Traumatic brain injury: patterns of failure of nonoperative management.

OBJECTIVE: The circumstances of failure for nonoperative management of blunt traumatic brain injury have been poorly defined. In this study, all trauma patients identified over a 12-year period with progression of neurologic injury requiring craniotomy were retrospectively reviewed. METHODS: Data collected included demographic information, mechanism of injury, field and admission vital signs, and Glasgow Coma Scale score, medications, associated injuries, and coagulopathy. Head computed tomographic scans were reviewed, and anatomic findings were correlated with clinical changes (change in mental status or elevation of intracranial pressure) that led to subsequent CT scan and craniotomy. RESULTS: Of 20,100 patients, there were 852 who had computed tomographic scans with acute intracranial injuries on admission; 462 patients were managed nonoperatively. Fifty-seven patients had progression of neurologic injury (34 < 24 hours = early; 23 > 24 hours = late) that required surgery. CONCLUSION: Of the variables investigated, only anatomic location of injury was found to be predictive of early failure of nonoperative management. Frontal intraparenchymal hematomas are particularly prone to early failure. Clinical examination and intracranial pressure monitoring are equally important in detecting failure and should be an integral part of nonoperative management.

Trauma in pregnancy: maternal and fetal outcomes.

OBJECTIVE: Pregnancy imposes significant physiologic demands that may confuse and complicate the evaluation, resuscitation, and definitive management of pregnant women who sustain trauma. Accurate prediction of fetal outcome after trauma remains elusive. The objective of this study was to characterize patterns of injury in pregnant women, to determine if pregnancy affects maternal morbidity and mortality after trauma, and to identify predictors of fetal death. METHODS: We performed a retrospective, case-control analysis of all injured pregnant patients admitted to the Trauma Service at the University of California San Diego Medical Center from 1985 to 1995. RESULTS: We identified 114 injured pregnant patients. Motor vehicle crashes accounted for 70% of injuries, and of these, 46% of patients were not using seat belts or helmets. Violence accounted for 12% of injuries. Injured pregnant women with Injury Severity Scores > 8 demonstrated similar mortality, morbidity, and length of stay to matched nonpregnant control patients. Pregnant women were more likely to sustain serious abdominal injury and were less likely to sustain severe head injury. Identified risk factors for fetal loss include maternal death, overall maternal injury severity, the presence of severe abdominal injury, and the presence of hemorrhagic shock. CONCLUSION: There appears to be a group of pregnant women in San Diego at high risk for traumatic injury who should be targeted for preventative strategies including improved seat belt use. Pregnancy does not increase mortality or morbidity after trauma but influences the pattern of injury. Maternal death, high Injury Severity Score, serious abdominal injury, and hemorrhagic shock are risk factors for fetal loss.

Lethal injuries and time to death in a level I trauma center.

BACKGROUND: The purpose of this study was to identify the causes and time to death of all trauma victims who died at a level I trauma center during an 11-year period. STUDY DESIGN: Autopsies were performed on all patients who died secondary to trauma. Retrospective review of these autopsies was carried out and appended to existing trauma registry data. Standard definitions were used to attribute the cause of death in each case. Preventable deaths were determined by a standardized peer review process. RESULTS: Between January 1985 and December 1995, a total of 900 trauma patients died. This represented 7.3% of all major trauma admissions (12,320). Seventy percent of these patients died within the first 24 hours of admission. Thoracic vascular and central nervous system (CNS) injuries were the most common causes of death in the first hour after admission to the hospital. CNS injuries were the most common causes of death within the 72 deaths after admission. Acute inflammatory processes (multiple organ failure, acute respiratory distress syndrome, and pneumonia) and pulmonary emboli were the leading causes of death after the first 72 hours. Overall, 43.6% (393 of 900) of all trauma deaths were caused by CNS injuries, making this the most common cause of death in our study. The preventable death rate was 1%. CONCLUSIONS: The first 24 hours after trauma are the deadliest for these patients. Primary and secondary CNS injuries are the leading causes of death. Prevention, early identification, and treatment of potentially lethal injuries should remain the focus of those who treat trauma patients.

Is operating room resuscitation a way to save time?

BACKGROUND: Direct admission to the operating room (OR) can shorten the time to incision. A protocol for operating room resuscitation was established with patient triage based on (1) cardiac arrest, (2) hypotension unresponsive to field fluid resuscitation, or (3) uncontrolled external hemorrhage. METHODS: Operating room resuscitation over 11 years was reviewed to determine whether the triage criteria correctly identified patients requiring operation. Survival was analyzed and compared with the probability of survival (Ps) determined at the scene. RESULTS: Operating room resuscitation patients were more likely to require a major operation regardless of mechanism of injury. Of 476 patients with penetrating injury, 170 patients had persistent low blood pressure (<90 mm Hg), and 146 (85.9%) of these required major operative intervention. The mean time to incision in this group was 21.7-67.5 minutes less than for patients not receiving OR resuscitation. Observed survival was significantly greater than that predicted for this group. CONCLUSIONS: Field triage criteria are able to reliably identify patients who require immediate major operative intervention. Direct admission to the OR results in a more timely initiation of operative therapy for patients requiring emergency surgical procedures.

Endotracheal intubation in the field improves survival in patients with severe head injury. Trauma Research and Education Foundation of San Diego.

OBJECTIVE: To measure the effect of prehospital endotracheal intubation on outcome in patients with severe head injury and the percentage of these patients intubated in the field under existing protocol. DESIGN: Retrospective case-control study. SETTING: Countywide urban trauma system. PATIENTS: Trauma patients with blunt injury and scene Glasgow Coma Score of 8 or less, transported by ground ambulance with advanced life support capabilities from January 1, 1991, to December 31, 1995. Severe head injury was defined as head or neck Abbreviated Injury Scale score of 4 or greater. Isolated severe head injury was defined as head or neck Abbreviated Injury Scale score of 4 or greater with no other Abbreviated Injury Scale component greater than 3. One thousand ninety-two patients met initial criteria; of these, 671 had severe head injury, and 351 had isolated severe head injury. INTERVENTIONS: None. MAIN OUTCOME MEASURES: Mortality and functional status sufficient for discharge to home. RESULTS: Field intubation was associated with significant decreases in mortality from 36% to 26% in the full study group, from 57% to 36% in patients with severe head injury, and from 50% to 23% in patients with isolated severe head injury. Rate of discharge to home was unaffected by field intubation. Between 50% and 60% of study patients were intubated under current paramedic protocol, compared with intubation rates of 85% to 92% for similar patients transported by aeromedical teams operating under expanded indications for intubation. CONCLUSIONS: Prehospital endotracheal intubation was associated with improved survival in patients with blunt injury and scene Glasgow Coma Score of 8 or less, especially those with severe head injury by anatomic criteria. Broadening indications for intubation by paramedical personnel has great potential to improve outcome in patients with severe head injury.

Analysis of heart-rate variability: a noninvasive predictor of death and poor outcome in patients with severe head injury.

BACKGROUND: Analysis of heart-rate variability (HRV) is a promising new technique for noninvasive quantification of autonomic function. We measured HRV in patients with severe head injury to assess its potential as a monitoring tool. METHODS: Analysis of HRV was prospectively done on all intensive care unit patients. Concurrent data on intracranial pressure (ICP) and cerebral perfusion pressure (CPP) were collected. Registry data were reviewed to identify patients with severe head injury, defined as Head/Neck Abbreviated Injury Scale score > or = 4. Mortality, likelihood of discharge to home, ICP, and CPP were compared between patients with abnormal HRV and those without. RESULTS: Low HRV was associated with increased mortality and decreased rate of discharge to home. Abnormal HRV was associated with episodes of increased ICP and decreased CPP. CONCLUSION: Assessment of HRV is a noninvasive method that can be widely used. Abnormal HRV was associated with poor outcome and altered cerebral perfusion. Monitoring of HRV may improve outcome by allowing earlier detection and treatment of intracranial pathology.

Elevated selectin levels after severe trauma: a marker for sepsis and organ failure and a potential target for immunomodulatory therapy.

UNLABELLED: Severe injury is frequently complicated by sepsis and organ failure. Activated neutrophils adherent to inflamed endothelia have been implicated in the pathogenesis of these complications. Identification of high-risk patients to target immunomodulatory therapy, however, remains an elusive goal. We postulated that (1) patients at risk for sepsis and organ failure could be identified by measuring shed selectin adhesions molecules as a marker of endothelial activation after injury and reperfusion, and (2) these elevated selectin levels would correlate with injury severity, shock, major complications, and mortality. METHODS: Blood samples were drawn from 50 patients with multiple trauma every 2 hours after admission for the first 24 hours, and every 6 hours for the subsequent 24 hours, and assayed for levels of shed E- and P-selectin. Patients were then stratified according to Injury Severity Score (ISS), presence or absence of shock, presence or absence of organ failure and/or infectious complications, and finally, death versus survival. RESULTS: Trauma patients who had ISS < 30, who did not develop shock, sepsis, or organ dysfunction, had minimal increase in circulating E- and P-selectin over admission levels. In patients who subsequently developed infectious complications, organ dysfunction, or both, or subsequently went on to die, elevated levels of E-selectin levels were evident by 36 hours, and in some cases, earlier. Differences between nonsurvivors and survivors was statistically significant. There was also a trend toward increased levels of P-selectin in the same group of patients, although these differences were not significant. There was no differentiation in either of the two selections when patients were stratified according to ISS or presence of shock. CONCLUSION: A subset of major trauma patients manifest increased levels of circulating E-selectin adhesion molecules after resuscitation. These patients seem to be at increased risk of death and possibly at risk for infections complications and organ failure. Selectin blockade is a potential new immunomodulatory strategy in this subgroup of patients.

A prospective evaluation of abdominal ultrasound in blunt trauma: is it useful?

OBJECTIVE: The purpose of this study is to evaluate the utility and feasibility of abdominal ultrasound (US) in blunt trauma patients. DESIGN: This prospective study examined the operational issues and the diagnostic accuracy of US in selected blunt trauma patients triaged to a Level 1 trauma center. MATERIALS AND METHODS: All patients were evaluated by an attending trauma surgeon and our usual criteria for objective evaluation of the abdomen were applied. US was performed by US technicians and interpreted by the trauma surgeon. We prospectively evaluated the availability (time to arrival), the ease with which the US could be integrated into the resuscitation (minutes to start after arrival), and the time required to perform the study. The US results were compared to diagnostic peritoneal lavage and computed tomography findings, clinical course, operative findings, and to repeat US examinations to determine sensitivity, specificity, and usefulness. MEASUREMENTS AND MAIN RESULTS: A total of 800 US studies were performed over 15 months. In four cases (0.5%), the US was incomplete for technical reasons. The results in the remaining 796 studies were as follows: [table: see text] The average time to arrival of the US was 17.3 minutes (range 0-120) and the average minutes to start after arrival was 7.0 (range 1-49). The average time required to perform the study was 10.6 minutes (range 2-26). CONCLUSIONS: This study demonstrates that US can be obtained rapidly, integrated into the resuscitation, and completed quickly. US provides a highly accurate, noninvasive method to evaluate the abdomen in the blunt trauma patient, and has supplanted the previously used methods at this institution.

Spectral analysis of heart rate variability in the ICU: a measure of autonomic function.

Beat-to-beat heart rate variability (HRV) is a measure of autonomic nervous system activity, which can be quantified using frequency domain analysis. Despite its potential utility, routine serial analysis of HRV in an ICU setting has rarely been attempted. We have developed an automated system for real-time spectral analysis of HRV and have utilized this system to study the effect of alterations in HRV on mortality in a surgical ICU population. HRV measurements were performed every 6 hr on all patients in the ICU. Total spectral power in the variability signal (TP, a measure of overall autonomic activity) and the ratio of high frequency to low frequency components (HF/LF ratio, a measure of parasympathetic/sympathetic balance) were calculated. Over a 6-month period 7994 automated HRV measurements were made in 742 patients. Both low TP (low autonomic tone) and high HF/LF ratio (relative lack of sympathetic tone) were associated with increased mortality. A low HF/LF ratio (relatively high sympathetic tone) was associated with increased survival, especially in patients with low autonomic tone. We conclude that serial spectral analysis of HRV is practical in an ICU setting and that HRV parameters appear to be a clinically relevant indication of autonomic activity. Low sympathetic tone and vagal predominance are associated with increased mortality, while sympathetic predominance favors survival. Monitoring of HRV parameters has the potential to detect physiologic deterioration or response to therapy.

Transient systolic hypotension. A serious problem in the management of head injury.

OBJECTIVE: To determine the frequency and clinical impact of transient systolic hypotension (systolic blood pressure < 100 mm Hg) in patients with severe anatomic head injury. DESIGN: Retrospective case-control study. SETTING: Urban level 1 trauma center. PATIENTS: Consecutive trauma patients admitted to the intensive care unit (ICU) with severe anatomic head injury, defined as Head and Neck Abbreviated Injury Scale Score of 4 or higher. One thousand thirteen trauma patients were admitted to the ICU during the study period, 157 of whom met inclusion criteria. MAIN OUTCOME MEASURES: Acute mortality, defined as death during initial ICU admission, and functional status of ICU survivors, assessed as level of function sufficient for discharge to home. RESULTS: One hundred fifty-seven patients with severe head injury had a total of 831 episodes of systolic hypotension. Fifty-five percent of the patients suffered at least one event. Patients were grouped by total number of low systolic blood pressure events and by average number of events per ICU day. The total number of hypotensive events was associated with increased mortality rates and decreased rate of discharge to home. Average daily frequency of events was associated with increased mortality rates. After stratification by admission Glasgow Coma Scale score, the effects were most dramatic in patients with an initial Glasgow Coma Scale score higher than 8. CONCLUSIONS: Transient hypotension is common in the ICU and is associated with increased acute mortality and decreased functional status in patients with head injury. The impact of this secondary insult is greatest in patients with less severe primary injury. Strict avoidance of hypotension through enhanced monitoring and active treatment appears to be important, especially in patients with higher presenting Glasgow Coma Scale scores.

A risk analysis of stress ulceration after trauma.

Prophylaxis for stress ulceration is considered standard care for most critically ill patients, but may be overutilized. We determined the incidence of stress ulceration in 33,637 major trauma patients treated in a regionalized trauma system from 1985 to 1991 using trauma registry data and chart review. Injury-related risk factors for stress ulceration and other associated infectious and organ failure complications were identified by regression analysis. Clinical stress ulceration developed in 57 patients (0.17%) despite prophylaxis. Eighteen patients (0.05%) developed severe ulceration with either gastroduodenal perforation (3 patients) or a > 2 U blood transfusion requirement (16 patients). Independent risk factors with odds ratios (OR) were identified as follows: Injury Severity Score (ISS) > or = 16, OR = 12.6; spinal cord injury, OR = 2.0; and age > 55, OR = 2.4. Other serious complications, including pneumonia, sepsis, and organ failure (adult respiratory distress syndrome and renal and hepatic failure), were significantly associated with the development of stress ulceration. Clinically significant stress ulceration after trauma is uncommon, but occurs despite prophylaxis. Severe injury (ISS > 16) and spinal cord injury were identified as independent injury-related risk factors. All patients with severe ulceration had either one of these injury-related risk factors or a significant infectious complication or organ failure. Standard prophylaxis may be inadequate in high-risk patients, who should be targeted for increased surveillance and aggressive prophylaxis. On the other hand, routine prophylaxis in low-risk patients may be overutilized.

Use of computed tomography of the head in the hypotensive blunt-trauma patient.

STUDY OBJECTIVE: To identify patients presenting with hypotension due to blunt trauma who should undergo computed tomography (CT) of the head before urgent chest or abdominal operation. DESIGN: Retrospective registry-based record review. SETTING: Urban Level I trauma center. PARTICIPANTS: Consecutive trauma patients with suspected head injury, blunt mechanism of injury, and hypotension who were discharged between January 1, 1989, and December 31, 1991. Patients who were dead on arrival or died within 15 minutes of arrival were judged unsalvageable and excluded. Review of 3,224 trauma patients identified 212 as the study population. INTERVENTIONS: Frequency of neurosurgical intervention or general surgical intervention within 6 hours of admission and the time required for completion of CT of the head were noted. RESULTS: Overall, 40 general surgical operations (19%) and 16 craniotomies (8%) were performed, with a mortality rate of 18%. Patients with Glasgow Coma Scale scores of less than 8 had a 19% rate of craniotomy, and those with scores between 8 and 13 had a 9% rate. Sixteen patients had CT before surgery, with an average delay of 68 minutes. No patient who responded to initial resuscitation experienced hemodynamic instability in the CT suite, including 15 patients with positive diagnostic peritoneal lavage. CONCLUSION: CT scan of the head before general surgical operation appears to be safe in patients who respond to initial resuscitation. The likelihood of craniotomy in patients with Glasgow Coma Scale scores of 13 or less is comparable to the likelihood of general surgical operation. Physicians should be encouraged to make CT of the head a high priority in this group.

Risk factors associated with pulmonary embolism despite routine prophylaxis: implications for improved protection.

Despite prophylaxis, pulmonary embolism (PE) remains a major cause of posttraumatic morbidity and mortality in high-risk patients. We studied injury-related risk factors associated with the occurrence of PE despite routine prophylaxis. A review of 9721 trauma patients discharged from January 1, 1985 through December 31, 1992, identified 36 patients (0.4%) who suffered clinically evident PE despite a policy of routine prophylaxis against deep venous thrombosis that included use of prophylactic inferior vena caval filters. (Twenty-nine patients had an inferior vena caval filter placed for prophylaxis against PE.) A detailed analysis of injury-related risk factors was performed. Four high-risk patterns of injury were identified, representing common combinations of significant risk factors. These patient groups have an absolute risk of PE despite prophylaxis ranging from 1.5% to 3.8%. The relative risk is approximately ten times that of control patients. Identification of appropriate high-risk groups is necessary to allow optimization of prophylactic measures, including placement of inferior vena caval filter.

Death in the operating room: an analysis of a multi-center experience.

To characterize causes of death in the operating room (OR) following major trauma, a retrospective review of admissions to eight academic trauma centers was conducted to define the etiology of death and challenges for improvement in outcome. Five hundred thirty seven OR deaths of 72,151 admissions were reviewed for mechanism of injury, physiologic findings, resuscitation, patterns of injury, surgical procedures, cause of death, and preventability. Blunt injuries accounted for 61% of all injuries, gunshot wounds (GSW) accounted for 74% of penetrating injuries. Sixty two percent of all patients arrived in shock. Average blood pressure (BP) was 52 mm Hg at the scene and 60 mm Hg on admission, with the period of shock > 10 minutes in 74%. Only 56% were resuscitated to a BP > 90 mm Hg before surgery. Average time to the OR was 30.1 minutes and mean best postresuscitation pH was 7.18. Mean best OR temperature was 32.2 degrees C. Recurrent injury patterns judged as the primary cause of patient death included head/neck injury (16.4%), chest injury (27.4%), and abdominal injury (53.4%). Actual cause of death was bleeding (82%), cerebral herniation (14.5%), and air emboli (2.2%). A different strategy for improved outcome was identified in 54 patients with the following conclusions: (1) delayed transfer to the OR remains a problem with significant BP deterioration during delay, particularly following interfacility transfer; (2) staged injury isolation and repair to allow better resuscitation and warming may lead to improved results; (3) combined thoraco-abdominal injuries, particularly with thoracic aortic disruption, often require a different sequence of management; (4) aggressive evaluation of retroperitoneal hematomas is essential; (5) OR management of severe liver injuries remains a technical challenge with better endpoints for packing needed; and (6) resuscitative thoracotomy applied to OR patients in extremis from exsanguination offers little.

Analysis of recurrent process errors leading to provider-related complications on an organized trauma service: directions for care improvement.

Complications in trauma care occur because of provider-related or patient disease-related events. Strictly defined standardized definitions of both types of complications are needed to develop strategies for problem resolution. The frequency and characteristics of 135 disease-related and provider-related complications were examined for a 3-year period in a level I university trauma service in all patients meeting Major Trauma Outcome Study (MTOS) criteria. Provider-related complications were analyzed for recurrent process errors to be targeted for corrective action. Complication events occurred in 2764 of 3327 patients, with provider-related complications in 759. Twenty-three percent (175) of complications were judged unjustified and 16 patterns of recurrent process-of-care errors were identified. Delay in trauma team activation was caused by insensitivity of field triage protocols and inadequate recognition of injury patterns. Delays in diagnosis or surgery were caused by inadequate performance of standard work-up, inadequate recognition of injury severity by providers, diagnostic procedure interpretation errors, and errors in prioritizing the order of diagnostic procedures. Errors in technique were attributed to inexperience, haste, unfamiliarity with devices, lack of developed institutional techniques, and failure of providers to use recognized endpoints. Errors in judgment were attributed to failure to access available patient information, proceeding despite available information, and failure to utilize available care guidelines. Further reduction in provider-related morbidity in an organized trauma system requires this type of analysis, which identifies the need to change the process of care through education or adjustment of protocols for standardization care delivery in addition to the traditional focus on outcomes.

A risk analysis of pulmonary complications following major trauma.

Varying institutional definitions and degrees of surveillance limit awareness of the true incidence of posttraumatic pulmonary complications. Prospective review with standardized definitions of 25 categories of pulmonary complications was applied to a university level I trauma service over 3 years to establish the true incidence. Potential injury-related predictors of individual complications were determined using multiple logistic regression analysis and adjusted odds ratios were calculated, thereby controlling for the effect of other covariants. Significance was attributed to p < 0.05. Of 3289 patients meeting MTOS criteria, pulmonary complications occurred in 368 (11.2%). Pulmonary complications account for one third of all disease complications. Significant associations with pneumonia included age, the presence of shock on admission, significant head injury, and surgery to the head and chest. Significant risk for atelectasis occurred in patients with blunt injury mechanism, ISS > 16, shock on admission, and severe head injury. Risks for development of respiratory failure included age > 55 years, the mechanism of "pedestrian struck", and the presence of significant head injury. Risk factors for ARDS included surgery to the head and a Trauma Score < 13 on arrival. Significant predictors for pulmonary embolism included ISS > 16, shock on admission, and extremity and pelvis injuries. The true incidence of pulmonary complications is established with this kind of analysis and focuses attention on (1) groups at high risk for developing complications, (2) groups for which current therapeutic modalities are still ineffective, and (3) defining the need to refocus on prospective research rather than ineffective processes of care.

Lack of effect of oxygen-radical scavenging systems in the preserved reperfused rabbit kidney.

Previous studies showed the beneficial effects of superoxide dismutase +/- catalase in perfusion-preserved rabbit kidneys but failed to show benefit in flush-cooled organs. The current studies undertook to evaluate scavengers, xanthine oxidase inhibitors, and agents that prevent the release of myeloperoxidase in 3 systems: kidneys preserved by perfusion or by flush cooling for 24 hr, studied immediately, and warm ischemia-injured kidneys evaluated after a 24-hr recovery period. In none of these groups could we demonstrate any protective effects against preservational or warm ischemic injury by the above modalities. Even though biochemical and other evidence from previous studies suggested free radical-induced injury to occur in preserved rabbit kidneys, these studies using renal function as the indicator did not do so.

Relaxation enhancement using liposomes carrying paramagnetic species.

Studies were performed to investigate the effects upon the relaxation times of mouse organs of intravenously administered Mn-DTPA entrapped in multilamellar liposomes, Mn-DTPA, 0.9% NaCl entrapped in liposomes, 0.9% NaCl. Manganese concentrations in injectates and tissues were assessed with 54Mn and atomic absorption. T1 and T2 of freshly excised tissues were measured in an NMR spectrometer at 20 MHz and 37% C with IR and CPMG sequences. Entrapment of Mn-DTPA in liposomes increased 54Mn accumulation in liver by 207% and in spleen 1208% and reduced 54Mn in heart by 20% and in kidney by 24% relative to free Mn-DTPA. Statistically significant increases in relaxation rates were produced. However, the increase in relaxation rate per unit concentration of Mn in tissue is reduced by 31% in liver and 62% in spleen when Mn is delivered inside liposomes. These observations have implications for the design of NMR contrast agents.