The Functional Independence Measure (FIM) is used by rehabilitation professionals to access disability. The FIM score combines both motor and cognitive parameters to assess a patient's level of required assistance in performing activities of daily living (ADL). The geriatric trauma patient is becoming an increasingly important cohort for trauma services. FIM has been shown to predict discharge outcomes and those at high risk for falls. We hypothesized pretrauma FIM scores may predict survival in the geriatric trauma population. This was a retrospective study of patients 65 years and older that were admitted to our Level I trauma center from July 1, 2006 to July 1, 2012. A total 941 patients underwent stepwise regression to identify those factors predicting survival. Age, Injury Severity Score, revised trauma score, body mass index, and pretrauma FIM scores (12-point scale) were studied. The primary outcome was survival. Statistical significance reached at P value <0.05. Multiple logistic regression analysis was then performed. A total of 1315 patients were identified and complete data were available on 941 patients. Mean age was 78 (SD ± 8.2), mean Injury Severity Score was 13(SD ± 8.7), and mean body mass index was 26. Overall mortality was 11 per cent. The odds ratio of survival was 3.532 (95% confidence interval = 2.191–5.718) times greater for every 1-point increase in the preadmission FIM expression score. Glasgow Coma Scale, revised trauma score, gender, and pretrauma FIM expression scores were predictive of survival in the geriatric trauma patient. Pretrauma FIM expression can be used to predict survival in the elderly trauma victim. Further study is needed to establish the role of FIM as part of trauma scoring systems.

The National Surgical Quality Improvement Project (NSQIP) uses data abstracted by highly trained personnel to evaluate risk-adjusted outcomes in surgical patients. It uses an odds-ratio method to compare morbidity and mortality rates among patients at individual hospitals with those in the national database. It has allowed models to be built to estimate risks for individual patients undergoing a specific procedure: the on-line NSQIP risk calculator. Some investigators have suggested that NSQIP may underestimate the impact of preoperative risk factors and/or may fail to account for some risk factors in their analysis.1 This problem may be more common in hospitals that care for patients with a higher prevalence of risk factors. As a result, these hospitals may suffer financially when Medicare determines that their risk of mortality after surgical procedures is higher than the national average. Relatively early in the expansion of NSQIP through the American College of Surgeons, some questions became apparent. Certain risks are subjective: the American Society of Anesthesiology score and the patient’s functional health status. Models with and without these variables produce different results, indicating the importance of inclusion in the risk model. In 2010, an adjustment to the risk model with the use of a novel procedure risk score obtained a higher concordance (C-statistic). It was then shown that for specific procedures, fewer variables could be used with similar predictive ability. In 2013, yet another improvement was made with the addition of the procedure mix adjustment. This used a hierarchical model with a shrinkage adjustment to stabilize estimates derived from small sample sizes. Clearly, the perfect risk model is a moving target. In 2014, Sherman et al.2 concluded that the NSQIP risk calculator underestimates proctectomy morbidity and offered a new validated risk model. A similar conclusion was reached by Saunder et al.3 concerning gastric cancer resections at several major academic centers including Ohio State, Emory, Johns Hopkins, University of Wisconsin, Stanford, Wake, and Washington University. There is also a difference in risk estimate reliability between elective and emergent procedures. The reasons for the perceived underestimation of risks are not clear. Perhaps there are other major risk factors which are not in the NSQIP database that have significant effects on outcomes but remain unaccounted for in the risk model. Even a small proportion of patients with a substantial unaccounted risk can have a dramatic impact on the assessment of the risk adjusted surgical outcomes. Neither time/distance of travel for care nor narcotic dependence is in the NSQIP risk adjustment model. An analysis of our own institution’s patient populations shows that 25 per cent of our colorectal patients were narcotic dependent on presentation. Table 1 is a list of NSQIP risk factors found in our patients who died compared with the NSQIP database for the period of 2012 through 2015. A review of Table 1 reveals that three times the percent of our patients who died were transferred from an outside emergency room, and 47 per cent of our patients who died presented in septic shock. Approximately 71 per cent of our patients who died were emergent compared with 53 per cent in the NSQIP database. As discussed previously, the risk adjustment is all important. Stepwise regression has two key flaws in the process. First, it may underestimate the importance of the interaction of variables and may overestimate sporadic associations between independent and dependent variables. The measures of performance of a risk model include Brier score, the goodness-of-fit statistic, and the C-statistic. There are several refinements to these measures, including net reclassification improvement and the integrated discrimination improvement. The C-statistic can be less informative and lead to incorrect conclusions as the case mix is restricted and patients become more uniform, as in the Address correspondence and reprint requests to Sandy L. Fogel, M.D., Department of Surgery, Virginia Tech Carilion School of Medicine, 3 Riverside Circle, Roanoke, VA 24016. E-mail: slfogel@carilionclinic.org.

Background: Victims of traumatic injuries represent a population at risk for a wide variety of complications. Contact isolation (CI) is a set of restrictions designed to help prevent the transmission of medically significant organisms in the healthcare setting. A growing body of literature demonstrates that CI can have significant implications for the individual isolated patient. Our goal was to characterize the use of contact isolation at our Level I trauma center and investigate the association of CI with infectious complications. Patients and Methods: An existing trauma database containing data on patients admitted at our Level I trauma center between January 1, 2011 and December 31, 2012, along with their contact isolation status, was queried. Demographics, injuries, and the presence of infections were collected. Diagnosis of pneumonia or UTI was based on clinical documentation in the patient's medical record. A chart review was performed to ascertain the reason for CI including specific organisms. Because of differences in patient demographics between the CI and non-CI groups, linear regression was performed to adjust for the effects of different variables. Results: A total of 4,423 patients were admitted over this period. Of these, 4,318 (97.6%) had complete records and were included in the subsequent analysis. The CI was in place in 249 (5.8%) patients; 4,069 (94.2%) were not isolated. The number who had CI initiated for MRSA nasal colonization was 173 (69.5%). Twenty-two (8.9%) had no reason for CI documented. Pneumonia occurred in 190 (4.4%), 54 (21.7) in the CI group versus 136 (3.3%) in the non-CI group. Urinary tract infection (UTI) was diagnosed in 166 (3.8%), 48 (19.3%) in the CI group versus 118 (2.9%) in the non-CI group. Using logistic regression and excluding patients placed on contact isolation for the development of a new resistant nosocomial infection, CI, Injury Severity Score, gender, length of stay, and mechanical ventilation were identified as common covariates for pneumonia (PNA) and UTI. Chronic obstructive pulmonary disease COPD was specifically identified for PNA. Spinal cord injury, vertebral column injury and pelvic-urogenital injury were also significant for UTI. Conclusions: The development of pneumonia and UTI in patients with trauma was significantly associated with the use of CI. Because the majority of these patients had CI precautions in place for asymptomatic colonization, the CI provided them no direct benefit. Because the use of CI is associated with multiple negative outcomes, its use in the trauma population needs to be carefully re-evaluated.

Background: Trauma patients are at risk for malnutrition due to metabolic needs associated with injuries and surgery. Ileus may result in improper withholding of vital enteral nutrition. Contact isolation precautions (CI) are a set of restrictions intended to prevent spread of certain organisms. Our goal was to study a possible association between CI and development of ileus among trauma patients. Methods: Our Level I trauma center's institutional trauma database was queried for all patients evaluated between January 1, 2011 and December 31, 2012. Data collected included demographics, comorbidities, and development of ileus. A separate infection control database was used to determine patients on CI. Unadjusted relationships were determined by chi-square. Logistic regression was then used to adjust for patient and injury characteristics. Results: A total of 4,423 trauma patients were evaluated during the study period; of these, 4,317 (97.6%) patients had complete records and were analyzed. CI was in place for 251 (5.8%) patients; 4,066 (94.2%) were not isolated. In the CI group, 14 (5.6%) had ileus vs. 74 (1.8%) in the non-CI group (p<0.0001; OR 3.19; 95% CI 1.77-5.73). Next, logistic regression was used to adjust for potential confounders. Gender, ISS, and CI were all statistically significant (p<0.05) in their association with ileus. Conclusion: The use of CI in trauma patients is significantly associated with the development of ileus. A growing body of evidence suggests that CI among this population, which is already at greater risk of malnutrition and caloric deficit, should be re-evaluated.

Venous thromboembolism (VTE) represents a major risk for patients who have sustained a traumatic injury. Several additional risk factors for adverse thrombotic events are well characterized and incorporated into risk stratification models to quantify VTE risk.1, 2 To date, these models have not included transfusion as a contributing factor. Although some studies have shown an association between transfusion or transfusion timing and VTE, these have focused primarily on large transfusion volumes or the age of stored blood products.3, 4 It is our hypothesis that even in small quantities, the transfusion of packed red blood cells (pRBCs) in orthopedic trauma patients is an independent risk factor for VTE, and that transfusion timing has a measurable effect on clotting risk. Clinical and demographic data were obtained from the Carilion Roanoke Memorial Hospital trauma database for patients with extremity injuries between January 1, 2008 and December 31, 2013. Variables included age, sex, Injury Severity Score (ISS), Abbreviated Injury Scale (AIS), obesity, surgery, hospital length of stay (HLOS), femoral or tibial fractures, and transfusion requirements. Patients with nonextremity Abbreviated Injury Scale >2 were excluded to limit the effects of nonorthopedic injuries. Furthermore, because risks associated with larger volume transfusions are well described, patients receiving $5 units were excluded. Pre-existing conditions that might significantly impact clot formation, including coagulopathies, thrombophilias, pregnancy, cancer, and atrial fibrillation were identified but not considered grounds for exclusion. Chart review was performed to identify the indication for blood products, perior intraoperative transfusion setting, hemoglobin (Hb) before transfusion and the use of anticoagulants for chemical VTE prophylaxis. The primary outcome was VTE, defined as deep vein thrombosis confirmed by venous duplex ultrasound or pulmonary embolism confirmed by CT angiography. Pearson’s chi-squared test and Student’s unpaired t test were used to analyze categorical and continuous variables, respectively. Backward stepwise logistic regression models were used to measure the independent effect of variables using effect likelihood ratio testing. Data were analyzed using JMP Pro 11 (SAS, Cary, NC). The threshold for significance was defined as P < 0.05. A population of 319 patients was established for analysis including 58 patients who were transfused and 261 who were not. Of those transfused, 13 (22.4%) had documented VTE formation versus 12 (4.6%) in the nontransfusion group for an adjusted odds ratio of 4.12 (95% confidence interval: 1.57–10.99.) Though this is in qualitative concordance with prior studies that examined transfusion volumes $5 units, it represents a larger associated risk than has been previously cited. Four patients died during their hospitalization, none related to VTE. There was little variability in chemoprophylaxis for those transfused as 54 of the 58 were anticoagulated, with 51 placed on enoxaparin sodium and three on heparin. Two of the four who received no chemoprophylaxis were not surgical patients and it was unclear why the remaining two were not anticoagulated. All patients placed on anticoagulants before surgery had doses held the day of their procedure. For all patients with available data who received pRBCs outside the operating room, including nonsurgical patients, 22 had Hb <7.0 g/dL and 22 had values $7.0 g/dL. There were seven (31.8%) and five (22.7%) deep vein thromboses recorded in each group, respectively. Though variable documentation in electronic medical records did not always clearly denote the indication for intervention, chart review suggested that at least 14 of the 22 patients transfused with Hg >7.0 g/dL were given pRBCs in the absence of a physiologically symptomatic anemia. An established transfusion threshold of Hg <7.0 g/dL and risks demonstrated in the current and prior studies would seem to suggest these patients were inappropriately transfused. Presented as a poster at the Southeastern Surgical Congress Annual Scientific Meeting, February 22, 2016, Atlanta, GA. Address correspondence and reprint requests to Mark E. Hamill, M.D., Department of Surgery, Carilion Clinic, 1906 BelleviewAvenue SW, Roanoke, VA 24014. E-mail: mehamill@carilionclinic.org.

BACKGROUND: Recent federal legislation driving transition from fee-for-service to alternative methods of payment makes risk recognition essential for determination of appropriate payment systems. Because negotiations will include bundled population cohorts, we compared risk and results of an urban safety net teaching hospital's surgical population with state and national cohorts.STUDY DESIGN: Deidentified summary data for 2013 and 2014 were analyzed to compare the safety net teaching hospital with a statewide collaborative and a national cohort from similar academic centers. Incidence of preoperative risk factors were compared, identifying those that were >50% higher than both state and national experiences. These were compared for change in incidence between years. Outcomes were evaluated by 30-day mortality, readmissions, return to operating room, length of stay, and adverse events incidence.RESULTS: For both years, incidence of smoking, ventilator dependence, and CHF within 30 days was >50% higher than in the state and national cohorts. In 2014, septic shock was added to this, along with increased diabetes (14.3% to 19.8%), CHF (1.9% to 2.8%), and hypertension (39.9% to 52.5%). Despite these changes, 30-day mortality, return to operating room, length of stay, and readmissions were within +/- 5% of state and national results. Unplanned intubation, ventilation longer than 48 hours, and acute renal failure were 10th decile outliers. Median and interquartile range for length of stay were similar for all 3 populations across both years.CONCLUSIONS: The incidence of comorbid conditions defines greater risk in this safety net teaching hospital population. Increased smoking-related pathology reflects local population disease burden, and increased ventilator support defines additional cost for this care. As disease-, procedure-, or population-based payment alternatives evolve, risk recognition, reduction, and resolution will be essential for determination of cost-efficient, optimal, surgical outcomes. (J Am Coll Surg 2016; 222: 505-512. (C) 2016 by the American College of Surgeons. Published by Elsevier Inc. All rights reserved.)

BACKGROUND Contact isolation (CI) is a series of precautions used to prevent the transmission of medically significant infectious pathogens in the health care setting. Our institution's implementation of CI includes limiting patient movement to the assigned room. Our objective was to define the association between CI and venous thromboembolism (VTE) at our Level I trauma center.METHODS Our institution's prospective trauma database was retrospectively queried for all patients admitted to the trauma service between January 1, 2011, and December 31, 2012. Data including demographics, Injury Severity Score (ISS), preexisting medical conditions, injury type, and VTE development were collected. CI status data were obtained from our institution's infection control database. (2) was used to examine the unadjusted relationship between CI status and VTE. As the groups were not equivalent, logistic regression was then used to examine the relationship between CI and VTE while adjusting for relevant covariates including sex, age, ISS, and comorbidities.RESULTS Of the 4,423 trauma patients admitted during the study period, 4,318 (97.6%) had complete records and were included in subsequent analyses. A total of 249 (5.8%) of the patients were on CI. VTE occurred in 44 patients (17.7%) on CI versus 141 patients (3.5%) who were not isolated (p < 0.0001; odds ratio, 6.0; 95% confidence interval, 4.1-8.6). With the use of lasso [least absolute shrinkage and selection operator] regression to adjust for patient risk factors, this relationship remained highly significant (p < 0.0001; odds ratio, 2.61; 95% confidence interval, 1.7-4.0).CONCLUSION CI, ISS, hospital length of stay, and cardiac comorbidity were associated with VTE. After adjustment for other risk factors, CI remained most strongly associated with VTE. Although any medical intervention may come with unintended consequences, the risks and benefits of CI in this population need to be reevaluated. Further study is planned to identify opportunities to mitigate this increased VTE risk.LEVEL OF EVIDENCE Prognostic/epidemiologic study, level III; therapeutic study, level IV.

Drake, Mack D. DO; Hamill, Mark E. MD; Bradburn, Eric H. DO; Taylor, Dallas A. BS, RN; Gilbert, Carol M. MD; Baker, Christopher C. MD, FACS; Kundzins, John R. BS; Ferrara, John J. MD, FACS; Collier, Bryan R. DO, FACS Author Information

The American College of Surgeons National Surgical Quality Improvement Program (NSQIP) data at our institution indicated that surgical mortality was significantly higher than expected. This study examines the effect of implementation of a strict, intensive preoperative screening and intervention process on postoperative mortality at our institution, as measured by the NSQIP. Carilion Roanoke Memorial Hospital (CRMH) is a 763-bed tertiary care hospital serving a population of one million people in southwest Virginia. Data were collected for NSQIP at CRMH from July 2007 to December 2012. In January 2010, a new preoperative process was implemented to include risk assessment and intervention for hypertension, cardiac disease, pulmonary disease, diabetes, renal disease, and obstructive sleep apnea. Before initiation of our preoperative program (July 2007 to January 2010), odds ratios (ORs) for 30-day mortality in general and vascular cases were significantly higher than expected (1.40, 1.43, 1.58, and 1.56 in successive reporting periods). Beginning with the first report after implementation of the preoperative screening program, CRMH showed a progressively decreasing OR for overall 30-day mortality (1.26, 1.19, 1.14, 0.86, 0.82, 0.84, 0.89) with similar reductions in both general (0.92) and vascular (0.92) surgery. The implementation of an intensive preoperative screening and intervention process in our institution was accompanied by a significant decrease in the 30-day mortality for general surgery and vascular procedures, as measured by the NSQIP.

We describe the implementation of a care pathway for patients with fractured neck of femur (NOF) using Lean and Six Sigma principles. After introduction of the Lean pathway, 32 patients out a total of 86 (37%) with fractured NOF were admitted to the Trauma Ward within 4 hours of presentation to the hospital; prior to implementation this was 16 patients out of a total of 59 (27%). Post-Lean an earlier mean theatre start time of 8.40am was achieved, resulting in a 38 minute increase in daily theatre time. An additional 52 patients (12%) received surgery within 24 hours of admission, resulting in 1 night length of stay reduction. Lean methodology proved an effective method to guide change resulting in an improved journey for the patient and significant workflow gains.

The use of computerized decision support systems (CDSS) in glucose control for critically ill surgical patients has been reported in both diabetic and nondiabetic patients. Prospective studies evaluating its effect on glucose control are, however, lacking. The objective of this study was to evaluate patient-specific computerized IV insulin dosing on blood glucose levels (BGLs) by comparing patients treated pre-CDSS with those treated post-CDSS.A prospective study was performed in 4 surgical ICUs and 1 progressive care unit comparing patient data pre- and post-implementation of CDSS. The primary outcomes measures were the impact of the CDSS on glycemic control in this population and on reducing the incidence of severe hypoglycemia.Data on 1,682 patient admissions were evaluated, which corresponded to 73,290 BGLs post-CDSS compared with 44,972 BGLs pre-CDSS. The percentage of hyperglycemic events improved, with BGLs of >150 mg/dL decreasing by 50% compared with 6-month historical controls during the 18-month study period from July 2010 through December 2011. This was true for all 5 units individually (p < 0.0001, by one sample sign test). In addition, severe hypoglycemia (defined as BGL <40 mg/dL) decreased from 1% to 0.05% after implementing CDSS (p < 0.0001 by 2-sided binomial test).Patients whose BGLs were managed using CDSS were statistically significantly more likely to have a glucose reading under control (<150 mg/dL) than in the 6-month historical controls and to avoid serious hypoglycemia (p < 0.0001).

The Surgical Care Improvement Project (SCIP) is aproject that focuses on improving surgical care by reducing surgical morbidity and mortality by 25 per cent by 2010. Starting in 2011, SCIP compliance affects Medicare and Medicaid reimbursement rates. Although SCIP reinforces better practices in surgical care, does compliance with SCIP measures actually result in a decrease in surgical morbidity and mortality? This study examined compliance with the SCIP surgical site infection (SSI) module (prophylactic antibiotic received within 1 hour before surgical incision) during 2009 to 2010 (n = 703) to determine whether patients compliant with SCIP data had a correlation with SSI rates as reported by National Surgery Quality Improvement Program (NSQIP) data for the same time period. We found no statistically significant association in patients that have failed SCIP INF1 in the years 2009 to 2010 (n = 43) and the rates of SSI (n = 0) for the same time period. These data suggest that SCIP compliance should not be used to determine Medicare and Medicaid reimbursement rates because there is no correlation between failure of SCIP INF1 and SSI. Instead, further effort should be placed on developing tools designed to acknowledge outcome measures that result in decreased morbidity/mortality and change practices accordingly such as NSQIP.

Duchesne, Juan C. MD, FACS, FCCP; McSwain, Norman E. Jr. MD, FACS; Cotton, Bryan A. MD, FACS; Hunt, John P. MD, MPH, FACS; Dellavolpe, Jeff MD; Lafaro, Kelly MD, MPH; Marr, Alan B. MD, FACS; Gonzalez, Earnest A. MD, FACS; Phelan, Herb A. MD, FACS; Bilski, Tracy MD, FACS; Greiffenstein, Patrick MD; Barbeau, James M. MD, JD; Rennie, Kelly V. MD; Baker, Christopher C. MD, FACS; Brohi, Karim MD, FRCS, FRCA; Jenkins, Donald H. MD, FACS; Rotondo, Michael MD, FACS Author Information

BACKGROUNDDamage control laparotomy (DCL) improves outcomes when used in patients with severe hemorrhage. Correction of coagulopathy with close ratio resuscitation while limiting crystalloid forms a new methodology known as damage control resuscitation (DCR). We hypothesize a survival advantage in DCL patients managed with DCR when compared with DCL patients managed with conventional resuscitation efforts (CRE).METHODSThis study is a 4-year retrospective study of all DCL patients who required >or=10 units of packed red blood cells (PRBC) during surgery. A 2-year period after institution of DCR (DCL and DCR) was compared with the preceding 2 years (DCL and CRE). Univariate analysis of continuous data was done with Student's t test followed by multiple logistic regression.RESULTSOne Hundred twenty-four and 72 patients were managed during the DCL and CRE and DCL and DCR time periods, respectively. Baseline patient characteristics of age, Injury Severity Score, % penetrating, blood pressure, hemoglobin, base deficit, and INR were similar between groups. There was no difference in quantity of intraoperative PRBC utilization between DCL and CRE and DCL and DCR study periods: 21.7 units versus 25.5 units (p = 0.53); however, when compared with DCL and CRE group, patients in the DCL and DCR group received less intraoperative crystalloids, 4.7 L versus 14.2 L (p = 0.009); more fresh frozen plasma (FFP), 18.2 versus 6.4 (p = 0.002); a closer FFP to PRBC ratio, 1 to 1.2 versus 1 to 4.2 (p = 0.002); platelets to PRBC ratio, 1:2.3 versus 1:5.9 (0.002); shorter mean trauma intensive care unit length of stay, 11 days versus 20 days (p = 0.01); and greater 30-day survival, 73.6% versus 54.8% (p < 0.009). The addition of DCR to DCL conveyed a survival benefit (odds ratio; 95% confidence interval: 0.19 (0.05-0.33), p = 0.005).CONCLUSIONThis is the first civilian study that analyses the impact of DCR in patients managed with DCL. During the DCL and DCR study period more PRBC, FFP, and platelets with less crystalloid solution was used intraoperatively. DCL and DCR were associated with a survival advantage and shorter trauma intensive care unit length of stay in patients with severe hemorrhage when compared with DCL and CRE.

BACKGROUNDBase deficit (BD) and lactate are used as markers of mortality, injury severity, and resource utilization in the general trauma population. No study has defined the role of these markers in the triage and management of the normotensive injured elderly patient.METHODSRetrospective cohort study of the trauma registry from a Level I trauma Center during the period of January 1, 2000 through December 31, 2006. Inclusion criteria were age > or = 65 years, initial systolic blood pressure > or = 90 mm Hg; blunt mechanism of trauma. Lactate was categorized as 0 to 2.4 mmol/L (normal), 2.5 to 4.0 mmol/L (moderately elevated), or > 4.0 mmol/L (severely elevated). BD was categorized as > 0 mEq/L (normal), 0 to -6 mEq/L (moderate), or < -6 mEq/L (severe). The primary outcome was inhospital mortality.RESULTSMean lactate was higher in nonsurvivors compared with survivors (2.8 mm/L +/- 1.8 mm/L vs. 2.0 mm/L +/- 1.0 mm/L, p < 0.001). Normal, moderately elevated, and severely elevated lactate was associated with mortality rates of 15% (95% confidence interval [CI] 12-18.8%), 23.4% (95% CI 2-32.4%), and 39.6% (95% CI 26.5-52.8%), respectively. Compared with the normal lactate group, patients in the severely elevated lactate group had 4.2 increased odds of death. BD was more abnormal in nonsurvivors compared with survivors (-2.3 mEq/L +/- 5.2 mEq/L vs. 0.28 mEq/L +/- 1.0 mEq/L, p < 0.001). Normal, moderate, and severe BD were associated with mortality rates of 14% (95% CI 10.3-17.1%), 27% (95% CI 20.1-34.2%), and 40% (95% CI 24.9-54.1%), respectively. Compared with the normal BD group, patients in the severe group had 4.1 increased odds of death.CONCLUSIONSBoth lactate and BD were associated with significantly increased mortality in normotensive elderly blunt trauma patients. However, because of the high baseline mortality rates in elderly trauma patients, "normal" lactate does not offer complete reassurance to the clinician.

BACKGROUND:Although hemostatic resuscitation with a 1:1 ratio of fresh-frozen plasma (FFP) to packed red blood cells (PRBC) after severe hemorrhage has been shown to improve survival, its benefit in patients with traumatic-induced coagulopathy (TIC) after >10 units of PRBC during operation has not been elucidated. We hypothesized that a survival benefit would occur when early hemostatic resuscitation was used intraoperatively after injury in patients with TIC. METHODS:A 7-year retrospective study of patients with emergency department diagnosis of TIC after transfusion of >10 units of PRBC in the operating room. TIC was defined as initial emergency department international normalized ratio > 1.2, prothrombin time > 16 seconds, and partial thromboplastin time > 50 seconds. Patients were divided into FFP:PRBC ratios of 1:1, 1:2, 1:3, and 1:4. Patients with diagnosis of TIC who received transfusion of both FFP and PRBC during surgery were included. Other variables evaluated included age, gender, mechanism of injury, initial base deficit, mean operative time, trauma intensive care unit length of stay (TICU LOS) and Injury Severity Score. The primary outcome measure evaluated was the impact of the early FFP:PRBC ratio on mortality. RESULTS:Four hundred thirty-five patients underwent emergency operations postinjury and received FFP with >10 units of PRBC in the operating room; 135 (31.0%) of these patients had TIC and 53 died (39.5% mortality). Mean operative time was 137 minutes (SD +/- 49). There were no differences with regard to age, gender, mechanism of injury, initial base deficit, or Injury Severity Score among all groups. A significant difference in mortality was found in patients who received >10 units of PRBC when FFP:PRBC ratio was 1:1 versus 1:4 (28.2% vs. 51.1%, p = 0.03). Intermediate mortality rates were noted in patients with 1:2 and 1:3 ratios (38% and 40%, respectively). From a linear regression model, 13 days of increased TICU LOS was observed among 1:4 group compared with 1:1 group (p < 0.01). CONCLUSION:TIC is common after severe injury and is associated with a high mortality in patients transfused with >10 units of PRBC during surgery. Early hemostatic resuscitation during first hours after injury improves survival with shorter TICU LOS in patients with TIC.

Charity Hospital (CH) was devastated by Hurricane Katrina and remains closed. Design and staffing of a new, temporary dedicated trauma hospital relied on data from prior experience at CH, updated census information, and a changed trauma demographic. The study objective was to analyze the new trauma program and evaluate changes in demographics, injury patterns, and outcomes between pre- (PK) and post-Katrina (POK) trauma populations. A retrospective review of trauma patients’ demographics, anatomical variables, and physiological variables 6 months PK and POK was performed under an approved Institutional Review Board protocol. Trauma activation triage criteria between study periods were also analyzed. Continuous data comparisons between the two time periods were made with Student's t test. Dichotomous data were analyzed using χ2 test. The demographic of trauma patients is different in the POK interval, reflecting changes in the New Orleans population. Modification of triage criteria by the exclusion of mechanism as an activation criterion resulted in an increase of patients with higher acuity and Injury Severity Score, lower initial Glasgow Coma Score, and a higher proportion of penetrating mechanism. Outcome measures reflect longer length of stay (4.4 vs 6.8 days, P < 0.0001) without a significant difference in mortality (6.0 vs 7.5, P = 0.227). Hospital data demonstrates that the POK trauma system was stressed by the increased acuity, penetrating injury, and number of procedures per patient (1.7 vs 3.4). Resources should be directed toward patients requiring multidisciplinary care by increasing intensive care unit beds and operating room capacity. Future resource planning in the recovery phases of large-scale natural disasters should take into account these observations.