Head Protection Device for Individuals at Risk for Head Injury due to Ground-Level Falls: Single Trauma Center User Experience Investigation.

BACKGROUND: Falls represent a large percentage of hospitalized patients with trauma as they may result in head injuries. Brain injury from ground-level falls (GLFs) in patients is common and has substantial mortality. As fall prevention initiatives have been inconclusive, we changed our strategy to injury prevention. We identified a head protection device (HPD) with impact-resistant technology, which meets head impact criteria sustained in a GLF. HPDs such as helmets are ubiquitous in preventing head injuries in sports and industrial activities; yet, they have not been studied for daily activities. OBJECTIVE: We investigated the usability of a novel HPD on patients with head injury in acute care and home contexts to predict future compliance. METHODS: A total of 26 individuals who sustained head injuries, wore an HPD in the hospital, while ambulatory and were evaluated at baseline and 2 months post discharge. Clinical and demographic data were collected; a usability survey captured HPD domains. This user experience design revealed patient perceptions, satisfaction, and compliance. Nonparametric tests were used for intragroup comparisons (Wilcoxon signed rank test). Differences between categorical variables including sex, race, and age (age group 1: 55-77 years; age group 2: 78+ years) and compliance were tested using the chi-square test. RESULTS: Of the 26 patients enrolled, 12 (46%) were female, 18 (69%) were on anticoagulants, and 25 (96%) were admitted with a head injury due to a GLF. The median age was 77 (IQR 55-92) years. After 2 months, 22 (85%) wore the device with 0 falls and no GLF hospital readmissions. Usability assessment with 26 patients revealed positive scores for the HPD post discharge regarding satisfaction (mean 4.8, SD 0.89), usability (mean 4.23, SD 0.86), effectiveness (mean 4.69, SD 0.54), and relevance (mean 4.12, SD 1.10). Nonparametric tests showed positive results with no significant differences between 2 observations. One issue emerged in the domain of aesthetics; post discharge, 8 (30%) patients had a concern about device weight. Analysis showed differences in patient compliance regarding age (χ12=4.27; P=.04) but not sex (χ12=1.58; P=.23) or race (χ12=0.75; P=.60). Age group 1 was more likely to wear the device for normal daily activities. Patients most often wore the device ambulating, and protection was identified as the primary benefit. CONCLUSIONS: The HPD intervention is likely to have reasonably high compliance in a population at risk for GLFs as it was considered usable, protective, and relevant. The feasibility and wearability of the device in patients who are at risk for GLFs will inform future directions, which includes a multicenter study to evaluate device compliance and effectiveness. Our work will guide other institutions in pursuing technologies and interventions that are effective in mitigating injury in the event of a fall in this high-risk population.

Surgical stabilization of rib fracture patients versus nonoperative controls treated by a multidisciplinary team in a single institution.

Introduction: Despite promising evidence, surgical stabilization of rib fractures (SSRF) is not ubiquitously offered in all trauma centers. Some centers struggle with patient selection while some struggle due to surgeon comfort with the technique. To address this issue, our trauma center developed a multidisciplinary SSRF approach between orthopedic and trauma surgery. Methods: This retrospective study compared 43 patients who underwent SSRF at a level 1 trauma center with 43 nonoperatively managed controls. Our study Indications were flail chest with >3 segments; non-flail with severe, bi-cortical displacement of >3 contiguous segments. Main outcome measures included mortality, ICU duration, hospital stay LOS, rates of ventilator-associated pneumonia (VAP) and ventilator days. Results: Results of SSRF included decreases in mortality (2% vs 16.3%; p = 0.03) and in ICU duration. Patients with SSRF had a significantly shorter duration in the ICU than the nonoperative group (8.72 vs 14 days; p = 0.013) but a similar hospital duration (LOS mean, 12.81 vs 15.2; p = 0.29). Less patients in the SSRF group developed VAP but the difference was not significant (2% vs 14%, p = 0.055). Discussion: SSRF patient outcomes supported prior evidence. The tandem approach had benefits as surgeons were able to leverage skills and expertise, increase collaboration between services, and complete more difficult reconstructions. Our experience may serve as a model for trauma centers interested in starting a new program or enhancing current service offerings.

Compliance Prediction of a Novel Head Protection Device for Elderly Patients With Fall Risk.

Brain injury from ground level falls (GLF) is common and has substantial morbidity. We identified a potential head protection device (HPD). This report describes predicted future compliance. 21 elderly patients were provided a HPD and were evaluated on admission and after discharge. Compliance, ease of use, and comfort were evaluated. Differences between categorical variables (gender, race, age group1, 55-77 years; group2, 78+) and compliance were tested using the chi-squared statistic. HPD compliance at baseline was 90% with 85% at follow-up (P = .33). No difference with HPD interaction (P = .72), ease of use (P = .57), and comfort (P = .77). Weight was a concern on follow-up (P = .001). Age group1 was more compliant (P=.05). At two months, patients were compliant with no falls recorded. The identified HPD with modifications has a high predicted compliance in this population. After the device is modified, effectiveness will be assessed.

Shift Length and Shift Length Preference Among Acute Care Surgeons.

BACKGROUND: Work hour restrictions have been imposed by the Accreditation Council for Graduate Medical Education since 2003 for medical trainees. Many acute care surgeons currently work longer shifts but their preferred shift length is not known. METHODS: The purpose of this study was to characterize the distribution of the current shift length among trauma and acute care surgeons and to identify the surgeons' preference for shift length. Data collection included a questionnaire with a national administration. Frequencies and percentages are reported for categorical variables and medians and means with SDs are reported for continuous variables. A chi-square test of independence was performed to examine the relation between call shift choice and trauma center level (level 1 and level II), age, and gender. RESULTS: Data from 301 surgeons in 42 states included high-level trauma centers. Assuming the number of trauma surgeons in the United States is 4129, a sample of 301 gives the survey a 5% margin of error. The median age was 43 years (M = 46, SD = 9.44) and 33% were female. Currently, only 23.3% of acute care surgeons work a 12-hour shift, although 72% prefer the shorter shift. The preference for shorter shifts was statistically significant. There was no significant difference between call shift length preference and trauma center level, age, or gender. CONCLUSION: Most surgeons currently work longer than 12-hour shifts. Yet, there was a preference for 12-hour shifts indicating there is a gap between current and preferred shift length. These findings have the potential to substantially impact staffing models.

Multiple gallstones causing ileus twenty years after cholecystectomy.

Gallstone ileus is an important form of small bowel obstruction that occurs in less than 0.5% of patients who present with obstruction. A biliary enteric fistula that evolves in the setting of chronic cholecystitis may allow the passage of a large gallstone into the gastrointestinal tract distal to the common duct. A single stone that is sufficient in size (at least 2-2.5 cm diameter) may then create a mechanical obstruction, most often at the ileocecal valve or the terminal ileum where the intestinal lumen narrows, and where peristalsis is less robust. We present an unusual case of gallstone ileus in a patient whose obstruction was caused by not one, but seven individual gallstones, collectively restricted in the jejunum at the point of a previous anastomosis and occurring twenty years after cholecystectomy.

Delayed Intracranial Hemorrhage After Blunt Head Injury With Direct Oral Anticoagulants.

INTRODUCTION: Patients presenting to the Emergency Department (ED) following head injury are frequently evaluated with an initial computed tomography scan (CT) of the brain. Imaging is particularly important in patients who are receiving medications that alter normal blood hemostasis. As an imaging modality, CT has a high negative predictive value when used to rule out clinically significant acute intracranial hemorrhage. Patients receiving anticoagulant or antiplatelet therapy have both an increased risk of initial hemorrhage, as well as an increased risk of mortality above nonanticoagulated patients, should they suffer hemorrhage. Multiple studies of delayed intracranial hemorrhage have placed the risk among the patients taking warfarin at the time of head injury in the range of 0.6-6.0%. However, data regarding the risk of delayed intracranial hemorrhage in patients taking the class of agents referred to as Direct-Acting Oral Anticoagulants (DOACs) remains limited. This study aims to estimate this risk. METHODS: A retrospective chart review was performed to identify patients on DOACs who presented to our Level I trauma center following blunt head injury between January 2017 and August 2018. Patients with a negative initial head CT were selected. From this subset, data regarding demographics, injury characteristics, anticoagulant use, and antiplatelet use were collected. RESULTS: Overall, 314 patients were included; 129 patients taking rivaroxaban, 182 patients taking apixaban, and four patients taking dabigatran. In approximately 29% of the patients, the sole indication for admission was close monitoring following head injury while taking an anticoagulant agent. The mechanism of injury for the majority of the patients was fall. Of the 314 patients, three were found to have delayed intracranial hemorrhage on the repeated head CT (0.95%). Two of these three patients were on concomitant antiplatelet medication. None of the three individuals required neurosurgical intervention. CONCLUSIONS: at the time of submission, this is the largest study estimating the risk of delayed intracranial hemorrhage among patients on DOACs. Based on the results of this study, patients who sustain a blunt head injury while taking only DOACs; that is, without concurrent antiplatelet medication, admission, and repeat head CT may not be necessary after confirming a negative initial CT scan.

Use of a Dog Visitation Program to Improve Patient Satisfaction in Trauma Patients.

Clinical staff members all recognize the importance of attaining high patient satisfaction scores. Although there are many variables that contribute to patient satisfaction, implementation of a dog visitation program has been shown to have positive effects on patient satisfaction in total joint replacement patients. This innovative practice had not previously been studied in trauma patients. The purpose of this quasi-experimental study was to determine whether dog visitation to trauma inpatients increased patient satisfaction scores with the trauma physicians. A team consisting of a dog and handler visited 150 inpatients on the trauma service. Patient satisfaction was measured using a preexisting internal tool for patients who had received dog visitation and compared with other trauma patients who had not received a visit. This study demonstrated that patient satisfaction on four of the five measured scores was more positive for the patients who had received a dog visit.

The Michigan Trauma Quality Improvement Program: Results from a collaborative quality initiative.

BACKGROUND: American College of Surgeons verified trauma centers and a third-party payer within the state of Michigan built a regional collaborative quality initiative (CQI). The Michigan Trauma Quality Improvement Program began as a pilot in 2008 and expanded to a formal program in 2011. Here, we examine the performance of the collaborative over time with regard to patient outcomes, resource utilization, and process measures. METHODS: Data from the initial 23 hospitals that joined the CQI in 2011 were analyzed. Performance trends from 2011 to 2015 were evaluated for outcomes, resource utilization, and process measures using univariate analysis. Risk-adjustment was performed to confirm results observed in the unadjusted data. To calculate the potential number of patients impacted by the CQI program, the maximum absolute change was multiplied by the number of trauma patients treated in the 23 hospitals during 2015. RESULTS: Membership in a CQI program significantly reduced serious complications (8.5 vs. 7.3%, p = 0.002), decreased resource utilization, and improved process measure execution in trauma patients over 5 years time. Similar results were obtained in unadjusted and risk-adjusted analyses. The CQI program potentially avoided inferior vena cava filter placement in 167 patients annually. Decreased venous thromboembolism rates mirrored increased compliance with venous thromboembolism pharmacologic prophylaxis. CONCLUSION: This study confirms our hypothesis that participation in a regional CQI improves patient outcomes and decreases resource utilization while promoting compliance with processes of care. LEVEL OF EVIDENCE: Economic/therapeutic care, level V.

Prophylactic Inferior Vena Cava Filter Placement Does Not Result in a Survival Benefit for Trauma Patients.

OBJECTIVE: Trauma patients are at high risk for life-threatening venous thromboembolic (VTE) events. We examined the relationship between prophylactic inferior vena cava (IVC) filter use, mortality, and VTE. SUMMARY BACKGROUND DATA: The prevalence of prophylactic placement of IVC filters has increased among trauma patients. However, there exists little data on the overall efficacy of prophylactic IVC filters with regard to outcomes. METHODS: Trauma quality collaborative data from 2010 to 2014 were analyzed. Patients were excluded with no signs of life, Injury Severity Score <9, hospitalization <3 days, or who received IVC filter after occurrence of VTE event. Risk-adjusted rates of IVC filter placement were calculated and hospitals placed into quartiles of IVC filter use. Mortality rates by quartile were compared. We also determined the association of deep venous thrombosis (DVT) with the presence of an IVC filter, accounting for type and timing of initiation of pharmacological VTE prophylaxis. RESULTS: A prophylactic IVC filter was placed in 803 (2%) of 39,456 patients. Hospitals exhibited significant variability (0.6% to 9.6%) in adjusted rates of IVC filter utilization. Rates of IVC placement within quartiles were 0.7%, 1.3%, 2.1%, and 4.6%, respectively. IVC filter use quartiles showed no variation in mortality. Adjusting for pharmacological VTE prophylaxis and patient factors, prophylactic IVC filter placement was associated with an increased incidence of DVT (OR = 1.83; 95% CI, 1.15-2.93, P-value = 0.01). CONCLUSIONS: High rates of prophylactic IVC filter placement have no effect on reducing trauma patient mortality and are associated with an increase in DVT events.

Results of a clinical practice algorithm for the management of thoracostomy tubes placed for traumatic mechanism.

INTRODUCTION: The management and removal of thoracostomy tubes for trauma-related hemothorax and pneumothorax is controversial. General recommendations exist; however, institutional data related to an algorithmic approach has not been well described. The difficulty in establishing an algorithm centers about individualized patients' needs for subsequent management after thoracostomy tube placement. In our institution, we use the same protocol for all trauma patients who receive a thoracostomy tube with minimal complications. PURPOSE: To present the clinical outcomes of patients who required a tube thoracostomy for traumatic injury and were managed by an institutional protocol. METHODS: A retrospective chart review of 313 trauma patients at a single level I trauma institution from January 2008 through June 2012 was conducted. Inclusion criteria were patient age ≥ 18 years, involvement in a trauma, and requirement of a thoracostomy tube. The patients' charts were reviewed for demographic data, injury severity score (ISS), length of stay (LOS), and chest-tube specific data. Thoracostomy tube complications were defined as persistent air leak, persistent pneumothorax, recurrent pneumothorax, and clotting of thoracostomy tube. The patients were managed per our institutional algorithm. Descriptive statistics were performed. RESULTS: Most of the patients who required a thoracostomy tube had blunt-related traumas (271/313; 86.6%), while 42 patients (13.4%) sustained penetrating injuries. There were 215 (68.7%) male patients. The average age at time of injury was 45.7 ± 21.1 years and the mean ISS was 24.9 ± 15.9 (mean ± SD). Elevated alcohol levels were found in 65 of the 247 patients who were tested upon admission (26.3%). Overall, 15 patients (4.8%) developed a thoracostomy tube related complication: persistent air leak in six patients, persistent pneumothorax in six patients, recurrent pneumothorax in two patients, and clotted thoracostomy tube in one patient. The average LOS was 10.4 ± 8.4 days, and the mean length of thoracostomy tube placement was 5.9 ± 4.3 days. CONCLUSIONS: Our algorithmic thoracostomy tube management protocol resulted in a complication rate of 4.8%. By managing thoracostomy tubes in a systematic manner, our patients have improved outcomes following placement and removal compared to other studies.

An 8-factor prognostic model predicted early death in patients with traumatic bleeding.

QUESTION: Does a prognostic model accurately estimate risk for early death in patients with traumatic bleeding? DESIGN: 2 cohort studies: Clinical Randomisation of an Antifibrinolytic in Significant Haemorrhage (CRASH-2) trial (derivation) and Trauma Audit and Research Network (TARN) dataset (validation). SETTING: 274 hospitals in 40 countries (derivation cohort) and 60% of hospitals that received trauma patients in England and Wales and some hospitals in Europe (validation cohort). PATIENTS: Derivation cohort: 20 127 trauma patients with, or at risk for, significant bleeding within 8 hours of injury (median age 30 y, median 2 h from injury). Validation cohort: 14 220 patients > 15 years of age (median age 39 y, median 1 h from injury) who arrived at hospital alive and had ≥ 1 of death from injury during admission, hospitalization > 3 days, need for intensive or high-dependency care, or need for interhospital transfer for specialist care. Patients who had isolated closed limb injuries or were > 65 years of age with isolated fractured neck of femur or pubic ramus fracture were excluded. DESCRIPTION OF PREDICTION GUIDE: The prognostic model included country (low, middle, or high income), age, time since injury, Glasgow coma score, systolic blood pressure, respiratory rate, heart rate, and type of injury (blunt or penetrating) (available at www.crash2.lshtm.ac.uk). OUTCOME: Early death (in-hospital death within 4 wk of injury). MAIN RESULTS: 15% and 12% of patients from the derivation and validation cohorts died. The operating characteristics of the model in the derivation and validation cohorts are in the Table. CONCLUSION: An 8-factor prognostic model predicted early death in patients with traumatic bleeding.

Chemical venous thromboembolic prophylaxis is safe and effective for patients with traumatic brain injury when started 24 hours after the absence of hemorrhage progression on head CT.

BACKGROUND: Venous thromboembolism (VTE) continues to be an important complication for patients with trauma, including patients with intracranial hemorrhage. We implemented a protocol starting chemical prophylaxis 24 hours after the absence of progression of hemorrhage on computed tomography (CT) to increase consistency with the use of chemical venous thromboembolic prophylaxis in this population. The objective of this study was to review the protocol of VTE prophylaxis for patients with traumatic brain injury at our institution to determine whether it has been effective and safe in preventing VTE without increasing intracranial hemorrhage. METHODS: A retrospective case series was conducted to study 205 patients with intracranial hemorrhage admitted to a Level I trauma center during a 24-month period. These patients were reviewed with respect to type of intracranial injury, need for surgery, injury severity, time to initiation of chemical prophylaxis, and progression of injury on brain CT. Patients with a hospital length of stay less than 3 days or nonstable CT were excluded in the analysis of administration of chemical prophylaxis. Time to chemical prophylaxis in relation to absence of progression on brain CT was examined as well as the subsequent risk of progression of hemorrhage and risk of VTE events. The overall rate of venous thromboembolism was compared with that of matched historical controls. RESULTS: All patients received mechanical prophylaxis in the form of sequential compression devices. One hundred sixty-two intracranial hemorrhages were identified in 122 patients who met the study's inclusion criteria. Of this group of patients who did not have progression of hemorrhage on follow-up CT, 76.2% received chemical prophylaxis during their hospitalization.No patients had progression of intracranial hemorrhage after initiation of chemical VTE prophylaxis, and no patients developed VTE. This represents a decrease of VTE from previous years. No other complications related to chemical VTE prophylaxis were identified. CONCLUSION: A protocol based on an early use of chemical venous thromboembolic prophylaxis after the absence of progression of tramatic intracranial hemorrhage does not result in increased progression of intracranial hemorrhage and reduced the rate of venous thromboembolic events at our institution.

Disorganized care: the findings of an iterative, in-depth analysis of surgical morbidity and mortality.

INTRODUCTION: Performance improvement driven by the review of surgical morbidity and mortality is often limited to critiques of individual cases with a focus on individual errors. Little attention has been given to an analysis of why a decision seemed right at the time or to lower-level root causes. The application of scientific performance improvement has the potential to bring to light deeper levels of understanding of surgical decision-making, care processes, and physician psychology. METHODS: A comprehensive retrospective chart review of previously discussed morbidity and mortality cases was performed with an attempt to identify areas where we could better understand or influence behavior or systems. We avoided focusing on traditional sources of human error such as lapses of vigilance or memory. An iterative process was used to refine the practical areas for possible intervention. Definitions were then created for the major categories and subcategories. RESULTS: Of a sample of 152 presented cases, the root cause for 96 (63%) patient-related events was identified as uni-factorial in origin, with 51 (34%) cases strictly related to patient disease with no other contributing causes. Fifty-six cases (37%) had multiple causes. The remaining 101 cases (66%) were categorized into two areas where the ability to influence outcomes appeared possible. Technical issues were found in 27 (18%) of these cases and 74 (74%) were related to disorganized care problems. Of the 74 cases identified with disorganized care, 42 (42%) were related to failures in critical thinking, 18 (18%) to undisciplined treatment strategies, 8 (8%) to structural failures, and 6 (6%) were related to failures in situational awareness. CONCLUSIONS: On a comprehensive review of cases presented at the morbidity and mortality conference, disorganized care played a large role in the cases presented and may have implications for future curriculum changes. The failure to think critically, to deliver disciplined treatment strategies, to recognize structural failures, and to achieve situational awareness contributed to the morbidities and mortalities. Future research may determine if focused training in these areas improves patient outcomes.

Tigecycline penetration into skin and soft tissue.

BACKGROUND: Tigecycline, a derivative of minocycline, has antibacterial activity against common pathogens associated with complicated skin and soft tissue infections (cSSTIs), including methicillin-resistant Staphylococcus aureus. At present, there is a paucity of data concerning its penetration into skin and soft tissue (SST). METHODS: This study evaluated the penetration of tigecycline into SST in 25 patients (mean age, 52 years) with cSSTIs requiring surgical intervention. After a 100-mg loading dose, each patient received 50 mg of tigecycline infused intravenously over 1 h every 12 h. Blood samples were obtained on the day of surgery at 1 h (peak), during surgery, and 12 h (trough) after the beginning of a 50-mg infusion. A viable SST sample was harvested at the infection site. Tissue and concomitant serum concentrations were grouped into three time intervals: 2-4 h (median, 3 h), 5-7 h (median, 7 h), and 8-10 h (median, 9h), and analyzed for tissue penetration. RESULTS: Tissue and blood samples were obtained one to six days (mean 2.5 days) after initiation of tigecycline treatment. The mean serum peak and trough concentrations of tigecycline were 0.56±0.25 mg/L and 0.26±0.12 mg/L, respectively. The mean tissue:serum ratios at the three study time periods were 3.8 (range 0.7-5.5), 5.2 (range 0.8-7.1), and 2.8 (range 0.8-8.8). CONCLUSIONS: In general, we found higher concentrations of tigecycline in SST than in the serum at the same time point.

Parental recognition of postconcussive symptoms in children.

Evaluation of current emergency department discharge instructions and parental recognition of symptomatology requiring further care for traumatic brain injury (TBI) is not well understood. A convenience sample of 105 parents of children aged 5 to 17 years who were seen and discharged from the pediatric emergency department with TBI was identified. Parents were surveyed by telephone 2 to 5 days after injury and a questionnaire was completed regarding identification of TBI symptoms. This study demonstrated that despite verbal and written discharge instructions, many parents with symptomatic children reported that their children were asymptomatic, and unable to identify postconcussive symptoms in their children.

Reduction of radiation dose in pediatric brain CT is not associated with missed injuries or delayed diagnosis.

Increased accuracy of CTs in the identification of traumatic injuries compared with physical examination or conventional radiography is well documented. Our goal was to identify the most effective strategy for decreasing radiation exposure while retaining the benefits of computerized imaging. Based on a literature review and our trauma registry, the mortality risk of untreated injuries was compared with that of patients who received treatment of injuries diagnosed by CT. Because automated exposure control of tube current is not routinely used with brain CT, this region was identified as the initial focus for a dose-saving algorithm. CT settings were adjusted for children studies and the new settings were implemented into four protocols based on age. Images were compared and reviewed by radiologists for the ability to identify traumatic injuries. Effective dose (ED) was estimated using Monte Carlo simulations. The lifetime incidence and mortality for thyroid cancer and leukemia were assessed. In-hospital mortality of unidentified injury in trauma patients is 8.0%. Forty dose-saving CTs were performed and no injuries were missed. The ED decreased by 5.2-, 4.5-, 2.62-, and 2.5-fold in each group. Decreasing the ED is achievable, theoretically decreases the cancer risk and does not increase the missed injury rate.

A disciplined approach to implementation of evidence-based practices decreases ICU and hospital length of stay in traumatically injured patients.

BACKGROUND: Evidence-based medicine has gained wide acceptance in practice of medicine since the 1990s. The objective of our study was to demonstrate the effect of evidence-based critical care practices on ICU and hospital length of stay in mechanically ventilated trauma patients. MATERIALS AND METHODS: Retrospective cohort using historic controls. During 2004, several different evidence-based practices were implemented, including low tidal volume ventilation, protocol driven trauma resuscitation, and a sepsis bundle. Outcomes in critically ill, mechanically ventilated patients who were ≥ 18 y old were compared between a historic control group (2000-2003) and the study group after implementation (2005-2008). Patients were identified using the institutional trauma registry (NATIONAL TRACS). Gender, age, ISS, mechanism of injury, and mortality were also examined to identify trends in epidemiology. RESULTS: From 2000 to 2003. there were 6920 trauma admissions and during 2005-2008 there were 8911 (increase of 28.8%). These included 217 and 337 (increase of 55.3%) admissions to the ICU of mechanically ventilated patients, respectively. The mean age was 43.9 y versus 45.9 y (P = 0.258). Males were 66.4% versus 71.8% (P = 0.610). The mean ISS was 29 versus 27 (P = 0.25). Blunt mechanism was 87% versus 89% (P = 0.913). Mortality rate was 36.4% versus 36.5% (P = 0.944). The mean number of ICU days and hospital days decreased from 7.6 versus 5.5 (P = 0.02) and 13.2 versus 9.7 (P = 0.03), respectively. CONCLUSION: The application of evidence-based critical care practices decreases length of ICU and hospital stay, but not mortality, in critically ill, mechanically ventilated trauma patients. Our trauma volume, including critically ill patients, increased during the study periods.