Impact of advanced age on outcomes following damage control interventions for trauma.

BACKGROUND: Damage control (DC) strategy has significantly contributed to mortality reduction in massively bleeding and critically injured trauma victims. However, there is a lack of literature validating the effectiveness of this approach in the elderly population. METHODS: The trauma registry of a Level I trauma center was utilized to identify all severely injured patients [Injury Severity Score (ISS) ≥16] from January 1996 to December 2007 who underwent initial DC procedures. Patients with a head Abbreviated Injury Scale (AIS) ≥3 were excluded from the analysis. Demographics, clinical and physiological parameters, and in-hospital outcome measures were compared between elderly (≥55 years) and younger (<55 years) patient cohorts subjected to DC procedures. RESULTS: Overall, 158 patients met the inclusion criteria. Among them, 34 patients (21.5%) were aged ≥55 years (range 55-85 years) and 124 patients (78.5%) were <55 years old (range 16-54 years). The overall in-hospital mortality rate was 10.1% (n = 16) with a significantly higher mortality rate for elderly patients than for younger patients: 29.4% vs. 4.8%; adjusted P = 0.001; adjusted odds ratio (OR) with 95% confidence interval (CI) 7.09 (2.30-21.74). When stratified by DC subgroups, the case-fatality rate was significantly higher for the elderly patients who underwent extremity DC procedures [19.2% vs. 3.2%; adjusted P = 0.032; adjusted OR with 95% CI 5.95 (1.16-30.30)] and DC laparotomy [55.6% vs. 7.1%; P = 0.005; OR and 95% CI 16.25 (2.32-114.06)]. Both cohorts required massive transfusion during the initial 24 h of admission (18.9 ± 2.9 vs. 15.1 ± 1.6 units of packed red blood cells; P = 0.290). Nevertheless, there were no statistically significant differences between the two groups regarding hospital and surgical intensive care unit lengths of stay or major in-hospital complications. CONCLUSIONS: The mortality rate for elderly trauma patients undergoing DC is excessive at 29%. Despite the significant burden of injury and the massive transfusion requirement, most of the elderly patients subjected to DC survived and experienced in-hospital morbidity measures comparable to those of the younger patients. Our results provide further support for damage control intervention in severely injured elderly patients.

Crystalloids after primary colon resection and anastomosis at initial trauma laparotomy: excessive volumes are associated with anastomotic leakage.

BACKGROUND: Recognition of preventable risk factors for suture line failure after colon anastomosis is important for optimizing anastomotic healing. The purpose of this study was to investigate the impact of crystalloids on the occurrence of anastomotic leakage after traumatic colonic injuries. METHODS: Retrospective review from January 2005 to August 2009 of severely injured patients who underwent primary colocolonic anastomosis and intensive care unit (ICU) admission for ≥72 hours. Demographics on hospital and ICU admission, amount of crystalloids, and blood component transfusions within the first 72 hours were assessed by multivariate analysis to explore independent associations with anastomotic leakage. RESULTS: Of a total of 123 patients with primary colocolonic anastomosis, 7 died within 72 hour and 24 were discharged before 72 hour from the ICU. The remaining 92 patients required ICU admission for ≥72 hour. Their mean Injury Severity Score was 20.8 ± 10.7, and they were 29.9 years ± 13.0 years old. Twelve patients (13.0%) developed an anastomotic leak. Demographics on hospital and ICU admission, intraoperative blood loss, and the volume of intraoperative fluids given did not differ statistically between patients with or without anastomotic leakage. However, the cumulative amount of crystalloids given over the first 72 hours significantly predicted anastomotic leakage (area under the receiver operating characteristic curve: 0.758 [95% confidence interval 0.592-0.924], p=0.009). By multivariate analysis, ≥10.5 L of crystalloids given over the first 72 hours was independently associated with anastomotic breakdown (odds ratio [95% confidence interval]: 5.26 [1.14-24.39], p=0.033). In addition, increasing age, hemorrhagic shock on admission, and a concomitant stomach injury were independent risk factors for an anastomotic leak (R=0.396). CONCLUSION: Increased use of crystalloids after primary colocolonic anastomosis at initial trauma laparotomy is associated with anastomotic leakage. A threshold of 10.5 L of crystalloid fluid infused over the first 72 hours is associated with a 5-fold increased risk for colocolonic suture line failure. The impact of crystalloid restriction on anastomotic failure in trauma patients warrants prospective investigation.

Thromboembolic prophylaxis with low-molecular-weight heparin in patients with blunt solid abdominal organ injuries undergoing nonoperative management: current practice and outcomes.

BACKGROUND: Low-molecular-weight heparins (LMWHs) are effective in preventing thromboembolic complications after trauma. In the nonoperative management (NOM) of blunt solid abdominal organ injuries, the timing of the administration of LMWH remains controversial because of the unknown risk for bleeding. METHODS: Retrospective study including patients aged 15 years or older who sustained blunt splenic, liver, and/or kidney injuries from January 2005 to December 2008. Patients were stratified according to the type and severity of organ injuries. NOM failure rates and blood transfusion requirements were compared between patients who got LMWH early (≤3 days), patients who got LMWH late (>3 days), and patients who did not receive LMWH. RESULTS: Overall, 312 (63.8%) patients with solid organ injuries had NOM attempted. There were 154 splenic, 144 liver, and 65 kidney injuries (1.2 organs injured per patient). Forty-one patients (13.2%) received LMWH early, 70 patients (22.4%) received LMWH late, and 201 (64.4%) patients did not receive LMWH. The early LMWH group was less severely injured compared with the late LMWH group. However, the distribution of the risk factors for failure of NOM (high-grade injury, large amount of hemoperitoneum, and contrast extravasation) was similar between the three LMWH groups. Overall, 17 of 312 patients (5.4%) failed NOM (7.8% spleen, 2.1% liver, and 3.1% kidney). All but one failure occurred before LMWH administration. After adjustment for demographic differences, the overall blood transfusion requirements for the early LMWH group was significantly lower when compared with patients with late LMWH administration (3.0±5.3 units vs. 6.4±9.9 units; adjusted p=0.027). Pulmonary embolism and deep venous thrombosis occurred in four patients. The mortality rate for patients with splenic, liver, and kidney injuries was 3.2% and did not differ with LMWH application. CONCLUSION: In patients with solid abdominal organ injuries undergoing NOM, early use of LMWH does not seem to increase failure rates or blood transfusion requirements.

Gunshot injuries in the elderly: patterns and outcomes. A national trauma databank analysis.

BACKGROUND: Trauma in the elderly (≥ 55 years) accounts for a significant proportion of admissions to trauma centers. Our understanding of the epidemiology and outcomes associated with penetrating injury in this age segment of the population, however, is severely limited. The aim of the present study therefore was to investigate the incidence and type of injuries sustained by elderly patients from firearms and the impact of age on outcomes. METHODS: This was a 5-year National Trauma Databank (NTDB) study. Injury demographics, mortality rates, and lengths of stay in the Intensive Care Unit (ICU) and the hospital were analyzed. Elderly patients ≥ 55 years old were assigned to one of three categorical strata: 55-64 years old, 65-74 years old, and ≥ 75 years old. RESULTS: During the study period, 98,242 patients were admitted for firearm-related injuries, and 3,190 (3.2%) of them were ≥ 55 years old. Within the elderly age segment of the population, 1,676 patients (52.5%) were 55-64 years of age, 727 (22.8%) were 65-74 years of age, and 787 (24.7%) were ≥ 75 years old. The incidence of severe trauma [Injury Severity Score (ISS) ≥ 16] in the elderly age strata was 43.3, 46.8, and 57.6%, respectively (p < 0.001). Patients ≥ 75 years old were significantly more likely than patients 55-74 years old to suffer self-inflicted injuries. The most commonly encountered injury in elderly patients was gunshot wounds to the head, which increased in a stepwise fashion with advancing age (25.8, 31.6, and 39.4% respectively; p < 0.001). The crude mortality rate in all patients sustaining gunshot wounds increased progressively with age. Within the elderly age segment, mortality ranged from 28.5% in the age stratum 55-64 years, to 55.4% in the stratum ≥ 75 years (adjusted p < 0.001). Intensive care unit and hospital length of stay increased with advancing age but peaked and remained stable among the elderly age groups. An admission Glasgow Coma Score (GCS) ≤ 8, an ISS ≥ 16, hypotension on admission, age, self-inflicted injury, and injury sustained by assault were factors independently associated with death in patients ≥ 55 years. CONCLUSIONS: Injury from firearms is not uncommon in the elderly patient population and is primarily a result of self-inflicted gunshot wounds to the head. These patients sustain a high burden of injury and a high rate of mortality, which increases with advancing age.

In-hospital small bowel obstruction after exploratory laparotomy for trauma.

BACKGROUND: The purpose of this study was to examine the incidence and risk factors of in-hospital small bowel obstruction (SBO) after exploratory laparotomy for trauma. METHODS: A retrospective review of patients surviving over 72 hours after an exploratory laparotomy for trauma. Patients with intestinal obstructive symptoms were reviewed by a consensus panel, which evaluated the clinical, laboratory, and radiologic findings to validate the diagnosis of SBO. RESULTS: A total of 571 patients met inclusion criteria. The incidence of early SBO was 3.9%, with 22.7% of these patients requiring surgical intervention. Patients with gastrointestinal (GI) perforation had a significantly higher incidence of SBO, compared with those with no GI perforation (5.7% vs. 1.3%, p = 0.007). A forward logistic regression identified the presence of a GI perforation as the only factor independently associated with early SBO (adjusted odds ratio: 4.39; 95% confidence interval: 1.28-15.15; p = 0.019). The overall hospital stay was significantly longer for SBO patients (27.0 days ± 26.7 days vs. 16.0 days ± 22.8 days; adjusted mean difference: 11.5; 95% confidence interval: 1.6-21.3; p = 0.022). Development of SBO increased the cost by 59.7%. CONCLUSION: The incidence of in-hospital SBO after laparotomy for trauma is significant at 3.9%. The presence of a GI perforation is independently associated with the development of this complication. Over a fifth of patients with early SBO will require a surgical intervention. The use of preventive strategies may be justified in selected, high-risk patients to reduce the burden associated with early SBO.

The impact of Acinetobacter baumannii infections on outcome in trauma patients: a matched cohort study.

OBJECTIVE: To determine the impact of Acinetobacter baumannii infection on the outcome of trauma patients. DESIGN AND SETTING: A retrospective 1:2-matched cohort study. Level I trauma intensive care unit patients with confirmed Acinetobacter baumannii infection were defined as cases. PATIENTS: Thirty-one Acinetobacter baumannii patients were matched to 62 controls with evidence of infection caused by other microorganisms. MEASUREMENTS AND MAIN RESULTS: There were 12 matching criteria, including focus of infection, demographics, severity, and characteristics of injury. In-hospital mortality rate, intensive care unit length of stay, and complications of Acinetobacter baumannii including multidrug-resistant strains in patients were compared to those of their controls; 81% had hospital-acquired pneumonia, 13% had bloodstream infections, and 6% had urinary tract infections in both groups. Acinetobacter baumannii cultures were multidrug resistant in 42% (13/31) of cases. The initial empirical antibiotic therapy was adequate in 71% (22/31). Although the in-hospital mortality was higher in the Acinetobacter baumannii group (16% vs. 13%; odds ratio, 1.23; 95% confidence interval, 0.38-4.36; p = .67), the difference did not reach statistical significance. Using the test of equivalence or clinical indifference, the impact of an Acinetobacter baumannii infection on mortality is inconclusive. This applies also to multidrug-resistant strains. Overall intensive care unit stay was prolonged for Acinetobacter baumannii when compared to controls (median, [range], 28 [7-181] days vs. 17 [2-130] days, respectively; p = .05). ARDS and acute liver failure were more frequent in the Acinetobacter baumannii group compared to the control group (35% vs. 15%; odds ratio, 3.24; 95% confidence interval, 1.17-5.48; p = .02 and 26% vs. 10%; odds ratio, 3.25; 95% confidence interval, 3.25-10.40; p = .04). CONCLUSIONS: In this single-center experience, Acinetobacter baumannii infection, including multidrug-resistant strains, has inconclusive impact on mortality in a cohort of trauma patients. Larger studies are needed to support a definite conclusion. Acinetobacter baumannii infection was, however, associated with a longer intensive care unit stay and a higher rate of organ failure.

Microbiological profile and antimicrobial susceptibility in surgical site infections following hollow viscus injury.

INTRODUCTION: The purpose of this study was to assess the microbiological profile, antimicrobial susceptibility, and adequacy of the empiric antibiotic therapy in surgical site infections (SSI) following traumatic hollow viscus injury (HVI). METHODS: This is a retrospective study of patients admitted with an HVI from March 2003 to July 2009. SSI was defined as a wound infection or intra-abdominal collection confirmed by positive cultures and requiring percutaneous or surgical drainage. RESULTS: A total of 91 of 667 (13.6%) patients with an HVI developed an SSI confirmed by positive culture. Mean age was 33.0 +/- 14.1 years, mean Injury Severity Score (ISS) was 17.7 +/- 9.6, 91.2% were male, and 80.2% had sustained penetrating injuries. The SSI consisted of 65 intra-abdominal collections and 26 wound infections requiring intervention. The most commonly isolated species in the presence of a colonic injury was Escherichia coli (64.7%), Enterococcus spp. (41.2%), and Bacteroides (29.4%), and in the absence of a colonic perforation, Enterococcus spp. and Enterobacter cloacae (both 38.9%). Susceptibility rates of E. coli and E. cloacae, respectively, were 38% and 8% for ampicillin/sulbactam, 82% and 4% for cefazolin, 96% and 92% for cefoxitin, with both 92% to piperacillin/tazobactam, and 100% to ertapenem. The initial empirical antibiotic therapy adequately targeted the pathogens in 51.6% of patients who developed an SSI. CONCLUSION: The distribution of the microorganisms isolated from SSIs differed significantly according to whether or not a colonic injury was present. Empiric antibiotic treatment was inadequate in upwards of 50% of patients who developed an SSI. Further investigation is warranted to determine the optimal empiric antibiotic regimen for reducing the rate of postoperative SSI.

The impact of platelets on the progression of traumatic intracranial hemorrhage.

BACKGROUND: The purpose of this study was to analyze the association of the initial platelet count with mortality and progression of intracranial hemorrhage (ICH) in blunt traumatic brain injured (TBI) patients. METHODS: All blunt trauma patients with severe TBI admitted from January 2006 to December 2007 were retrospectively identified. Patients with a chest, abdomen, or extremity AIS score >3 were excluded to minimize the impact of concomitant injuries on the outcomes of the patients. All brain computed tomography scans were reviewed to analyze ICH progression. Discrete platelet cutoff values were entered into a multiple regression model to detect critical thresholds associated with ICH progression and mortality. RESULTS: Of 626 TBI patients, 310 (49.5%) had a minimum of two brain computed tomography scans and were able to have ICH progression evaluated. Patients with platelets <175,000/mm3 had a significantly increased risk for ICH progression (OR [95% CI]: 2.09 [1.07-4.37]; adjusted p = 0.043). ICH progression was associated with increased need for craniotomy (OR [95% CI]: 3.27 [1.28-8.33]; adjusted p = 0.013) and mortality (OR [95% CI]: 3.41 [1.11-10.53]; adjusted p = 0.033). A platelet count <100,000/m3 was an independent predictor for mortality (OR [95% CI]: 9.5 [1.3-71.4]; adjusted p = 0.029). CONCLUSION: A platelet count <100,000/mm3 is associated with a ninefold adjusted risk of death, and a platelet count <175,000/mm3 is a significant predictor of ICH progression. The impact of early correction of the admission platelet count warrants further validation.

Serial white blood cell counts in trauma: do they predict a hollow viscus injury?

BACKGROUND: The significance of serial white blood cell (WBC) counts in trauma patients with a suspected hollow viscus injury (HVI) is unknown. The purpose of this study was to examine the role of serial WBC counts in the diagnosis of a HVI. METHODS: After institutional review board approval, all injured patients admitted to a Level I trauma center from January 2003 to December 2007 with at least one WBC measurement were included in a retrospective analysis. The WBC profiles for patients with a HVI were compared against those without HVI. All WBC counts are reported as [x10(3)/microL]. RESULTS: The mean WBC count of the overall study population (n = 5,950) on admission was 11.6 +/- 5.3. Overall, 59.2% had an elevated WBC count on admission. A significant relationship between increasing Injury Severity Score and increasing WBC count on admission was found by linear regression. When comparing patients with HVI (n = 267) with patients without HVI (n = 5,683), no significant difference was found for admission WBC count. The highest WBC count within the first 24 hours for patients with HVI was 16.7 +/- 4.7. This was significantly higher than that for the 4,520 patients without any intraabdominal injury (13.0 +/- 5.2, adjusted p < 0.001). Penetrating injury, a concomitant severe thoracic trauma (chest Abbreviated Injury Scale value >or=3), and highest WBC count >or=20.0 in the first 24 hours were independent risk factors for HVI. A maximal WBC count or=20.0 are independently associated with a HVI, whereas counts

Decompressive craniectomy: surgical control of traumatic intracranial hypertension may improve outcome.

INTRODUCTION: The purpose of this study was to assess the role of decompressive craniectomy (DC) inpatients with post-traumatic intractable intracranial hypertension (ICH) in the absence of an evacuable intracerebral haemorrhage. METHODS: Retrospective study at LAC+USC Medical Centre including patients who underwent DC for post-traumatic malignant brain swelling or ICH without space occupying haemorrhage, during the period 01/2004 to 12/2008. The analysis included the effect of DC on intracranial pressure (ICP) and timing of DC on functional outcomes and survival. RESULTS: Of 106 patients who underwent DC, 43 patients met inclusion criteria. Of those, 34 were operated within the first 24 h from admission. DC decreased the ICP significantly from 37.8 ± 12.1 mmHg to 12.7 ± 8.2 mmHg in survivors and from 52.8 ± 13.0 to 32.0 ± 17.3 mmHg in non-survivors. Overall 25.6%died (11 of 43), and 32.5% (14 of 43) remained in vegetative state or were severely disabled. Favourable outcome (Glasgow Outcome Scale 4 and 5) was observed in 41.9% (18 of 43). No tendency towards either increased or decreased incidence in favourable outcome was found relative to the time from admission to DC.Six of the 18 patients (33.3%) with favourable outcome were operated on within the first 6 h. CONCLUSIONS: DC lowers ICP and raises CPP to high normal levels in survivors compared to non-survivors.The timing of DC showed no clear trend, for either good neurological outcome or death. Overall, the survival rate of 74.4% is promising and 41.9% had favourable neurological outcome.