The effect of vein repair on the risk of venous thromboembolic events: A review of more than 100 traumatic military venous injuries
Article Outline
Background
The management of venous trauma remains controversial. Critics of venous repair have cited an increased incidence of associated venous thromboembolic events with this management. We analyzed the current treatment of wartime venous injuries in United States military personnel in an effort to answer this question.
Methods
From December 1, 2001, to October 31, 2005, all United States casualties with named venous injuries were evaluated. A retrospective review of a clinical database was performed on demographics, mechanism of injury, associated injuries, treatment, outcomes, and venous thromboembolic events. Data were analyzed using the Fisher exact test, analysis of variance, and logarithmic transformation.
Results
During this 5-year period, 82 patients sustained 103 named venous injuries due to combat operations. All patients were male, with an average age of 27.9 years (range, 20.3-58.3 years). Blast injuries accounted for 54 venous injuries (65.9%), gunshot wounds for 25 (30.5%), and motor vehicle accidents for 3 (3.6%). The venous injury was isolated in 28 patients (34.1%), and 16 (19.5%) had multiple venous injuries. The venous injury in two patients was associated with acute phlegmasia, with fractures in 33 (40.2%), and 22 (28.1%) sustained neurologic deficits. Venous injuries were treated by ligation in 65 patients (63.1%) and by open surgical repair in 38 (36.9%). Postoperative extremity edema occurred in all patients irrespective of method of management. Thrombosis after venous repair occurred in six of the 38 cases (15.8%). Pulmonary emboli developed in three patients, one after open repair and two after ligation (P > .99).
Conclusion
In the largest review of military venous trauma in more than three decades, we found no difference in the incidence of venous thromboembolic complications between venous injuries managed by open repair vs ligation. Blast injuries of the extremities have caused most of the venous injuries. Ligation is the most common modality of treatment in combat zones. Long-term morbidity associated with venous injuries and their management will be assessed in future follow-up studies.
The advantages of venous repair may not be immediately apparent when compared with arterial repair; however, there has been considerable interest in the management of extremity venous injury since the Vietnam War. A number of reports emphasize venous repair over ligation to avoid the potential for early limb loss from venous hypertension or long-term disability from chronic edema. Surgeons deployed to combat zones continue to demonstrate the value of this practice.
Routine venous ligation was the accepted practice of during the first half of the 20th century, but animal and human studies have suggested the possibility of an associated increased morbidity. Ligation of major veins may result in increased venous hypertension, extremity phlegmasia, and compromised arterial flow.1, 2, 3, 4 There is assumed to be an increased rate of complications such as stasis ulceration in patients with prior ligation of a major vein in the extremity, although this is poorly documented. Although durability of complex venous repairs appears to be poor, the hope is that a period of patency will allow for collateral veins to form and may decrease the adverse effects of venous hypertension.3 Despite these theoretic concerns about venous hypertension and long-term sequelae, however, many surgeons continue to advocate ligation.5, 6 A major concern is that repair of venous injuries will result in vein thrombosis and subsequent pulmonary emboli, although support for this scenario is somewhat anecdotal.7
Although ligation had good early success, the evolving knowledge of venous physiology and improving surgical technique stimulated some increased enthusiasm for venous repair by the time of the Korean War.8 During the Vietnam War era, increasing efforts were made to repair disrupted veins, and surgeons often reported the value of these repairs in the conflict. In the four decades since the Vietnam War, civilian surgeons have sporadically reported the use of prosthetic interposition grafts and temporary intravascular shunts in the management of civilian venous injuries.9, 10 During the current wars in Iraq and Afghanistan, also referred to as the Global War on Terrorism (GWOT), management of wartime venous injuries parallels that of the Vietnam experience. About one-third of venous injuries are managed with repair by venorrhaphy or interposition graft.11
The purpose of this study was (1) to describe the rates of venous injuries and the patterns of venous injuries in the United States (US) military personnel in the current wars in Iraq and Afghanistan, (2) to evaluate the contemporary vascular surgical management of wartime venous injuries, and (3) to assess the effects of open surgical repair of traumatic venous injuries vs ligation with respect to the early risk of venous thrombotic events.
Methods
A retrospective review was conducted on a clinically compiled database encompassing all named venous injuries evaluated at the Walter Reed Army Medical Center and National Naval Medical Center from December 2001 to October 2005. The research protocol #06-21005EX was reviewed and granted exemption from a full review by the Institutional Review Board. The evaluation included all patients who were US military casualties with named venous injuries resulting from the wars in Iraq and Afghanistan. Patients treated with endovascular techniques or those who had a venous injury but required immediate limb amputation owing to a nonsalvageable limb were excluded from this analysis. Basic demographic data collected included age, sex, the country where the injury occurred, date of injury, and prior surgical treatment.
Also documented were the specific locations of vascular injury, the type of vessel that was injured, the mechanism of vascular injury, the presence of associated neurologic or orthopedic injuries, limb phlegmasia, and the presence of venous thromboembolism. Named venous injuries were studied with respect to anatomic distribution, association with arterial injury, type of repair performed, and the use of autologous or prosthetic conduit. The initial surgical outcome as well as complications and the need for reintervention were documented. All patients underwent re-evaluation after evacuation to the United States, and undiscovered occult injuries were documented and treated as required.
Patients were clinically evaluated for limb edema and phlegmasia. All patients had radiologic evaluation upon arrival to the United States. Evaluation of the patency of the venous repair was by computed tomography (CT), ultrasonography, or contrast venography, or a combination of these. Clinical suspicion of pulmonary embolus was confirmed with CT pulmonary arteriography.
Patients were divided into three groups according to surgical therapy: treatment by ligation only, venous repair only, and both ligation and venous repair. The percentage of the clinical outcomes of pulmonary embolism, thrombosis of the injured vein after surgical repair, phlegmasia, and deep vein thrombosis (DVT) are presented with a 95% confidence interval for each treatment group. The presence of a DVT was defined as a thrombosis of a vein that was not involved in a ligation or repair. Venous thrombosis that occurred at the site of venous ligation as a result of the treatment of the vein injury was not categorized as a DVT. Thrombosis at the site of venous repair was considered to be a thrombosis complication after surgical repair.
As appropriate, demographic, injury, and clinical data, and outcome variables of the treatment group were compared using the Fisher exact test, analysis of variance, and logarithmic transformation. Logistic regression analysis was performed for the outcome variables to identifying possible risk factors. The analysis was done with SPSS 15.0 software (SPSS Inc, Chicago, Ill). A value of P ≤ .05 was considered to be statistically significant.
Results
Demographics
During the 5-year period studied, 82 patients (all men) sustained 103 named venous injuries due to combat operations. Their average age was 27.9 ± 8.1 years (range, 20.3-58.3 years). The mean time of arrival to our stateside tertiary medical center was 5.0 ± 2.5 days (range, 2-19 days). Treatment was by ligation only in 48 patients with 59 venous injuries, surgical repair in 29 patients with 32 venous injuries, and a combination of ligation and repair in five patients with 12 venous injuries. There was no difference in the mean age or transfer times amongst these three groups (Table I).
Table I. Demographics, injury, and treatment variables of patients by each treatment group
| Variablea | All patients | Ligation | Repair | Ligation + repair | P |
|---|---|---|---|---|---|
| Patients, No. | 82 | 48 | 29 | 5 | |
| Age, years | 27.9 ±8.1 | 27.7±8.4 | 28.7±7.8 | 25.2±6.5 | .65b |
| Transfer time, days | 5.0±2.5 | 5.2±2.9 | 4.8±1.8 | 4.6±1.9 | .68b |
| Anticoagulation, days | 69±58 | 58±48 | 88±68 | 67±58 | .36c |
| Mechanism of injury | .25d | ||||
| Blast | 54(65.9) | 29(60.4) | 22(75.9) | 3(60.0) | |
| Gunshot wound | 25(30.5) | 18(37.5) | 5(17.2) | 2(40.0) | |
| MVA | 3(3.6) | 1(2.1) | 2(6.9) | 0(0) | |
| Associated injuries | |||||
| Arterial injury | 54(65.9) | 34(70.8) | 17(58.6) | 3(60.0) | .55e |
| Boney fractures | 33(40.2) | 17(35.4) | 12(41.4) | 4(80.0) | .17e |
| Nerve injury | 23(28.1) | 14(29.2) | 7(24.1) | 2(40.0) | .72e |
| Adjunctive treatments | |||||
| Arterial shunts | 8(9.9) | 6(12.5) | 2(7.1) | 0(0) | .83e |
| Fasciotomies | 41(50.0) | 27(56.3) | 12(41.4) | 2(40.0) | .46e |
| Associated infections | |||||
| Wound infections | 56(68.3) | 32(66.7) | 19(65.6) | 5(100) | .38e |
| Bacteremia | 40(48.8) | 19(39.6) | 18(62.1) | 3(60) | .14e |
aContinuous data are presented as mean ± standard deviation; categoric data are presented as number (%). |
bAnalysis of variance. |
cLog transformation before analysis of variance to adjust for heterogenous variances. |
dFisher exact test for a 3 × 3 cross tabulation. |
eFisher exact test. |
Mechanism of injury
Blast injuries accounted for 54 of the 82 injured patients (65.9%), high-velocity gunshot wounds (GSW) occurred in 25 (30.5%), and three (3.5%) were involved in motor vehicle rollovers while engaged in combat. These vehicles were struck by roadside blast. The mechanism of injury for each treatment group, which is summarized in Table I, was not statistically different for the three surgical treatment groups (P = .25).
Injury pattern
Concomitant venous and arterial injury was present in 54 of the 82 patients (65.9%), 28 (34.1%) had isolated venous injury, and 16 (19.5%) had multiple venous injuries. As summarized in Table II, 69 of the 103 venous injuries (67.0%) occurred in the extremities: 59 (57.3%) were lower extremity venous injuries, with 24 injuries involving the femoral vein, and 10 (9.7%) were upper extremity venous injuries (4 axillary, 6 brachial). There were 19 (18.4%) truncal and 15 (14.6%) cervical venous injuries. The iliac vein was the most common truncal venous injury, including 12 iliac, 3 subclavian, 2 inferior vena cava, 1 hepatic, and 1 adrenal. In the neck, there were 11 internal jugular and four external jugular vein injuries.
Table II. Location of all venous injuries and type of surgical treatment performed
| Location/Vein | Total | Ligation | Primarya | Patchb | AVGc |
|---|---|---|---|---|---|
| Patients, No. | 103 | 65 | 24 | 3 | 11 |
| Neck | |||||
| External jugular | 4 | 4 | |||
| Internal jugular | 11 | 8 | 3 | ||
| Trunk | |||||
| Subclavian | 3 | 3 | |||
| Hepatic | 1 | 1 | |||
| Adrenal | 1 | 1 | |||
| Inferior vena cava | 2 | 1 | 1 | ||
| Iliac | 12 | 8 | 4 | ||
| Upper extremity | |||||
| Axillary | 4 | 2 | 1 | 1 | |
| Brachial | 6 | 5 | 1 | ||
| Lower extremity | |||||
| Great saphenous | 4 | 3 | 1 | ||
| Common femoral | 6 | 2 | 2 | 1 | 1 |
| Profunda femoral | 8 | 7 | 1 | ||
| Femoral | 24 | 9 | 7 | 1 | 7 |
| Popliteal | 13 | 8 | 2 | 1 | 2 |
| Tibial | 4 | 4 |
aPrimary repair or lateral venorrhaphy. |
bAutogenous vein patch repair. |
cAutogenous vein graft. |
Associated injuries
Owing to the nature of the high-velocity gunshots and high-explosive blast munitions, all patients had concomitant soft tissue injury, and 54 of 82 (65.9%) had combined venous and arterial injuries. Although patients treated in the ligation-only group were more likely to have concomitant arterial injury, it was not statistically different from those in the other two surgical treatment groups, as listed in Table I. Fractures were associated with venous injuries in 33 patients (40.2%). Venous injuries associated with fractures were most commonly in the lower extremities, occurring in 25 of 45 leg injuries (55.6%), which was significantly higher (P = .005) compared with the fracture rate in the neck (7.7%, 1 of 13), trunk (41.7%, 5 of 12), or upper extremities (11.1%, 1 of 9). Neurologic injuries were seen in 23 patients (28.1%). In the upper extremity, 77.8% of patients (7 of 9) with venous injuries had neurologic injuries, which was significant (P = .015) compared with a neurologic injury rate of 21.1% in the neck (3 of 13), 16.7%, in the trunk (2 of 12), and 22.2% in the lower extremities (10 of 45).
Anticoagulation therapy
All but two of the patients received prophylactic low-molecular-weight heparin anticoagulation with enoxaparin (30-mg subcutaneous injection twice daily), starting the day after injury and continuing until discharge from the hospital. Two patients were treated with sequential compression devices only: one with an intracranial hemorrhage and a single common femoral vein ligation and the other with a retroperitoneal hemorrhage and a single common femoral vein lateral venorrhaphy. Fifteen patients received 6 months of anticoagulation with warfarin for DVT treatment or after complications of vein graft thrombosis. The average length of anticoagulation for all patients was 69 ± 58 days. A log regression analysis showed no difference in the duration on anticoagulation therapy amongst the three surgical treatment groups (P = .36; Table I).
Surgical management
Eighty-two patients sustained 103 venous injuries, of which 65 injuries (63.1%) were treated with ligation and 38 (36.9%) were repaired. Two-thirds of the veins injuries that were repaired were primary repairs, consisting of lateral venorrhaphy in 24, vein patch in three, and vein graft in 11. With the exception of one axillary vein graft repair, all other vein patch or autogenous venous graft repairs were performed in the lower extremities (Table II).
Adjunctive treatments
Eight of the 54 patients with combined arterial and venous injury received temporary arterial shunts and one venous shunt. No arterial shunts were used in patients treated with combined venous ligation and repair (Table I). Forty-one extremity fasciotomies were performed, with arm injuries accounting for 9.8% (4 of 41) and leg injuries for 90.2% (37 of 41). However, the use of fasciotomies was not significant, with 40.5% in the repair-only group vs 56.3% in the ligation-only group (P = .46).
Associated infections
Positive wound cultures were found in 68.3% (56 of 82). All patients treated with combined ligation and repair had positive wound cultures, although this was not statistically significant different than in the ligation-only (66.7%) and the repair-only (65.6%) groups (P = .37). Bacteremia occurred in 58.8% (40 of 82) patients. However, patients in the repair-only group had the highest incidence of bacteremia at 62.1%, compared with ligation only at 39.6% and combined ligation and repair at 60%. Again, there was no difference in the incidence of amongst treatment groups (P = .14). Of the 40 patients with bacteremia, 34 (85.0%) had associated positive wound cultures. All patients were treated with the appropriate intravenous antibiotics for 14 days or until the results of repeat blood cultures were negative.
Outcome
Deep vein thrombosis, unrelated to the site of surgical repair, was found in 12.2% (10 of 82) of patients (Table III). All DVT occurred after surgical treatment, and the diagnosis was confirmed with imaging by duplex ultrasound or contrast CT. Deep vein thrombosis occurred in 6.9% (2 of 29) patients in the repair group, in 14.6% (7 of 48) in the ligation patients, and in 20.0% (1 of 5) of patients in the combined ligation and repair group, although this difference was not statically significant (P = .37).
Table III. Outcomes of each treatment group with respect to complications
| Variable | All patients (N = 82) | Ligation (n = 48) | Repair (n = 29) | Ligation and repair (n = 5) | P |
|---|---|---|---|---|---|
| % (95% CI) | % (95% CI) | % (95% CI) | % (95% CI) | ||
| Deep vein thrombosis | 12.2(6.0–21.2) | 14.6(6.1–27.8) | 6.9(0.9–22.8) | 20.0(0.5–71.6) | .37a |
| Phlegmasia | 2.4(0.3–8.5) | 2.1(0.1–8.5) | 0(0–9.8) | 20.0(0–45.1) | .12a |
| Thrombosis after repair | NA | NA | 20.7(8.0–39.7) | 0(0–45.1) | .56a |
| Pulmonary embolism | 3.7(0.8–10.3) | 4.2(0.1–11.1) | 3.4(0–9.8) | 0(0.5–71.6) | >.99 |
aFisher exact test. |
All patients with extremity injury had edema of the affected limb; however, phlegmasia developed after vein ligation in 2.4% (2 of 82) of patients, each with gunshot injuries of the common femoral vein. Phlegmasia did not develop in patients treated by vein repair. One patient was treated with a femoral vein-to-common femoral vein interposition graft and the other patient with a great saphenous vein transposition (Fig). Although the incidence of phlegmasia was 0% to 20% amongst the three treatment groups, no group had a statistically significant higher risk of phlegmasia (P = .12).
Figure.
Great saphenous vein transposition onto the common femoral vein for the treatment of left leg phlegmasia. This patient had a common femoral vein injury from a gunshot wound that was initially ligated.
In six patients with blast injuries and each with a single venous injury, thrombosis occurred after repair, but phlegmasia did not develop. This included primary repair (1 common femoral vein, 2 femoral veins, and 1 popliteal vein) and interposition graft repair (1 femoral vein and 1 popliteal vein). Thrombosis occurred in six of the 38 vein repairs (15.8%), all of whom were patients who were treated by repair only. This complication did not develop in any patients in the combined ligation and repair group. However, the Fisher exact test showed no increased risk between the two groups (P = .56), as summarized in Table III.
Pulmonary emboli (PE) developed in three patients: one patient after open repair of a SFV repair with an interposition vein graft and in two patients after ligation of an injured vein, one patient from a brachial vein ligation and the other from an iliac vein ligation. Of the 82 patients, 29 patients underwent venous repair only, with PE developing in one. In comparison, 48 patients were treated with ligation only, with PE developing in two. Five patients had combined ligation and repair of their multiple venous injuries, but no PE developed. When the vein repair-only group was compared with the vein ligation-only group, no significant difference was found for PE (3.4% vs 4.2%, P > .99), as shown in Table III.
Discussion
The management of venous trauma remains topic of considerable interest. The first recorded lateral venorrhaphy performed in humans was by Schede for the treatment of venous trauma back in the 1870s.12 In 1879, the Russian surgeon Eck, performed the first anastomosis between two blood vessels, a side-to-side connection between the portal vein and the vena cava.13 Subsequently in the 1880s, Kummel and Dorfler performed the first end-to-end venous repair in humans.14 It was not until the second half of the 20th century that autogenous venous interposition grafts and panel or spiral composite vein grafts were used in the treatment of venous trauma.15, 16 However, despite these advancements in repair of venous injuries, vein repair was not common during any war of the 20th century.
Venous ligation is considered the most common modality of treatment in the patient in extremis. Ligation is expedient and allows the surgeon to address other injuries in a patient likely to be hypothermic, coagulopathic, and acidotic. Advocates for ligation cited concern for DVT and subsequent PE in repaired veins, although these complications by far are not documented.17 Venous ligation after trauma is associated with increased acute venous hypertension, however, and may contribute to an increased early amputation rate.8 The long-term problems are associated with chronic venous insufficiency.
Repair of venous injuries caused by trauma was established during the Korean War. During the Vietnam War, venous repair became popularized but constituted only one-third of all venous injuries reported by Rich et al.18 In a study on the management of 110 popliteal vein injuries without associated arterial injury, thrombophlebitis and PE were not significantly increased with venous repair. The incidence of lower extremity edema was 50.9% after ligation vs 13.2% after repair.19 More than three decades later, in the current wars in Iraq and Afghanistan, the rate of traumatic venous repair remains similar to the Vietnam experience. In this series, 39% (42 of 106) of venous injuries were repaired.
Many studies evaluate for incidence of lower extremity edema after venous repair vs ligation as an end point. In our study, all patients with lower extremity injury had edema of varying degrees, but it was difficult to quantify the degree of edema given that 79.2% of lower extremity injuries had fasciotomies and 38% had associated fractures. In addition, many had soft tissue injury or loss for which we did not quantify in this study.
Phlegmasia occurred in two patients who both had combined femoral and profunda femoral vein injuries that were initially ligated. The phlegmasia resolved after venous reconstruction; however, the chronic edematous changes persisted, and the patients continue to have edema. When we look at watershed areas for phlegmasia, this usually involves injuries to the iliofemoral vein segments, including the external iliac, common femoral, popliteal veins, and combined femoral and profunda femoral vein injuries. There were five common femoral vein injuries (one patient treated with ligation) and six combined femoral vein and profunda femoral vein injuries (three patients with both veins ligated). The incidence of phlegmasia was 50% (2 of 4) of those who were treated initially with ligation. In contrast with civilian venous trauma, Timberlake and Kerstein20 reported 29 femoral veins treated with ligation in 45 patients, with no evidence of phlegmasia or long-term edema at a mean follow-up of 33.6 months.
The patency of venous repair, especially complex venous reconstruction, has been a concern amongst surgeons. Studies of short-term patency have been published, but long-term studies are few and usually involve only a small cohort:
Most patients had wound infections associated with the site of their venous injuries. Nearly half also had an associated bacteremia. Infection is a known risk factor for venous thromboembolism.24 Bacterial toxins and inflammatory soluble mediators (thrombin and tissue factors) contribute to activated blood coagulation.25 Several case reports and small case series have described venous thrombosis in association with musculoskeletal infections, especially due to Staphylococcus aureus.26, 27, 28 Neither wound infection nor bacteremia appears to be an independent risk factor for thromboembolism, however. Patients in our series who were treated by ligation or venous repair did not have significantly different infection rates.
Deep vein thrombosis and PE are common sequelae after major trauma. Patient with major trauma such as head injuries, abdominal injuries, vascular injuries, extremity, pelvic fractures, and spinal cord injuries are in a hypercoagulable state due to a systemic response to trauma.29 The incidence of venous thromboembolism can be as high as 58% to 63%, in the absence of thromboprophylaxis.30 A large, randomized, controlled trial of 344 patients with major trauma without overt bleeding or intracranial injuries comparing low-dose unfractionated heparin (UNF) with low-molecular-weight heparin (LMWH) showed that LMWH lowered the overall risk of thromboembolism to 30% vs 41% with low-dose UNF (P = .014).31 In our series, the overall risk of thromboembolism (DVT, thrombosis after vein repair, and PE) was 23.2% (19 of 82); this is commensurate with other reports.
Pulmonary embolus associated with DVT in trauma is well documented,32, 33, 34, 35 and occurs in 2% to 22% of trauma patients.36, 37, 38 Several studies suggested that major venous injury or venous repair results in an increased risk of DVT.39, 40 More important, there is risk of PE as a complication of DVT.35 Few studies have examined the risk of PE associated with venous repair in civilian trauma. Sue et al41 reported 12 patients with iliac or common femoral venous injuries, of which seven underwent primary repair. The DVT/PE complication rate in this small group was at least 43% (3 of 7), but no DVT/PE occurred amongst those who were treated with venous ligation.41 The studies that have evaluated civilian trauma mostly involved a small number of patients.
To our knowledge, no studies have been done on the risk of PE in venous repair after military trauma. In this study, the overall incidence of PE is three in 85 patients (3.5%), and three in 106 vein injuries (2.8%) amongst all venous injuries. The risk of PE in venous repairs was 3 in 39 vein repairs (7.7%). Of the 40 iliofemoral venous injuries, there were 19 repairs with one PE (5.3%) and 21 ligations with one PE (4.8%). The third PE was in a patient with a ligated brachial vein. Thus in our series, the risk of PE in venous repair is low compared with the small series of civilian venous trauma.
This study contained inherent limitations. Although to our knowledge this is the largest series on venous injuries in US military personnel since the Vietnam War, the number of sample size is still too small with insufficient number of cases to provide statistical power for comparing the effects of different surgical treatments on the risk of DVT, phlegmasia, PE, or thrombosis after repair. Advanced statistical procedures such as multiple logistic regressions were not performed to identify the risk factors associated with the above outcome variables because few cases were available for the analysis. Other associated risk factors, such as lymphedema, injury severity scores, blood transfusion, and length of surgery and numbers of procedures, were not accounted for. Intermediate and long-term follow-up is currently not available. However, the results may provide useful information to contribute to the design of future larger multi-center study.
Conclusion
The management of vein repair vs ligation for traumatic venous injury remains a controversy. In an ideal setting, venous injuries should be repaired when possible and tolerated by the patient. In a watershed area, such an iliofemoral vein segment, popliteal, and internal jugular vein injury with an absent contralateral internal jugular, repair is especially encouraged to ameliorate the high risk of leg phlegmasia or facial edema. The use of interposition vein graft is effective in venous injuries caused by high-energy trauma, and the risk of graft thrombosis is similar to that of civilian low-velocity injuries. For the patient in extremis, ligation is the preferred modality. If ligation is performed, prophylactic fasciotomies are required to prevent phlegmasia. Although several prior small series studies have suggest that venous repair has a high association with PE, we have found this to be the contrary; in fact, the risk of PE is low and the risk of repair is equivalent to that of ligation.
Author contributions
References
- . Extremity venous ligation: clinical and hemodynamic correlation. Ann Surg. 1975;41:203–208
- . Management of venous trauma. Surg Clin North Am. 1988;68:809–821
- Femoral venous trauma: techniques for surgical management. Am J Surg. 1983;146:220–224
- . Importance of venous occlusion in arterial repair failure: an experimental study. Ann Surg. 1972;175:223–227
- . The natural history following venous ligation for civilian injuries. J Trauma. 1982;22:827–832
- . Venous ligation of the lower extremities and pelvis: repair versus ligation. J Trauma. 1992;33:532–538
- . The natural history of extremity venous repair performed for trauma. Am Surg. 1999;65:116–120
- . Vascular trauma. In: 2 edition. New York, NY: McGraw-Hill; 2004;p. 309–404
- . On the vascular lesion produced by gunshot injuries and their results. Br J Surg. 1916;3:353–421
- . L’hemostase d’urgence en chirurgie de guerre. Helvet Med Acta. 1941;8:3–21
- . Acute vascular trauma in Korea war casualties: an analysis of 180 cases. Surg Gyn Obstet. 1954;99:91–101
- . Historical aspects of direct venous reconstruction. In: Bergen JJ, Yao JTS editor. Symposium on venous problems in Honor of Geza de Takats. Chicago, IL: Year Book Medical Publishers Inc; 1977;p. 50
- . Blood vessel surgery and its applications. (A reprint) In: Harbison SP, Fisher B editor. Pittsburgh, PA: Univ of Pittsburgh Press; 1959;p. 2
- . History of arterial grafting. J Cardiovasc Surg. 1963;4:152
- . Autogenous venous interposition grafts in repair of major vein injuries. J Trauma. 1977;17:512–520
- . Principles and indications for venous repair. Surgery. 1982;91:492–496
- . Long-term results of lower-extremity venous injuries. Arch Surg. 1997;132:963–968
- . Management of venous injuries. Ann Surg. 1970;171:724–730
- . The effect of acute popliteal venous interruption. Ann Surg. 1976;183:365–368
- . Venous injury: to repair or ligate, the dilemma revisited. Am Surg. 1995;61:139–145
- The early fate of venous repair after civilian vascular trauma. Ann Surg. 1987;206:458–464
- Outcomes of venous reconstruction in patients with trauma. J Vasc Surg. 1997;25:398–404
- Management and short-term patency of lower extremity venous injuries of various repairs. Am J Surg. 2003;186:631–635
- . Coagulation in sepsis. Int Care Med. 2004;30:1032–1040
- . Thrombin/inflammation paradigms: A closer look at arterial and venous thrombosis. Am Heart J. 2005;149:S19–S31
- . Venous thrombosis and thromboembolism in children with osteomyelitis. J Pediatr. 2006;149:537–541
- . Deep venous thrombosis associated in pediatric musculoskeletal sepsis. J Pediatr Ortop. 2002;22:329–332
- . Staphylococcus aureus infections. N Engl J Med. 1998;339:520–532
- . Venous thromboembolism in trauma: a local manifestation of systemic hypercoagulability?. J Trauma. 2003;54:224–231
- . A prospective study of venous thromboembolism after major trauma. N Engl J Med. 1994;331:1601–1606
- . A comparison of low-dose heparin with low molecular weight heparin as prophylaxis against venous thromboembolism after major trauma. N Engl J Med. 1996;335:701–707
- . The iliofemoral venous segment as a source of pulmonary emboli. Lancet. 1967;871:4
- . Pulmonary embolism in acute liofemoral venous thrombosis. Br J Surg. 1985;72:912–915
- . Free floating iliofemoral thrombosis: a risk of pulmonary embolism. Arch Surg. 1985;120:806–808
- Unexpected high prevalence of silent pulmonary embolism in patients with deep venous thrombosis. Chest. 1989;95:498–502
- . Deep vein thrombosis and pulmonary embolism in trauma patients. J Intensive Care Med. 1988;3:87–98
- . Pulmonary embolism in major trauma patients. J Trauma. 1990;30:748–750
- . The efficacy of of sequential compression devices in multiple trauma patients with severe head injuries. J Trauma. 1994;37:205–208
- . Venous thromboembolism in patients with major trauma. Am J Surg. 1990;150:365–369
- . Thromboembolism following multiple trauma. J Trauma. 1992;32:2–11
- . Iliofemoral venous injuries: an indication for prophylactic caval filter placement. J Trauma. 1995;39:693–695
Competition of interest: none.
PII: S0741-5214(07)01762-4
doi:10.1016/j.jvs.2007.10.056
© 2008 The Society for Vascular Surgery. Published by Elsevier Inc. All rights reserved.
