Search

Background: Duodenal leak is a feared complication of repair, and innovative complex repairs with adjunctive measures (CRAM) were developed to decrease both leak occurrence and severity when leaks occur. Data on the association of CRAM and duodenal leak are sparse, and its impact on duodenal leak outcomes is nonexistent. We hypothesized that primary repair alone (PRA) would be associated with decreased duodenal leak rates; however, CRAM would be associated with improved recovery and outcomes when leaks do occur., Methods: A retrospective, multicenter analysis from 35 Level 1 trauma centers included patients older than 14 years with operative, traumatic duodenal injuries (January 2010 to December 2020). The study sample compared duodenal operative repair strategy: PRA versus CRAM (any repair plus pyloric exclusion, gastrojejunostomy, triple tube drainage, duodenectomy)., Results: The sample (N = 861) was primarily young (33 years) men (84%) with penetrating injuries (77%); 523 underwent PRA and 338 underwent CRAM. Complex repairs with adjunctive measures were more critically injured than PRA and had higher leak rates (CRAM 21% vs. PRA 8%, p < 0.001). Adverse outcomes were more common after CRAM with more interventional radiology drains, prolonged nothing by mouth and length of stay, greater mortality, and more readmissions than PRA (all p < 0.05). Importantly, CRAM had no positive impact on leak recovery; there was no difference in number of operations, drain duration, nothing by mouth duration, need for interventional radiology drainage, hospital length of stay, or mortality between PRA leak versus CRAM leak patients (all p > 0.05). Furthermore, CRAM leaks had longer antibiotic duration, more gastrointestinal complications, and longer duration until leak resolution (all p < 0.05). Primary repair alone was associated with 60% lower odds of leak, whereas injury grades II to IV, damage control, and body mass index had higher odds of leak (all p < 0.05). There were no leaks among patients with grades IV and V injuries repaired by PRA., Conclusion: Complex repairs with adjunctive measures did not prevent duodenal leaks and, moreover, did not reduce adverse sequelae when leaks did occur. Our results suggest that CRAM is not a protective operative duodenal repair strategy, and PRA should be pursued for all injury grades when feasible., Level of Evidence: Therapeutic/Care Management; Level IV., (Copyright © 2023 Wolters Kluwer Health, Inc. All rights reserved.)

Background: The American College of Surgeons National Surgical Quality Improvement Program collects information related to procedures in the form of the work relative value unit (RVU) and current procedural terminology (CPT) code. We propose and evaluate a fully automated nonparametric learning approach that maps individual CPT codes to perioperative risk., Study Design: National Surgical Quality Improvement Program participant use file data for 2005-2006 were used to develop 2 separate support vector machines (SVMs) to learn the relationship between CPT codes and 30-day mortality or morbidity. SVM parameters were determined using cross-validation. SVMs were evaluated on participant use file data for 2007 and 2008. Areas under the receiver operating characteristic curve (AUROCs) were each compared with the respective AUROCs for work RVU and for standard CPT categories. We then compared the AUROCs for multivariable models, including preoperative variables, RVU, and CPT categories, with and without the SVM operation scores., Results: SVM operation scores had AUROCs between 0.798 and 0.822 for mortality and between 0.745 and 0.758 for morbidity on the participant use file used for both training (2005-2006) and testing (2007 and 2008). This was consistently higher than the AUROCs for both RVU and standard CPT categories (p < 0.001). AUROCs of multivariable models were higher for 30-day mortality and morbidity when SVM operation scores were included. This difference was not significant for mortality but statistically significant, although small, for morbidity., Conclusions: Nonparametric methods from artificial intelligence can translate CPT codes to aid in the assessment of perioperative risk. This approach is fully automated and can complement the use of work RVU or traditional CPT categories in multivariable risk adjustment models like the National Surgical Quality Improvement Program., (Copyright © 2011 American College of Surgeons. Published by Elsevier Inc. All rights reserved.)

The National Surgical Quality Improvement Program (NSQIP), as administered by the American College of Surgeons, became available to private sector hospitals across the United States in 2004. The program works to improve surgical outcomes by providing high-quality, risk-adjusted data to surgeons at a given hospital to stimulate discussion and define target areas for improvement. Although the NSQIP began in the early 1990s with Veterans Administration hospitals and expanded to private sector hospitals nearly 5 years ago, the "how to" process for NSQIP implementation has been left to individual institutions to manage on their own. The NSQIP was instituted at a large tertiary hospital in 2005, identifying through experience 12 critical steps to help surgeons and hospitals implement the NSQIP.

Background: Few good surgical options exist for the repair of complex anterior abdominal wall defects, particularly those in which bacterial contamination is present. The use of prosthetic mesh increases complication rates when the mesh is placed directly over viscera or when the surgical site is contaminated from a pre-existing infection or enteric spillage. The use of an acellular dermal matrix (ADM), which becomes vascularized and remodeled into autologous tissue after implantation, may represent a low-morbidity alternative to prosthetic mesh products in these complex settings. This study examined our experience with ADM in the reconstruction of contaminated abdominal wall defects., Methods: Patients undergoing abdominal wall reconstructions in the face of contamination with ADM between May 2002 and December 2005 underwent retrospective chart review. Demographics, indications for ADM placement, plane of implantation, complications, and follow-up data were evaluated., Results: Sixty-seven patients were identified. The indications for ADM placement included incarcerated hernias, infected mesh, fistulae, early/delayed abdominal wall reconstruction after intra-abdominal catastrophe or trauma, dehiscence/evisceration, and spillage of enteric contents. The ADM was positioned either above the fascia or beneath the fascia or was sutured directly to the fascial edges. Sixteen patients developed a wound infection; the majority of these were superficial and required only local wound care, 5 required some further surgical intervention, and 2 required removal of the ADM. Twelve patients developed recurrent hernias. The mean follow-up time for the study population was 10.6 months., Conclusions: ADM can be used safely and effectively as an alternative to traditional mesh products for abdominal wall reconstructions, even in the setting of contaminated fields.

We undertook this retrospective review to examine the appropriateness of a protocol for the selective emergency department (ED) workup of asymptomatic penetrating truncal injuries. Records of consecutive patients presenting to our urban Level I trauma center with penetrating truncal injuries between January 1, 1997 and September 2000 were reviewed. Data obtained included: patient demographics, ED workup, ED disposition, complications, and follow-up. Selective ED workup included hospital triple-contrast CT, admission for observation, and local wound exploration for selected anterior abdominal stab wounds. Four hundred fifty-five patients presented with penetrating truncal wounds during the study period. One hundred ninety-four patients were taken directly to the operating room, 136 were discharged based solely on physical examination and plain radiographs, 18 were admitted for observation without ED workup, and 107 had selective ED workup. Sixty-two patients (58% of those selectively worked up) were discharged home after negative ED workup, 18 were managed operatively, and 27 were managed nonoperatively. There were two missed injuries that were later identified and managed with no complications. Follow-up was available on 66 per cent of ED workup patients (range 1-42 months). We conclude that selective management of certain penetrating truncal injuries appears appropriate. Patients having a negative selective ED workup can be safely discharged thereby avoiding the cost and resource utilization associated with hospital admission.

With decreasing violent crime and an increase in the use of nonoperative management techniques the viability of urban trauma centers has come into question. In addition the workload and productivity for surgeons at such centers may be threatened. The current study examines the changing characteristics of patients admitted to an urban Level I trauma center over a 5-year period and examines factors that may affect trauma surgeon utilization. We reviewed all trauma registry admissions from January 1995 through December 1999. Data were collected regarding patient demographics, mechanism of injury, diagnostic workup, injury character and severity, operative procedures, intensive care unit (ICU) length of stay (LOS), hospital LOS, and patient disposition. Admissions declined 23 per cent over the 5-year period. Fewer patients were admitted to general practice units whereas more patients required ICU admission. Over the study period both mean patient age and mean Injury Severity Score increased significantly. Gunshot wound admissions declined by 45 per cent, but the percentage of those admitted who required operation rose 17 per cent. Number of operations for trauma performed by general surgeons was unchanged over time. Hospital LOS declined over time, and ICU LOS was unchanged. Although trauma center admissions--particularly those due to violent crime--are on the decline the operative productivity of trauma surgeons has remained unchanged. Patients admitted to the hospital are older and more severely injured; they undeniably require a higher level of care and service coordination. Urban trauma centers remain viable and are in fact more efficient in caring for sicker patients.

The debate over the use of diagnostic angiography (DA) to exclude arterial injury in penetrating extremity trauma (PET) continues. This review evaluates our current protocol for PET and identifies indications for DA. Patients presenting to our urban Level I trauma center between January 1997 and September 2000 with PET were included. Demographic data, emergency department (ED) course, and patient follow-up were reviewed. ED evaluation directed by physical examination (PE) included Doppler pressure indices (DPI) and DA if indicated. A total of 538 patients had PET injuries. Twenty (4%) patients with hard signs of vascular injury were taken to the operating room. Ninety-one (17%) patients without vascular compromise underwent operative procedures or were admitted for other injuries. One hundred twenty-three (23%) patients with nonproximity wounds were discharged. Four DAs were performed for abnormal DPI with no change in management. Three hundred patients with a negative PE and normal DPI were discharged from the ED. Follow-up was available on 51 per cent of these patients (range 1-49 months) with no missed injuries identified. We conclude that PE with DPI is an appropriate way to identify significant vascular injuries from PET. Patients with normal PE and DPI can be safely discharged. DA is only indicated for asymptomatic patients with abnormal DPI.

Between December 1, 1994 and April 1,1998, 44 thoracoscopic procedures were performed in 42 patients following chest injuries. Indications included exploration in 15, retained haemothorax in 10, continued bleeding after chest tube placement in 3, air leak in 5 and empyema in 11. Video thoracoscopy was used in 24 cases and rigid thoracoscopy in 20, including 14 patients in whom video thoracoscopy was contraindicated. There was no difference in the operative times, length of stay or incidence of complications. Two formal and 3 "mini" thoracotomies were used in the video thoracoscopy group. Three "mini" thoracotomies were required in the rigid thoracoscopy group. Rigid thoracoscopy is an effective tool that, in selected cases, increases the utility of thoracoscopy in the management of chest trauma and its complications.

Background: The goals were (1) to characterize physiologic changes after a combined insult of hemorrhage plus sepsis in a large animal model and (2) to determine whether transient inhibition of the neutrophil CD18 adherence receptor during fluid resuscitation impairs host defense during recovery from this injury., Methods: Two series of experiments were performed in anesthetized swine. In the first series (n = 22), the cecum was ligated and incised immediately before 35% hemorrhage. After 1 hour, shed blood plus supplemental fluid was administered to restore and stabilize hemodynamics. On the basis of these results, a second series examined effects of anti-CD18 (2 mg/kg R15.7; n = 9) or its saline placebo (n = 10) administered during fluid resuscitation., Results: In the first series the mortality rate was 41% (9 of 22). Early deaths occurred 3.0 +/- 0.8 days after injury and were distinguished by significantly lower neutrophil counts on resuscitation. Those alive a 7 days had intraabdominal abscesses and bacteremia. Alveoli and peribronchial spaces were congested, with edema and fibrin deposition in capillaries and alveoli. Livers were congested with biliary stasis. Despite these pathologic findings, hemodynamic, electrolyte, and serum enzyme changes were minimal. In the second series the mortality rate at 4 days was 30% with placebo (3 of 10) versus 33% with anti-CD18 (3 of 9). Lung changes (i.e., pneumonia, pleuritis, thrombosis, and edema) were similar in both groups, but liver congestion and hemorrhage were attenuated by anti-CD18. Some aspects of host defense were altered by anti-CD18. At 24 and 48 hours the oxidative burst potential for circulating granulocytes was 208% +/- 57% and 383% +/- 73% with placebo versus 1273% +/- 351% and 762% +/- 226% in anti-CD18. At 72 hours the granularity of circulating neutrophils was unchanged from baseline with placebo but was reduced to 82% +/- 5% by anti-CD18. At 48 hours lipopolysaccharide-evoked tumor necrosis factor production in vitro was reduced to 62% +/- 22% with placebo but was increased to 148% +/- 16% with anti-CD18., Conclusions: Anti-CD18 during fluid resuscitation did not increase vulnerability to endogenous pathogens because the transient inhibition of neutrophil demargination was balanced by enhanced oxidative burst, degranulation, and production of tumor necrosis factor in circulating cells later during recovery. Thus a single administration of antiadhesion therapy does not worsen posttrauma outcome even if given during ongoing sepsis.

Objective: There were two aims of this study. The first was to evaluate the application of helical computed tomography of the thorax (HCTT) for the diagnosis of blunt aortic injury (BAI). The second was to evaluate the efficacy of beta-blockers with or without nitroprusside in preventing aortic rupture., Summary Background Data: Aortography has been the standard for diagnosing BAI for the past 4 decades. Conventional chest CT has not proven to be of significant value. Helical CT scanning is faster and has higher resolution than conventional CT. Retrospective studies have suggested the efficacy of antihypertensives in preventing aortic rupture., Methods: A prospective study comparing HCTT to aortography in the diagnosis of BAI was performed. A protocol of beta-blockers with or without nitroprusside was also examined for efficacy in preventing rupture before aortic repair and in allowing delayed repair in patients with significant associated injuries., Results: Over a period of 4 years, 494 patients were studied. BAI was diagnosed in 71 patients. Sensitivity was 100% for HCTT versus 92% for aortography. Specificity was 83% for HCTT versus 99% for aortography. Accuracy was 86% for HCTT versus 97% for aortography. Positive predictive value was 50% for HCTT versus 97% for aortography. Negative predictive value was 100% for HCTT versus 97% for aortography. No patient had spontaneous rupture in this study., Conclusions: HCTT is sensitive for diagnosing intimal injuries and pseudoaneurysms. Patients without direct HCTT evidence of BAI require no further evaluation. Aortography can be reserved for indeterminate HCTT scans. Early diagnosis with HCTT and presumptive treatment with the antihypertensive regimen eliminated in-hospital aortic rupture.

Background: The purpose of this experimental study was to test whether transfusion potentiated physiologic changes associated with fluid resuscitated trauma in controlled conditions., Methods: Anesthetized and ventilated mongrel pigs were subjected to soft-tissue injury plus 35% hemorrhage and 1 hour shock and then were resuscitated with either autologous (shed) or heterologous (cross-transfused) fresh whole blood. Leukocyte differential counts, T-lymphocyte subsets, neutrophil adherence molecule (CD18) expression, granulocyte oxidative burst, plasma cortisol, and serum chemistries were monitored in awake animals with indwelling catheters on 3 consecutive days. Changes were referenced to preinjury baseline values and to a control group that received heterologous transfusion but no shock. To determine whether these changes might have influenced host defense, a low-dose challenge with Escherichia coli endotoxin (lipopolysaccharide [LPS]; 1 to 2 micrograms/kg for 30 minutes) was administered on day 4., Results: During recovery, neutrophil counts, neutrophil CD18 expression, and granulocyte oxidative burst were generally increased, but the changes were not potentiated by transfusion. Lymphocyte subpopulations remained relatively constant. Serum enzyme markers were elevated with trauma plus shed blood or trauma plus cross-transfusion, but they remained essentially unchanged after heterologous transfusion only. Plasma cortisol, a nonspecific index of stress, peaked at 3 to 6 times higher than baseline. The increases tended to be higher and later with heterologous transfusion only, relative to trauma plus shed blood or trauma plus cross-transfusion. The delayed LPS challenge evoked profound but transient pulmonary hypertension and leukopenia, followed by subsequent hypoxemia; the time courses and magnitude of these changes were similar in all groups., Conclusions: If these measured variables before and after LPS challenge are a valid index of host defense in this species, then a 35% transfusion does not potentiate the risk for posttrauma immune dysfunction when the magnitude of injury is constant. Thus the predisposition to infection after human trauma might be due to cold storage of blood; separation of blood into components, or other transfusion-related practices rather than to transfusion per se.

Objective: This study was conducted to evaluate those factors associated with popliteal artery injury that influence amputation, with emphasis placed on those that the surgeon can control., Summary Background Data: Generally accepted factors impacting amputation after popliteal artery injury include blunt trauma, prolonged ischemic times, musculoskeletal injuries, and venous disruption. Amputation ultimately results from microvascular thrombosis and subsequent tissue necrosis, predisposed by the paucity of collaterals around the knee., Methods: Patients with popliteal artery injuries over the 10-year period ending November 1995 were identified from the trauma registry. Preoperative (demographics, mechanism and severity of injury, vascular examination, ischemic times) and operative (methods of arterial repair, venous repair-ligation, anticoagulation-thrombolytic therapy, fasciotomy) variables were studied. Severity of extremity injury was quantitated by the Mangled Extremity Severity Score (MESS). Amputations were classified as primary (no attempt at vascular repair) or secondary (after vascular repair). After univariate analysis, logistic regression analysis was performed to identify the independent risk factors for limb loss., Results: One hundred two patients were identified; 88 (86%) were males and 14 (14%) were females. Forty injuries resulted from blunt and 62 from penetrating trauma. There were 25 amputations (25%; 11 primary and 14 secondary). Patients with totally ischemic extremities (no palpable or Doppler pulse) more likely were to be amputated (31% vs. 13%; p < 0.04). All requiring primary amputations had severe soft tissue injury and three had posterior tibial nerve transection; the average MESS was 7.6. Logistic regression analysis identified independent factors associated with secondary amputation: blunt injury (p = 0.06), vein injury (p = 0.06), MESS (p = 0.0001), heparin-urokinase therapy (p = 0.05). There were no complications with either heparin or urokinase., Conclusions: Minimizing ischemia is an important factor in maximizing limb salvage. Severity of limb injury, as measured by the MESS, is highly predictive of amputation. Intraoperative use of systemic heparin or local urokinase or both was the only directly controllable factor associated with limb salvage. The authors recommend the use of these agents to maximize limb salvage in association with repair of popliteal artery injuries.

The efficacy of prophylactic vena caval filters (VCF) in reducing morbidity and mortality from pulmonary embolism (PE) in high-risk trauma patients has been shown, but minimal follow-up data is currently available. VCFs were prophylactically placed in 110 patients between August 1991 and June 1995. There was an early VCF complication rate of 7%. Twenty-two patients died; the remaining 88 patients formed the basis for the follow-up study. Forty-five patients were located and interviewed by phone, and 30 of these patients (34%) returned for evaluation. The mean follow-up time was 18 months (range, 4-42 months). There was no incidence of caval thrombosis on follow-up. Eleven patients had physical findings, and duplex evidence consistent with postphlebitic syndrome. An additional three patients had evidence of old deep venous thrombosis (DVT) by duplex, but no significant symptomatology. VCF are effective in preventing PE related deaths and have few major complications. The long-term morbidity associated with posttraumatic venous thrombosis is significant. This morbidity is related not to PE or VCF, but to the underlying DVT. Improved strategies against DVT are necessary.

The majority of pancreatic injuries are minor in nature and can be managed easily and definitively with external drainage. The complexity of management increases significantly when a pancreatic ductal injury is present. It is requisite that thorough preoperative and intraoperative work-up be complete and systematic if injuries are to be properly recognized and managed. Once an injury has been detected, management guidelines based on injury classification can help to provide uniform results with minimal complications. In general, a conservative management scheme is indicated, the goals of such being preservation of pancreatic tissue and minimization of pancreaticoenteric anastomoses. Specific technical maneuvers may vary, but strict adherence to the basic concepts of hemorrhage control, contamination control, accurate pancreatic assessment, judicious resection, and adequate drainage can help to reduce the frequency of complications from these complex injuries.

Objective: The incidence, associated injury pattern, diagnostic factors, risk for adverse outcome, and efficacy of anticoagulant therapy in the setting of blunt and carotid injury (BCI) were evaluated., Summary Background Data: Blunt carotid injury is considered uncommon. The authors believe that it is underdiagnosed. Outcome is thought to be compromised by diagnostic delay. If delay in diagnosis is important, it is implied that therapy is effective. Although anticoagulation is the most frequently used therapy, efficacy has not been proven., Methods: Patients with BCI were identified from the registry of a level I trauma center during an 11-year period (ending September 1995). Neurologic examinations and outcomes, brain computed tomography (CT) results, angiographic findings, risk factors, and heparin therapy were evaluated., Results: Sixty-seven patients with 87 BCIs were treated. Thirty-four percent were diagnosed by incompatible neurologic and CT findings, 43% by new onset of neurologic deficits, and 23% by physical examination (neck injury, Horner's syndrome). There were 54 intimal dissections, 11 pseudoaneurysms, 17 thromboses, 4 carotid cavernous fistulas, and 1 transected internal carotid artery. Thirty-nine patients had follow-up angiograms. Mortality rate was 31%. Of 46 survivors, 63% had good neurologic outcomes, 17% moderate, and 20% bad. Logistic regression analysis demonstrated heparin therapy to be associated independently with survival (p < 0.02) and improvement in neurologic outcome (p < 0.01)., Conclusions: Blunt carotid injury is more common than appreciated, seen in 0.67% of patients admitted after motor vehicle accidents. Therapy with heparin is highly efficacious, significantly reducing neurologic morbidity and mortality. Heparin therapy, when instituted before onset of symptoms, ameliorates neurologic deterioration. Liberal screening, leading to earlier diagnosis, would improve outcome.

Objective: Ventilator-associated pneumonia (PN) is difficult to distinguish from trauma-induced systemic inflammatory response syndrome (SIRS), especially in patients with multiple injuries. Previous work using bronchoscopy and quantitative cultures demonstrated significant bacterial growth in about one-third of patients with clinical evidence of PN. In this prospective study, antibiotic therapy for PN was based solely on quantitative bronchoalveolar lavage (BAL) cultures., Methods: Mechanically ventilated trauma patients underwent bronchoscopy with BAL when they developed clinical evidence of PN: fever (temperature > 100.5 degrees F), white blood cells > 10,000 or > 10% immature forms, purulent sputum, and new or changing infiltrate on chest roentgenogram. Patients with other infections or those receiving antibiotics for any other reason were excluded. Empiric antibiotic therapy for PN was started at the time of BAL. If the quantitative cultures revealed > or = 10(5) colony-forming units (CFU)/mL, that patient was defined as having PN and was treated. If the cultures revealed < 10(5) CFU/mL, that patient was defined as having SIRS, and empiric therapy was stopped., Results: Forty-three patients (88% blunt, 12% penetrating) underwent bronchoscopy with BAL 55 times. Mean age was 40 and Injury Severity Score was 25. Twenty patients had > or = 10(5) CFU/mL (47%) and 23 had < 10(5) CFU/mL (53%). There were no differences in age, Injury Severity Score, temperature, white blood cell count, or ventilator days before BAL between groups. Sixty-five percent of those with SIRS improved after empiric therapy was stopped (average 3.3 days), and 35% underwent repeat BAL. Three patients with the initial diagnosis of SIRS developed PN (13% of SIRS). Mortality for PN was 15%, compared with 17% for SIRS; no deaths were related to antibiotic therapy., Conclusions: SIRS, which can mimic PN, is common in trauma patients. These entities can be distinguished by bronchoscopy with BAL. Basing antibiotic therapy solely on quantitative BAL cultures is efficacious in trauma patients.

Background: A number of retrospective studies recently have been published concerning nonoperative management of minor liver injuries, with cumulative success rates greater than 95%. However, no prospective analysis that involves a large number of higher grade injuries has been reported. The current study was conducted to evaluate the safety of nonoperative management of blunt hepatic trauma in hemodynamically stable patients regardless of injury severity., Methods: Over a 22-month period, patients with blunt hepatic injury were evaluated prospectively. Unstable patients underwent laparotomies, and stable patients had abdominal computed tomography (CT) scans. Those with nonhepatic operative indications underwent exploration, and the remainder were managed nonoperatively in the trauma intensive care unit. This group was compared with a hemodynamically matched operated cohort of blunt hepatic trauma patients (control subjects) who had been prospectively analyzed., Results: One hundred thirty-six patients had blunt hepatic trauma. Twenty-four (18%) underwent emergent exploration. Of the remaining 112 patients, 12 (11%) failed observation and underwent celiotomy--5 were liver-related failures (5%) and 7 were nonliver related (6%). Liver related failure rates for CT grades I through V were 20%, 3%, 3%, 0%, and 12%, respectively, and rates according to hemoperitoneum were 2% for minimal, 6% for moderate, and 7% for large. The remaining 100 patients were successfully treated without operation--30% had minor injuries (grades I-II) and 70% had major (grades III-V) injuries. There were no differences in admission characteristics between nonoperative success or failures, except admission systolic blood pressure (127 vs. 104; p < 0.04). Comparing the nonoperative group to the control group, there were no differences in admission hemodynamics or hospital length of stay, but nonoperative patients had significantly fewer blood transfusions (1.9 vs. 4.0 units; p < 0.02) and fewer abdominal complications (3% vs. 11%; p < 0.04)., Conclusions: Nonoperative management is safe for hemodynamically stable patients with blunt hepatic injury, regardless of injury severity. There are fewer abdominal complications and less transfusions when compared with a matched cohort of operated patients. Based on admission characteristics or CT scan, it is not possible to predict failures; therefore, intensive care unit monitoring is necessary.