat-ing room. Clinical suspicion or proximity injuries wereevaluated with computed tomography angiogram (CTA).Blunt injuries with suspicion of vascular injury were evalu-ated with duplex ultrasonography when anatomically feasi-

ble during the day and CTA at night. Hard clinical findingsof vascular injury were evaluated in the operating room(OR) with traditional angiogram or in the interventionalradiology (IR) suite for pelvic injuries. The APACHE II,RTS, and TRISS trauma scores were calculated from lab

values and vital signs upon initial presentation in the traumabay. The ISS system, however, is an anatomic system andthe score calculated after injuries was diagnosed.

Calculation of trauma scores. Calculation of the scor-ing systems has been previously described. Briefly, for the ISS,

AIS from 1 (minor) to 5 (critical) is assigned to the mostsevere injury in each of six anatomic regions (head and neck,face, chest, abdomen, extremity, and external).11-13 If any

injury is deemed unsurvivable, it is assigned a value to 6. Thetop three injuries are squared and summed to calculate theISS, which can range from3 (best) to 75(worst). A score of6in any category automatically gives an ISS of 75.8

The RTS is calculated by assigning a value of 0 (worst) to4 (best) to the GCS, systolic blood pressure (SBP), andrespiratory rate (RR). The score is calculated based on a

weighted formula with GCS carrying the most weight fol-lowed by SBP. The score ranges from 0 (worse) to 7.84 (best)and the resultant survival curve is logarithmic.9 Unlike theother systems, higher RTS scores indicated less severity.

The APACHE II score was calculated by obtaining 14physiologic and laboratory parameters and assigning a value

based on deviation from normal. The values are summed toachieve the final score. The score ranges from 0 (best) to 75(worst) and the resultant survival curve is logarithmic.10

The TRISS score was calculated by utilizing the ISS andRTS scores in a predetermined equation that takes intoaccount the age of the patient and whether the injury wasblunt or penetrating. The resultant score is then used tocalculate the probability of survival.14

Statistical analysis. Means, standard deviations, and/orstandard errors of the mean were calculated. Continuous

variables were analyzed using Studentttest and categoricalvariables were analyzed using 2 with continuity correctionand Fisher exact tests for smaller samples. APvalue less than

.05 was considered significant. Receiver operator characteris-tic curves (ROC) were constructed to test the discriminatoryability of all four trauma scoring systems. The area under thecurve (AUC) or C-statistic was calculated to determine con-cordance or discordance. Statistical analysis was performed

using SPSS statistical software (SPSS, Chicago, Ill).RESULTS

Patient characteristics. During the study period, a to-tal of 2157 patientswere evaluatedby thetrauma service. Fiftypatients (2.3%) with vascular trauma were identified duringthe study period with an additional 50 ISS-matched, nonvas-cular trauma patients as controls. Both groups were similar inage and gender. Mechanism of injury was slightly differentbetween the two groups but not statistically significant. In the

VT group, penetrating and pedestrian struck mechanismsaccounted for38 (75%) of the injuries while themechanism inthe NVT group was more evenly distributed (Table I).

In the VT group, 13 (26%) had more than one vascular

injury and 36 (72%) had an injury to a major vascular struc-ture. Arterial injuries accounted for 42 (82%) of vessels in-

volved. Diagnosis was made on clinical grounds in 20 (40%)and found on angiogram in 23 (46%). The remaining seven(14%) were diagnosed on computed tomography angiogra-phy (CTA) or duplex ultrasound. Thirty-five (70%) patientshad severe to critical injuries posing an imminent danger tothe patients life. Patient characteristics are summarized inTable I.

Vascular surgery consultation. Operative vascularintervention was indicated in 28 (56%) patients. Of those, a

vascular surgeon was involved in 22 (79%). In the remain-ing 6 (21%), the injury was repaired primarily by the trauma

service. Injuries repaired by the trauma service involved theinferior vena cava, portal vein, anterior jugular vein, com-mon carotid artery, and radial artery. Only two patients(7%) were treated with endovascular techniques precludingany meaningful analysis. Both cases were aortic transections

with placement of a thoracic endograft. Furthermore, themortality rate of VT patients requiring operative interven-tion was 28.6% compared with 18.2% for those VT patientsnot requiring an operation (P .512). While the overallrate for a vascular surgery consult was 58% (29 patients),that rate increased as injury severity increased. Vascularsurgery consult was obtained in 3/9 (33%) patients withmoderate, 8/13 (62%) severe, and 18/22 (82%) with

critical injury severities. Those patients with minor or un-survivable injuries did not involve a vascular surgeon.Outcomes and mortality. There was no significant

difference observed in functional outcome between the twogroups at the time of discharge (Table I). There was a trendtoward higher mortality in the VT group compared with theNVT group (24% vs 12%,P .108,Fig 1). The number of

vascular injuries did not appear to impact mortality (single23.08% vs multiple 24.32%, P .6); however, the vesselinvolved, type, and location of the vascular injuries were notsubstratified.

The mortality rates for each of the mechanisms ofinjuries were also examined to exclude the possibility of

JOURNAL OF VASCULAR SURGERY

February 2011360 Loh et al


3/8

increased mortality in the VT trauma as a result of themechanism of injury. In the VT patients, gunshot woundsand the pedestrian struck mechanism were more prevalent

while falls from height and motor vehicle accidents were

found more often in NVT patients. However, the pedes-trian struck and falls from height both carried high mortal-ity rates (26.32% and 26.67%) while gunshot wounds andmotor vehicle accidents carried low mortality rates (7.14%

Table I. Comparison of vascular and nonvascular trauma patients

Vascular trauma Nonvascular trauma Pvalue

Total patients 50 50Age 37.5 17.7 42.8 17.7 .129

Gender .128Male 36 (72%) 43 (85%)Female 14 (28%) 7 (14%)

Mechanism of injury .420Gun shot wound 9 (18%) 5 (10%)Stab wound 11 (22%) 12 (24%)Pedestrian struck 18 (36%) 12 (24%)Motor vehicle crash 2 (4%) 6 (12%)Fall from height 7 (14%) 12 (24%)Blunt trauma 3 (6%) 3 (6%)

Multiple vascular injuriesYes 13 (26%) n/aNo 37 (74%) n/a

Injury locationIntracranial 3 (6%) n/aNeck 4 (8%) n/aChest 5 (10%) n/aAbdomen 13 (26%) n/aPelvis 11 (22%) n/aUpper extremity 6 (12%) n/aLower extremity 8 (16%) n/a

Vessel typeArtery 41 (82%) n/aVein 8 (16%) n/aBoth 1 (2%) n/a

Type of injuryLaceration 21 (42%) n/aTransection 13 (26%) n/aDissection pseudoaneurysm 5 (10%) n/aSpasm 2 (4%) n/aThrombosis 2 (4%) n/aFistula 1 (2%) n/aExternal hematoma 4 (8%) n/a

Compression 2 (4%) n/aSeverity of injury

Minor 4 (8%) n/aModerate 9 (18%) n/aSevere 13 (26%) n/aCritical 22 (44%) n/aUnsurvivable 2 (4%) n/a

Mode of diagnosisClinical 20 (40%) n/aAngiogram 23 (46%) n/aCT scan 5 (10%) n/aDuplex 2 (4%) n/a

Vascular consultYes 29 (58%) n/aNo 21 (42%) n/a

Operation required for vascular injury

Yes 28 (56%) n/aNo 22 (44%) n/a

Outcome .378Baseline function 11 (22%) 14 (28%)Mild functional deficit 17 (34%) 22 (44%)Moderate functional deficit 7 (14%) 4 (8%)Severe functional deficit 3 (6%) 4 (8%)Death 12 (24%) 6 (12%) .108


Volume 53, Number 2 Loh et al 361


4/8

and 12.5%). The prevalence of mechanisms of injury carry-ing high mortality was evenly distributed between VT andNVT patients.

Trauma scoring systems and mortality. The ISS,RTS, APACHE II, and TRISS scores were calculated todetermine the predicted mortality in the presence or ab-sence of VT and to validate their usefulness and accuracy inpredicting mortality.

Injury severity score. By design, the two groups werematched based on ISS; therefore, there was no difference inthe mean ISS between the two groups (24.2 1.8 vs 25.51.9, P .63, Table II). Based on these scores, the composite

predicted mortality, adjusted for age, was 14.3% (VT) and15.6% (NVT). However, the actual mortality of 24% in the

VT group was higher than predicted. There appeared to be atrend toward a lower mean ISS for VT fatalities compared

with NVT (37.83 4.49 vs 47.33 6.68,P .13,Fig 2).For the less severely injured patients (ISS 30, 80% pre-dicted survival), the VT group had a higher mortality but thisdidnotreachstatistical significance (10.8% vs 3%, P .17, Fig3). Interestingly, for those with a predicted mortality greaterthan 15%, VT carried a significantly higher mortality thanNVT (61% vs 29%,P .04).

Revised trauma score. The RTS was similar betweenthe two groups (6.6 0.3 vs 7.0 0.2,P.21,Table II).

The predicted mortality was 4% (VT) and 3% (NVT). In thesubset of patients with greater than 80% predicted survival(RTS 5), the presence of vascular injuries was associated

with significantly higher mortality (26.2% vs 2.2%,P .007,Fig 3). The predicted mortality of 4% was less than the actualmortality of 24% in the VT group. The mean RTS for VTfatalities was significantly more favorable than in the NVTgroup (6.2 0.6 vs 3.8 0.5,P .01,Fig 4)which is anunderestimation of severity of injury in the presence of VT.

APACHE II score. The APACHE II scores weresignificantly different between the VT and NVT groups(12.3 1.3 vs 8.8 1.2, P .05, Table II). This

correlated to a predicted mortality of 15.1% (VT) and 9.6%(NVT). Nevertheless, this was still markedly lower than theactual mortality of 24% in VT. Themean APACHE II scorefor VT fatalities was significantly lower compared with theNVT group (17.0 2.4 vs 25.8 2.2, P .02,Fig 5),

which again underestimates the injury severity in VT pa-tients.For those patients with less severe injuries and pre-dicted survival of greater than 80% (APACHE II 14), theactual mortality was statistically higher in the VT group(18.2% vs 0%,P.02,Fig 3).

Trauma scoreinjury severity score. The TRISSmean calculated probability of survival was similar betweenthe two groups (0.811 0.082 vs 0.842 0.107, P .28,

Fig 1. Overall mortality between vascular and nonvascular trauma.

Table II. Trauma scoring systems in vascular and nonvascular trauma patients with predicted mortality

Vascular trauma Predicted Nonvascular trauma Predicted Mean SEM mortality Mean SEM mortality Pvalue

Injury Severity Score (ISS) 24.22 1.8 14.3% 25.47 1.86 15.6% .630Revised Trauma Score (RTS) 6.64 0.26 4.0% 7.06 0.21 3.0% .209APACHE II score 12.26 1.28 15.1% 8.75 1.18 9.6% .046TRISS 0.811 0.082 18.9% 0.842 0.107 15.8% .283

APACHE II,Acute Physiology and Chronic Health Evaluation II; TRISS, Trauma ScoreInjury Severity Scores.




5/8

Table II). The predicted mortality was 18.9% (VT) and

15.8% (NVT). In the subset of patients with greater than80% predicted survival (TRISS 0.800), the presence of

vascular injuries was associated with significantly higher mor-tality (13.9%vs 0%, P .05, Fig3). Thepredicted mortality of18.9% was less than the actual mortality of 24% in the VTgroup. The mean TRISS for VT fatalities was significantlymore favorable than in the NVT group (0.659 0.096 vs0.240 0.08, P .002, Fig 6), which is an underestimationof severity of injury in the presence of VT.

ROC analysis. An ROC analysis of all four scoringsystems showed that ISS was better than APACHE II andTRISS in predicting mortality in vascular trauma patients(Table III). The RTS fared poorly compared with the other

systems. The AUC of the ISS was 0.870 with a 95% CI of

0.746-0.993. The APACHE II and TRISS scoring systemshad similar AUC of 0.758 (95% CI, 0.626-0.889) and0.786 (95% CI, 0.641-0.932). The RTS had an AUC of0.618 (95% CI, 0.432-0.805).

DISCUSSION

Despite the steady increase in vascular trauma over thepast five decades, major vascular injuries still represent asmall percentage of traumatic injuries1, 2 and little is knownabout the epidemiology of vascular trauma outside of themilitary experience.1,2 Furthermore, vascular injuries leadto increased utilization of medical resources.2 In fact, pa-tients with VT have the highest utilization of medical

Fig 2. Mean injury severity score of fatalities between vascular trauma and nonvascular trauma.

Fig 3. Mortality in patients with greater than 80% predicted survival.




6/8

resources among trauma patients.2 A key limitation topreviously published studies is the lack of comparison withsimilarly injured NVT patients.

Since the initial studies, major advances have beenmade in pre-hospital treatment, rapid hospital transport,triage, and resuscitation.15-17All of these factors have led toimproved survival of patients with vascular injuries.16 Atour institution a vascular surgeon was involved in 75% ofpatients with severe to critical injuries. Vascular surgery wasnot involved in those patients with minor or unsurvivableinjuries. Furthermore, almost 80% of patients requiring

vascular operative intervention involved a vascular surgeon.Future studies will better assess the emerging role of endo-

vascular treatments in trauma patients.Our finding that VT patients have higher mortality rates

compared with NVT (24% vs 12%) is consistent with a priorreport by Galindo et al who showed a mortality of 20.8% in

VT patients compared with 4.5% in NVT patients.18 Thehigher mortality of theNVT group in ourstudy is attributableto the matching of ISS with the VT group. The ISS of theNVTgroup was 8.4in theGalindo study compared with 25.5in this study. Other studies of vascular trauma, such as the

Fig 4. Mean revised trauma score of fatalities between patients with vascular trauma and nonvascular trauma.

Fig 5. Mean APACHE II score of fatalities between patients with vascular trauma and nonvascular trauma.




7/8

large urban study by Mattox in 1998,16 did not utilize acontrol cohort with matched injury severity.

The traditional trauma scoring systems such as the ISS,RTS, and TRISS were designed to predict mortality calcu-lated across the entire spectrum of trauma patients.8,9 Sim-ilarly, the APACHE II score is even broader as a measure ofcritically ill patients regardless of mechanism.10 All foursystems failed to accurately predict mortality in the VT

group (14.3% [ISS], 4% [RTS], 15.1% [APACHE II], and18.9% [TRISS] vs 24% [actual]). The TRISS scoring sys-tems was the most accurate in predicting mortality in the

VT patients but still underestimated the actual mortality.The ISS, APACHE II, and TRISS systems were reasonablyaccurate in predicting mortality in NVT patients (15.6%,9.6%, and 15.8% vs 11.8%). The increased mortality of

vascular trauma patients is likely diluted by its low incidencein the general trauma population. The RTS dismal ability topredict mortality may stem from its design as an immediateclinical assessment tool utilizing parameters that may notmanifest derangement immediately. Interestingly, in lessseverely injured patients (greater than 80% predicted sur-

vival) almost all of the mortalities were in vascular traumapatients. This clearly highlights the fallacies of the scoringsystems causing an underappreciation of severity in VTpatients with a lower calculated trauma score.

The ROC analysis performed on the scoring systemsalso supported the conclusions that these scoring systems

were not ideal to predict mortality in VT patients. The ISSwas the best of the four systems tested with an AUC of0.870, which makes it an adequate, but not superb, trauma

scoring system for VT patients. The TRISS and APACHEII systems had AUCs of 0.786 and 0.758 indicating thatthese studies are not ideal for use in vascular trauma pa-tients. Finally, the RTS had an AUC of 0.618 making it acompletely inadequate scoring system for VT patients.

Closer examination of physiologic scoring systems suchas the RTS, APACHE II, and TRISS reveal that GCScarries more weight in determining the score than any otherfactor. In the RTS system GCS carries 30% more weightthan SBP. In the APACHE II system GCS accounts for16% of the total possible points and combined with age,both make up almost 30%. The heavy weight of neurologicparameters may explain the lack of utility of these systems

when applied to VT patients. Anatomic scoring systemssuch as the ISS may suffer from an inability to recognizeinjury to vascular structures without advance imaging. Inaddition, grading of injury severity is subjective and may

vary based on provider. In all cases, initial assessment maynot be accurate and physiologic parameters specific tohemorrhage should be evaluated. An accurate scoring sys-tem to compare and predict mortality in vascular traumapatients would need to combine physiologic parametersspecific for hemorrhage with a defined anatomic scoringsystem based on vessel injured. The TRISS system com-bines anatomic and physiologic parameters and thereforebetter predicts mortality in VT patients compared with the

Fig 6. Mean TRISS score of fatalities between patients with vascular trauma and nonvascular trauma.

Table III. Receiver operator characteristic curves C-statistic for trauma scoring systems in vascular traumapatients

Trauma scoring systemC-statistic

(AUC)95% confidence

interval

Injury Severity Score (ISS) 0.870 0.746-0.993Revised Trauma Score (RTS) 0.618 0.432-0.805APACHE II score 0.758 0.0626-0.0889

TRISS 0.786 0.641-0.932APACHE II,Acute Physiology and Chronic Health Evaluation II; TRISS,

Trauma ScoreInjury Severity Scores.




8/8

other systems examined. However, TRISS does not weightfactors specific for vascular injuries thus still underestimatesmortality in this patient population.

Endovascular surgery promises the ability to treatvascular injuries with significantly less morbidity and

mortality.

19

In fact, the use of endovascular techniquesto treat VT has increased 2.5 times in the past decade;however, few trauma surgeons possess an advanced cath-eter based skill set.20 Endovascular therapies offergreater versatility and the ability to treat anatomicallydifficult areas (eg, supra-renal aorta, retro-hepatic infe-rior vena cava, high cervical carotid, and vertebral arteryinjuries). The boundaries of endovascular therapy areconstantly moving forward and the reduced morbidityand mortality of endovascular surgery have the potentialto change the overall outcomes and ultimately mortalityof vascular trauma patients.

Limitations. There is no scoring system that is univer-sally accepted; therefore, our decision to use ISS was based

on its design as an anatomic rather than physiologic system.Functional disability, as a result of trauma, is especiallyimportant in young active individuals; however, our study

was not powered to examine functional outcome. Thesmall sample size also limited our ability for subset analysis.In particular, we recognize that injury to particular ana-tomic areas and vessels carries increased severity. Futurestudies should examine larger patient numbers to makemeaningful conclusions on immediate and long-term out-comes in vascular trauma.

CONCLUSION

We examined mortality in vascular trauma patients com-

pared with a matched cohort of nonvascular trauma patients.In patients with similar injury severity, the presence of VT wasassociated with a trend toward increased mortality. Further-more, the ISS, RTS, TRISS, and APACHE II scoring systemsunderestimated mortality when VT was present renderingthem inaccurate when examining VT patients. New scoringsystems should be developed to allow meaningful evaluationof the current care of vascular trauma patients. These patientsshould be critically evaluated expeditiously regardless of cal-culated score. In addition, vascularsurgery should be involvedearlygivenits expertisein both open surgical andendovasculartechniques.

The authors would like to acknowledge Maria McGee

and Sally Jacko, RN for their invaluable contribution indata collection.

AUTHOR CONTRIBUTIONS

Conception and design: SL, FM, HPAnalysis and interpretation: SL, CR, FMData collection: SL, CC

Writing the article: SLCritical revision of the article: SL, CR, TM, MA, NS,

HP, FM

Final approval of the article: SL, CR, CC, TM, MA, NS,HP, FM

Statistical analysis: SL, CRObtained funding: Not applicableOverall responsibility: FM

REFERENCES

1. MacKenzieEJ, Fowler CJ. Epidemiology.In: Moore EE, FelicianoDV,

Mattox KL, editors. Trauma. 5th ed. New York, NY: McGraw-Hill;

2004. p. 28-37.

2. Caps MT. The epidemiology of vascular trauma. Semin Vasc Surg

1998;11:227-31.

3. Asensio JA, Chahwan S, Hanpeter D, Demetriades D, Forno W, Gam-

baro E, et al. Operative management and outcome of 302 abdominal

vascular injuries. Am J Surg 2000;180:528-533; discussion 533-4.

4. Lopez-Viego MA, Snyder WH III, Valentine RJ, Clagett GP. Penetrat-

ing abdominal aortic trauma: a report of 129 cases. J Vasc Surg 1992;

16:332-5; discussion 335-6.

5. Tyburski JG, Wilson RF, Dente C, Steffes C, Carlin AM. Factors

affecting mortality rates in patients with abdominal vascular injuries.

J Trauma 2001;50:1020-6.

6. Feliciano DV. Abdominal vascular trauma. In: Moore EE, Feliciano

DV, Mattox KL, editors. Trauma. 5th ed. New York, NY: McGraw-

Hill; 2004. p. 755-75.

7. Frykberg ER, Schinco MA. Peripheral vascular injury. In: Moore EE,

Feliciano DV, Mattox KL, editors. Trauma. 5th ed. New York, NY:

McGraw-Hill; 2004. p. 969-1002.

8. Baker SP,ONeill B, Haddon W Jr,LongWB. Theinjury severity score:

a method for describing patients with multiple injuries and evaluating

emergency care. J Trauma 1974;14:187-96.

9. Champion HR, Sacco WJ, Copes WS, Gann DS, Gennarelli TA, Flana-

gan ME. A revision of the Trauma Score. J Trauma 1989;29:623-9.

10. Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE II: a

severity of disease classificationsystem.Crit CareMed 1985;13:818-29.

11. Moore EE, Cogbill TH, Jurkovich GJ, McAninch JW, Champion HR,

Gennarelli TA, et al. Organ injury scaling. III: Chest wall, abdominal

vascular, ureter, bladder, and urethra. J Trauma 1992;33:337-9.

12. MooreEE, Malangoni MA,Cogbill TH, Peterson NE,Champion HR,Jurkovich GJ, et al. Organ injury scaling VII: cervical vascular, periph-

eral vascular, adrenal, penis, testis, and scrotum. J Trauma 1996;41:

523-4.

13. Moore EE, Malangoni MA, Cogbill TH, ShackfordSR, ChampionHR,

Jurkovich GJ, et al. Organ injury scaling. IV: thoracic vascular, lung,

cardiac, and diaphragm. J Trauma 1994;36:299-300.

14. Boyd CR, Tolson MA, Copes WS. Evaluating trauma care: the TRISS

method. Trauma Score and the Injury Severity Score. J Trauma 1987;

27:370-8.

15. Bole PV, Purdy RT, Munda RT, Moallem S, Devanesan J, Clauss RH.

Civilian arterial injuries. Ann Surg 1976;183:13-23.

16. MattoxKL, FelicianoDV, BurchJ, Beall ACJr, Jordan GLJr, De Bakey

ME. Five thousand seven hundred sixty cardiovascular injuries in 4459

patients. Epidemiologic evolution 1958 to 1987. Ann Surg 1989;209:

698-705; discussion 706-7.

17. MinowaY, SchreiberR, Moore EE.Regionaltrauma systems. J Trauma1996;40:1054-5.

18. Galindo RM, Workman CR. Vascular trauma at a military level II

trauma center(1). Curr Surg 2000;57:615-8.

19. Hershberger RC, Aulivola B, Murphy M, Luchette FA. Endovascular

grafts for treatment of traumatic injury to the aortic arch and great

vessels. J Trauma 2009;67:660-71.

20. Burkhardt GE, Rasmussen TE, Propper BW, Lopez PL, Gifford SM,

Clouse WD. A national survey of evolving management patterns for

vascular injury. J Surg Educ 2009;66:239-47.

Submitted Apr 25, 2010; accepted Aug 24, 2010.



existing trauma and critical care scoring systems

Documents