Annotated bibliography of the”>American Journal of Emergency Medicine (2006) 24, 517 – 539

The Literature of Emergency Medicine

Trauma: an annotated bibliography of the recent literature–2005

Charles J. McCabe MD^a,*, Ralph L. Warren MD^b

^aDepartment of Emergency Medicine, Massachusetts General Hospital, Boston, MA 02114, USA

^bDepartment of Surgery, Gallup Indian Medical Center, Gallup, NM 87301, USA

1. General

Demetriades D, Kimbrell B, Salim A, et al. Trauma deaths in the mature urban Trauma system: is ”trimodal” distribution a valid concept? J Am Coll Surg 2005;201:343-348.

The ”classic” trimodal distribution of trauma deaths was first described in 1982 by Dr Donald Trunkey and has been taught as one of the cornerstones of the advanced trauma life support curriculum eversince. The 3 ”modes” or peaks were (1) immediate at-the-scene deaths, which accounted for approximately half of the total and were primarily secondary to brain or major cardiovascular injuries; (2) the second peak occurred between 1 and 4 hours after injury, accounted for a third of the deaths, and were secondary principally to cardiovascular and Neurologic injuries; and (3) late deaths, 20% of the total, occurred a week out, a result of multiorgan failure. The authors of this article sought to determine whether this classic trimodal distribution of deaths remains valid a quarter of a century later. They reviewed all 4151 traumatic deaths in the county of Los Angeles over a 3-year period. Penetrating injuries, most often gunshot wounds, were responsible for half of the deaths, although they caused only 20% of the injuries. Motor vehicle injuries were the second most common cause of death (19.9%), followed by pedestrian vs auto injuries. The most commonly critically injured body area from blunt mechanisms were the heads and extremities, whereas for penetrating trauma, they were the chest and abdomen. The authors found that it was indeed true that 50% of the deaths occurred at the scene or within 1 hour of the injury–the first peak. The second time peak of deaths, between 1 and 6 hours after injury, accounted for 18% of the total. The new finding was that there was no third peak: the late deaths, occurring a week or more after injury and due to multiorgan failure, accounted for only 7.6% of the deaths: the temporal distribution of deaths after the first 6 hours has been evened out. The second important finding was that the timing of death depended strongly on which body area was critically injured. Critical chest injuries caused death mainly in the first hour and during the early period from 1 to 6 hours after injury, but there was no peak after that. Deaths from critical head injuries had basically a decaying exponential-shaped curve (although the authors’ graph made the distribution look flat because of the nonlinearity of their time (x-axis). The important implications of these findings are the following: (1) we are doing something (or more likely many things) right to decrease the incidence of multiorgan failure; (2) because half of all deaths occur immediately (within 1 hour) after injury, the only way to decrease that

4 Corresponding author.

number is by prevention; and (3) a focus on Novel methods of resuscitation and early aggressive intervention in the field and in the ED is warranted to affect the number of deaths in the first few hours.

Cinat ME, Wilson SE, Lush S, et al. Significant correlation of trauma epidemiology with the economic conditions of a community. Arch Surg 2004;139:1350-1355.

The purpose of this study was to evaluate the correlation between trauma epidemiology, unemployment, and violent crime in the authors’ community of Orange County, CA, with the hypothesis that depressed economic conditions would be associated with an increased incidence of trauma and violent crime. Crime statistics from 1993 to 2002 were obtained from the state Department of Justice and from the Federal Bureau of Investigation. Data regarding the annual population, labor force, and unemployment rates for 2 urban communities were obtained from county and US Census sources. The authors found that the unemployment rate was highly correlated with both the Federal Bureau of Investigation’s crime rates and with the percentage of penetrating trauma seen in the area hospitals. Moreover, three fourths of penetrating Trauma victims had no health care coverage or had state or county aid only.

Mahoney EJ, Harrington BT, Eiffi WL, et al. Lessons learned from a night- club fire: institutional disaster preparedness. J Trauma 2005;58:487-491.

Many advances in medicine and, especially, in surgery and emergency medicine (EM), have been a direct result of learning from war and disasters. This is an excellent chronicle of a level I trauma center’s experience after a nightclub fire in Rhode Island. As everyone in the United States probably remembers, the band’s pyrotechnics ignited the club’s furnishings, walls, and ceiling, at a time when the club was overcrowded with more than 400 patrons. The exits were blocked (a recurring theme in manmade disasters). The inferno caused 96 deaths at the scene (the ultimate death count was 100), and 215 patients were evaluated at area hospitals. The article describes the Rhode Island hospital’s management of this mass casualty. The trauma surgeon became the Commander in Chief of the ED, and all prepared for a large number of patients. The main ED was immediately cleared by transferring all critical/severe patients to the pediatric section of the ED, and the trauma intensive care unit (ICU) was cleared of patients and converted into a dedicated burn ICU by transferring all of the other patients to other ICUs. An entire 22-bed medical-surgical floor was also cleared of

0735-6757/$ – see front matter doi:10.1016/j.ajem.2006.05.001

patients by transferring its patients to 2 other medical/surgical floors that head been previously closed. All surgical house staff were recalled to the ED, and trauma teams consisting of a senior surgical house officer, nurse, and a respiratory therapist were created to immediately assess the arriving patients, establish an airway, and triage the patient after evaluation. Operating rooms (ORs) were cleared for management of the acutely burned patients who required burn debridement and grafting. Of the 64 patients who were evaluated at Rhode Island Hospital, 47 were admitted, 12 patients required bedside escharotomy within the first 12 hours, and 8 patients were transferred to another level I trauma/burn facility. Over the next several weeks, an additional 20 burned patients were transferred in from other hospitals for further care. Nine bedside tracheostomies and 184 bronchoscopies were performed in the ICU. Probably, the greatest stress in

trauma. In the past, radiology was primarily a diagnostic tool, whereas now, radiology has become more involved in therapy. With the better and better computed tomographic (CT) machines, CT provides better initial screening, so fewer diagnostic angiographic procedures are done. On the other hand, because the CTs are of better quality, more angiograms are positive, and more are needed to actively control hemorrhage via angioembolization or even emergency placement of endovascular stents. Moreover, with the evolution of minimally invasive techniques, CT- and ultrasound (US)- guided drainage of fluid collections is replacing open operative drainage. The importance of this evolution is that radiology departments must prepare to be more involved with round-the-clock ”24/7” service, requiring more personnel and sophisticated equipment.

this disaster was that put upon the nursing staff: burn care nursing is intense

and exhausting. The Rhode Island Hospital did ask for and did receive additional nursing staff from the Federal Emergency Management Agency.

This is a must-read article for everyone involved in hospital and ED disaster planning.

Roettger RH, Taylor SM, Youkey JR, et al. The general surgery model: a more appealing and sustainable alternative for the care of trauma patients. Am Surg 2005; 71:633-639.

Being a trauma surgeon has its good points and its bad points. It is exciting, requires one to be a true general surgeon, comfortable with all parts of human anatomy and physiology, and the results–bringing a patient back from near death to full health–are often extremely gratifying. On the other hand, it always happens at night (usually weekend nights), often involves a lot of nonoperative work which is both time-consuming and poorly compensated, and it can wreak havoc with one’s elective and office schedules. The Academic Department of Surgery in Greenville, SC, decided in 2002 to change their staffing system from that of the ”contemporary trauma/critical care model,” in which the trauma service staff are subspecialists who do only trauma and critical care, to a ”general surgery” model, in which the trauma staff have a general surgery practice and also take trauma call, whereas critical care is provided by a separate group of intensivists. They found that the new model not only produced superior financial results and a better quantitative surgical operative experience but also resulted in superior trauma and ICU overall outcomes. There are pros and cons of each system. In the former, surgeons lose their nontrauma general surgery skills (and income); in the latter, they lose their critical care skills. Which model is best for your hospital depends completely on the preferences of your surgeons–chacun a son gout.

Rogers FB, Madsen L, Shackford S, et al. A needs assessment for regionalization of trauma care in a rural state. Am Surg 2005;71:690-693.

Most trauma literature comes from urban trauma centers, most trauma training occurs at urban centers. However, there are real differences between trauma encountered in rural areas and that seen in urban areas. The Division of Trauma at the University of Vermont is one of the few academic leaders who are in a rural area. They reviewed all the trauma deaths occurring in their area during a 1-year period, with the aim of determining how they could improve their trauma system. They found that almost two thirds of trauma deaths occurred at the scene before any medical intervention and that suicide accounted for almost one third of deaths. Their main conclusion was that rural trauma system design should focus on improving discovery and earlier/ quicker access to the emergency medical system, as well as on suicide prevention training. Having moved from an urban trauma center to a rural one, I fully agree with Dr Rogers and colleagues.

Pryor JP, Braslow B, Reilly PM, et al. The evolving role of interventional radiology in trauma care. J Trauma 2005;59:102-104.

This interesting evaluation describes the evolution over the last few decades of the role of the radiology department in the care of patients with

Demetriades D, Gkiokas G, Velmahos GC, et al. Alcohol and illicit drugs in traumatic deaths: prevalence and association with type and severity of injuries. J Am Coll Surg 2004;199:687-692.

This is an interesting documentation of something most ED physicians are already quite familiar with: Intoxicated patients with trauma. The authors of this study reviewed all the trauma deaths at the Los Angeles County–University of Southern California Trauma Center over a 3 1/2- year period, in an attempt to ascertain just how big the effect of drug use is on trauma. There were 931 trauma deaths, but toxicology testing was done on only 600 of these. Of the 600, 43% of the victims tested positive for alcohol or illicit drugs, or both. Males were almost twice as likely to be intoxicated than females, and victims of penetrating trauma were intoxicated almost twice as often as victims of blunt injury. Victims older than 50 years were much less likely to be intoxicated, although there was still a 15% intoxication rate even for the elderly. More disturbing is that among the 15- to 20-year olds, 37% tested positive for illicit drugs and 26% positive for alcohol. Forty-two percent of the intoxicated victims had gunshot wounds, 15.3% were pedestrians struck by motor vehicles, 14% were falls, and 13% were occupants in Motor vehicle collisions. The severity of the injuries was similar for intoxicated and unintoxicated victims. Despite this fact, intoxicated victims were more likely to arrive at the ED with no vital signs. The authors note that in a previous study of all their patients with trauma, those who survived and those who died, 71% of the patients screened positive for alcohol, drugs, or both; cocaine and opiates accounted for 91% of the positive drug screens.

Blondell RD, Dodds HN, Looney SW, et al. Toxicology screening results: injury associations among hospitalized trauma patients. J Trauma 2005;58:561-570.

Extending the results of the above article are the findings described in this report. It has become accepted and demonstrated in most EDs that alcohol abuse is commonly associated with traumatic injury, whereas the relationship between trauma and other drug use is less clear. The purpose of the present study was to determine the alcohol- and drug-specific associations of toxicology screening results with injury characteristics and outcome. The authors determined that cocaine was associated positively with violent injuries and negatively with nonviolent injuries. The type of injury associated with alcohol also depended upon what other drugs had been abused.

Alcohol and cocaine were additive in association with violence-related injury. On the other hand, opiate use was commonly associated with nonviolent injuries and burns.

Weinsheimer RL, Shermer CR, Malcoe LH, et al. Severe intimate partner violence and alcohol use among female trauma patients. J Trauma 2005;58:22-29.

intimate partner violence (formerly known as spouse abuse) and alcohol abuse are 2 major contributors to injury and death in the United States. Estimates of the incidence of IPV over a woman’s lifetime in the

United States range from 18% to 50%; each year; 1% to 13% of US women experience physical assault from an intimate partner. Alcohol is the drug most commonly associated with IPV, both by the aggressor and the victim. This study examined the relationship of female alcohol use, male partner alcohol use, and male-female IPV among female patients with trauma. The authors found that IPV among women seen in trauma centers is greater than that found in national surveys, that alcohol problems among abused women and their partners are greater than those among nonabused women, and that female problem-drinking increases a woman’s risk of IPV.

Problem-drinking women need to be offered interventions while hos- pitalized so that they and their partners can be referred to alcohol therapy.

Hundley JC, Kilgo PD, Miller PR, et al. Non-helmeted motorcyclists: a burden to society? A study using the National Trauma Data Bank. J Trauma 2004;57:944-949.

The Highway Safety Act of 1966 gave the Department of Transpor- tation authority to withhold federal highway safety and construction funds to states if they did not comply with mandatory motorcycle helmet use. This law has generated tremendous controversy, and it has been rescinded in many states. Opponents of mandatory helmet laws claim that the helmets reduce peripheral vision and hearing and increased the incidence of neck injuries, and that differences in morbidity and mortality can be attributed to errors in statistical modeling that fail to control for the effects of alcohol and drugs. The present article used the National Trauma Data Bank to determine the impact of helmet use on injuries, outcome, use of Hospital resources, and discharge status. It is a retrospective study covering an 8-year period. The authors found that mortality was significantly higher for nonhelmeted motorcyclists when compared with helmeted motorcy- clists, and that the injuries in nonhelmeted motorcyclists were more severe, as reflected in the higher injury severity score (ISS), the lower Glasgow Coma Scale score, and worse, Revised Trauma Score. Nonhelmeted motorcyclists had significantly longer ICU and hospital stays than helmeted motorcyclists and were more than likely to have either no health insurance or government-funded insurance. Their discharge was often to rehabilitation hospitals or nursing homes. Critics of legislated helmet use claim that these laws restrict personal freedom and individual rights; this article shows that the significantly higher charges for care of nonhelmeted motorcyclists is borne by the general public. The authors suggest that perhaps we should force the critics of the helmet law to accept a law mandating that motorcyclists carry health insurance whether they want to wear a helmet or not.

Smith LM, Pittman MA, Marr AB, et al. Unsafe at any age: a retrospective review of all terrain vehicles injuries in two level I trauma centers from 1995-2003. J Trauma 2005;58:783-788.

All-terrain vehicles are very popular recreational devices. Approximately 7000000 ATVs are in use in the United States. They weigh up to 600 lb and can attain speeds of up to 75 miles per hour. Stability of ATVs on various terrains is quite variable. Not generally recognized is the number of deaths and injuries that occur with the use of these devices. There were more than a million ED visits and approximately 500 deaths from ATV use in the year 2001. Yet, there are few and inconsistent safety regulations regarding ATV use in the United States. Twenty-one states have helmet and safety equipment regulations, 23 states have machine-related safety requirements, and 3 states and the District of Columbia have no regulatory statues. This study is a retrospective review from 2 level I trauma centers, 1 in Texas and 1 in Louisiana, of all ATV-related trauma. The most common injuries were to the face, including fractures, lacerations, contusions, and abrasions. More than a quarter of the patients had major head injury, whereas minor head injuries occurred in an additional 14%. thoracic injuries were the next most common, occurring in 24% of patients. Twenty-three percent had extremity injuries including fractures, lacerations,

contusions, abrasions, and degloving injuries. Abdominal injuries occurred in 12% of patients. There were 12 deaths in 208 patients, by far, most caused by head injuries. Three patients died of rupture of the thoracic aorta, and 1 had severe liver injuries. Driving an ATV requires muscle strength, psychomotor skills, and good judgment. The authors believe that both training on the use of an ATV and helmet use should be required. They also recommend the following regulations: a minimum age of 12 years to drive an ATV; adult supervision of riders between the ages of 12 and 16 years; minimum age of 16 years for ATVs with engines more than 90 cc; mandatory safety course and mandatory wear of helmet, goggles, and padded clothing for all riders; licensing at a state level for ATV use; and no passengers, no riding on paved roads, and no use of alcohol or drugs before ATV use. Of course, if we here in the United States cannot even get motorcycle helmet laws passed, the chances of getting these recommen- dations approved are slim at best.

Fonseca AH, Ochsner MG, Bromberg WJ, et al. All-terrain vehicle injuries: are they dangerous? A 6-year experience at a level I trauma center after legislative regulations expired. Am Surg 2005;71:937-941.

Here is a completely separate article from the one above, but highly related. All-terrain vehicles hit the American scene in the early 1970s and rapidly grew in popularity. Because of a near epidemic in injuries stemming from ATV use, rider safety regulations were published in 1988. These were voluntary regulations, however, and they expired without being renewed in 1998. The authors of this article reviewed their trauma registry from January 1998 to August 2004, comparing ATV-associated injured patients to motorcycle rider trauma victims. Compared with 352 injured motorcycle riders, the 221 injured ATV riders were more often pediatric or female and used a helmet much less frequently (9% vs 65%) and, thus, of course, had a higher incidence of closed head injury. Overall ISS and mortality were similar in the 2 groups. Review of previous records revealed that there were about 7 patients per year injured in ATV crashes during the 10 years the regulations were in place vs 32 per year after the regulations lapsed. All-terrain vehicles are dangerous. If bicycle riders have to wear helmets, surely, ATV riders need to as well: ”there should be a law ”

Demetriades D, Murray J, Brown C, et al. High-level falls: type and severity of injuries and survival outcome according to age. J Trauma 2005;58:342-345.

The purpose of this study was to evaluate the affect of age on the incidence and severity of specific injuries and death after falls. The authors evaluated 10 years’ worth of experience at their level I trauma center, specifically investigating all victims of falls greater than 15 ft. The patients were divided into 4 age groups: less than 14 years, 15 to 55 years, 56 to 65 years, and older than 65 years. One floor in a building was considered 10 ft; many of the heights were estimates. The most common body area with severe trauma was the head, followed by the chest and the extremities. The most common intracranial injuries were cerebral contusion and subarach- noid hemorrhage, followed by subdural hematomas and epidural hemato- mas. Spinal injuries were diagnosed in 24%, with the lumbar spine, by far, the most common injured site. Solid Organ injuries occurred in 3.7%, most commonly the liver, followed by the kidney and spleen. pelvic fractures occurred with increased incidence with age and were distinctly uncommon in patients younger than 15 years. The severity of injuries and the mortality increased with age.

The height, Body position on impact, and impact surface are all important factors in determining the injury to falls from any height. Spinal injuries are extremely common especially in patients older than 50 years, the lumbar spine being, by far, the most commonly injured area followed by thoracic and then the cervical spine. In the authors’ experience, aortic injuries were uncommon. As demonstrated in many studies, age plays a significant role in the severity of the trauma and mortality associated with it.

Lapostolle F, Gere C, Borron SW, et al. Prognostic factors in victims of falls from height. Crit Care Med 2005;33:1239-1242 [France].

When the paramedics bring a patient into the ED after a fall from height, what information about the fall is helpful? These authors from France reviewed the charts of 287 adult patients seen over the previous 33 months who had fallen greater than 3 m (roughly equivalent to 1 floor or 10 ft) and tried to correlate details of the fall with outcomes. The overall mortality was 34%. Multivariate analysis revealed the following indepen- dent variables to affect mortality: older age, height of fall, hard vs soft landing surface, and head impact first vs front landing vs lateral landing (in decreasing order of danger). Two interesting tidbits for use on rounds, one’s velocity at the end of a fall from height is roughly (2gh)^1/2, where g is the acceleration due to gravity, 32 ft/s², and a parachute can attenuate deceleration forces by a factor of 36 or more.

Kluger Y, Peleg K, Daniel-Aharonson L, et al. The special injury pattern in terrorist bombings. J Am Coll Surg 204;6:875-889.

Bombings have become the weapon of choice of terrorist groups, and no country has more experience in dealing with terrorist bombings than Israel. The authors of this article, the Israeli Trauma Group, have noticed some important characteristics of patients hospitalized after bombings and the distribution of their injuries. Nine hundred six victims of terrorist bombings, seen over a 3-year period, were compared with 55033 individuals injured by nonterrorist trauma, all from the Israeli National trauma registry. Bomb explosions resulted in significantly higher injury severity and complexity. Only 10% of conventional trauma victims had ISS of 16 or greater, whereas in terrorist bombings, the ISS was 16 or greater in 30%. Patients had a low GCS score 4 times more often, and they were twice as likely to be hypotensive and unstable upon hospital arrival. Bombing victims had more body regions injured than victims of conventional trauma and required surgical interven- tion much more often. Finally, both immediate and late mortality were higher among the bombing casualties. The diagnosis and management of these patients must differ from those of patients with conventional, severe trauma. The severity and lethality of their injuries requires appropriate triage to facilities that are prepared to provide immediate care.

Schecter WP, Fry DE. The Governor’s Committee on Blood Borne Infection and Environmental Risk of the American College of Surgeons. The Surgeon and acts of civilian terrorism: Chemical agents. J Am Coll Surg 2005;200:128-135.

Weapons of mass destruction include bombs, both conventional and nuclear; chemical poisons; and biologic toxins. terrorist attacks anywhere could combine any of these in any combination. The medical responders to terrorist attacks, both EM physicians and surgeons, must therefore be prepared to deal with patients having had any combination of trauma and exposure to chemical and biologic agents. This article focuses on chemical weapons; the next article deals with biologic toxins. Neither article can really be summarized or distilled; both must be read, but we can give you a flavor of them here.

There are 4 categories of chemical weapons available for use in a civilian/terrorist event: nerve agents, cyanide, vesicants, and pulmonary agents. A cardinal principle in dealing with victims of chemical or biologic agent exposure is decontamination, both to prevent further absorption of the agent by the patient and to prevent exposure of the care providers. Nerve agents are organophosphates, which are cholinesterase inhibitors and thus cause Cholinergic crisis. Many commercial insecticides are organophos- phates, and Insecticide poisoning causes the same syndrome as do nerve agents. Military nerve agents include tabun, sarin, soman, GF and VX, but terrorists could use ordinary insecticides. Cholinergic crisis–excess acetylcholine causing excess stimulation of both muscarinic and nicotinic receptors–is manifest as excessive sweating, excessive salivation, and production of bronchial mucus, constricted pupils, abdominal cramps,

muscle fasciculation and weakness leading to paralysis, and delirium progressing to seizures and coma. The antidote of choice is atropine, which acts as a competitive inhibitor of acetylcholine at its receptors. Treatment of mild to moderate exposure begins with the administration of atropine 2 mg IM, followed immediately by 600 mg of 2-pralidoxime chloride, which acts to prevent stabilization of the nerve agent cholinesterase complex. Diazepam is used for seizures. Mechanical ventilation may be required for severe exposures. Cyanide interrupts mitochondrial oxidative metabolism. Cyanide has an odor of bitter almonds, but only half of normal people can actually smell it. Victims are classically described as having a cherry-red cutaneous flush, but this finding is not universal. Cyanide causes disruption of the mitochondrial cytochrome energy production system and, thus, results in severe metabolic acidosis. Patients with Cyanide poisoning should be treated with amyl nitrite or sodium nitrite intravenously. Cyanide poisoning kits are somewhat complex, and should be found in every ED. Vesicants, such as Mustard gas, are liquids that cause cutaneous and mucous membrane burns and bone marrow depression. The priority in caring for patients exposed to vesicants is to remove any residual agent as quickly as possible while taking great care to protect caregivers. Pulmonary agents (chlorine, phosgene) are highly lethal chemicals that cause hypoxic death by direct injury to the pulmonary parenchyma. The treatment for patients with inhalation injury often requires intubation and ventilation.

Fry DE, Schecter WP, Parker JS, et al. The Surgeon and acts of civilian terrorism: biologic agents. J Am Coll Surg 2005;200:291-302.

The use of biologic agents by terrorist groups is perhaps the greatest cause of concern at the present time. This article focuses on the biologic agents that are available for acts of civilian terrorism. Biologic agents can be bacteria, viruses, or toxins. They require certain characteristics to be useful, including ease of production, stability in storage to allow adequate inventories to accumulate, appropriate Particle size so as to be easily disseminated, and finally, great efficacy in producing either lethal or incapacitating effects in the infected host. Agents most fitting these criteria include those causing anthrax (Bacillus anthracis), plague (Yersinia pestis), and tularemia (Francisella tularensis), as well as smallpox virus and botulinum toxin. Less likely agents include those causing cholera, brucellosis, glanders and Q fever. This is an excellent, well-organized article, and it will serve as both an excellent introduction and a good reference.

Rhee P, Nunley MK, Demetriades D, et al. Tetanus and trauma: a review and recommendations. J Trauma 2005;58:1082-1088.

Continuing with articles on infectious/noxious agent that can complicate acute trauma, this is an excellent review of tetanus. Tetanus is a real problem in countries where immunization is not practiced. It is estimated that 200000 to 1000000 deaths from tetanus occur worldwide each year! Some of these deaths are manifested as neonatal tetanus due to the high incidence associated with the use of nonsterile instruments or poultices on the umbilical cord. Tetanus is caused by 2 toxins secreted by the Clostridium tetani, a gram- positive, spore-forming obligate anaerobic bacillus. Wounds contaminated by soil and with low oxygen tension are optimal locations for the germination of Clostridium tetani. The main feature of clinical tetanus is painful muscular contractions that are sometimes violent enough to cause fractures. The name tetanus itself is derived from the Greek word for ”stretching” or ”rigidity.” Tetanus may cause spasm of the masseter muscles or lockjaw; risus sardonicus, spasms of the facial muscles which produces what appears to be a sardonic smile; or opisthotonus, rigid hyperextension of the spine. Muscles spasms are managed with Muscle relaxants. The disease may progress to airway obstruction and/or a reduction in chest wall compliance causing respiratory failure. Tetanus-prone wounds are considered to be Crush injuries, Open fractures, punctures, abscesses, and wounds with devitalized tissue or contaminated with dirt or rust. But even minor wounds can harbor clostridia and give rise to tetanus, which is one of the main reasons why tetanus immunization is so important. Everyone knows that tetanus toxoid needs to be readministered every 10 years, and in those with wounds, it

should be given if the last booster was more than 5 years previously. Some experts are now recommending that if the last immunization was more than 10 years before, then tetanus immune globulin should be administered, regardless of the severity of the wound. I personally feel this last recommendation would be an overkill, in light of the extremely low incidence of tetanus in the United States over the past several decades while following the old guideline that tetanus immune globulin be given only to those with especially tetanus-prone wounds.

Stawicki SP, Grossman MD, Cipolla J, et al. Deep venous thrombosis and pulmonary embolism in trauma patients: an overstatement of the problem? Am Surg 2005;71:387-391.

That patients with trauma are at high risk of developing Venous thromboembolic disease (VTE) has been known for decades. But just how high a risk is debatable: many studies report quite high numbers that do not jibe with many clinicians’ personal experience. The authors of this article performed a 10-year retrospective analysis of the Pennsylvania State Trauma Outcome Study, finding more than 73000 patients with trauma who were considered at high risk of developing VTE (pelvic or lower extremity fracture, severe head injury, or spinal cord injury). The incidence of VTE among these patients was just below 2% for Deep venous thrombosis , and 0.5% for pulmonary embolism. Patients with isolated lower extremity fracture had a 1.3% incidence of VTE; patients who had combined severe head injury and pelvic and lower extremity fractures had the highest incidence at 5.4%. In logistic regression analysis, only ISS was consistently predictive of VTE. These figures are much lower than those generally reported in other studies for reasons which are not entirely clear. My interpretation is that the present study represents data collected (1) retrospectively, (2) from many different hospitals, and (3) using many different techniques for diagnosis. I infer that the incidence of VTE is higher than reported here if you look for it carefully. On the other hand, these data may be more representative of the proportion of DVT which are clinically significant. Most calf DVTs, even if left untreated, do not progress to proximal DVT or pulmonary embolism.

Velmahos GC, Petrone P, Chan LS, et al. Electrostimulation for the prevention of deep venous thrombosis in patients with major trauma: a prospective randomized study. Surgery 2005;137:493-498.

Current prophylactic measures against VTE, including low-dose or low- molecular-weight heparin, lower extremity sequential compression devices, and inferior Vena cava filters, are far from perfect. This article reports a preliminary study of 47 patients with trauma, half of whom were randomized to have muscle electrostimulation of the lower extremities for 30 minutes twice a day for 1 or 2 weeks. All 47 had heparin and/or sequential compression boots as well. The muscle electrostimulation was well tolerated, but unfortunately, it was ineffective, as evaluated by venography or duplex US. A good effort, but back to the drawing board.

DeMaio A, Torres MB, Reeves RH. Genetic determinants influencing the response to injury, inflammation, and sepsis. Shock 2005;23:11-17.

One of the most important factors affecting one’s health is how you pick your parents. Not just Hemophilia A or familial polyposis but all the ills that human flesh is heir to–even infectious diseases–are strongly influenced by one’s genetic makeup. The same is true for one’s response

Ciesla DJ, Moore EE, Johnson JL, et al. The role of the lung in postinjury multiple organ failure. Surgery 2005;138:749-768.

Central to our current understanding of the pathophysiology of systemic inflammatory response syndrome and multi-organ dysfunction syndrome is the release from the gut of Inflammatory mediators, which then go directly via the venous circulation to the pulmonary vasculature. Inflammation in the lung, in turn, leads to release of more mediators, which travel through the pulmonary venous return to the arterial side of the circulation then to the various body organs. The trauma/critical care group at the University of Colorado Health Sciences Center prospectively studied 1344 adult patients with trauma with ISS of higher than 15 and found that 75% of them developed at least 1 organ dysfunction. Lung dysfunction occurred in 94% of patients with 1 or more organ dysfunctions and in 99% of patients with 2 or more organ dysfunctions. Lung dysfunction always preceded other organ dysfunctions, and the severity of the lung dysfunction correlated with both the number and the severity of other organ dysfunctions. These findings strongly support the idea that the injured lung is indeed the source of noxious substances that wreak havoc on rest of the body’s organ systems.

1. Prehospital

Hannan EL, Farrell SL, Cooper A, et al. Physiologic triage criteria in adult trauma patients: are they effective in saving lives by transporting patients to trauma centers. J Am Coll Surg 2005;200:584-592.

This study examined the performance of various physiologic trauma triage criteria by comparing mortality rates for patients who are treated in specialized trauma centers vs those treated in nonspecialized facilities. Mortality rates of patients flagged by the criteria for triage to a trauma center were compared among patients who actually did get taken to a trauma center vs those who were taken instead to other hospitals.

The database used in the study was the New York State Trauma Registry for the years 1996 through 1998. Patients in this study included all adult patients in the registry outside New York City who were identified in prehospital care reports as having met 1 or more of the following criteria for triage to a trauma center: a GCS score of less than 14, systolic Blood pressure less than 90 mm Hg, respiratory rate less than 10 or greater than 30 breaths per minute, or pulse less than 50 or greater than 120 beats per minute. The authors found that mortality rates for patients meeting the triage criteria were indeed lower at regional trauma centers, compared with area centers and nontrauma centers, confirming the validity of these triage criteria.

Codner P, Obaid A, Porral D, et al. Is field hypotension a reliable indicator of significant injury in trauma patients who are normotensive on arrival to the emergency department? Am Surg 2005;71:768-771.

A retrospective analysis of all patients seen in the University of California, Irvine (UC-Irvine) trauma unit over a period of a year yielded a pool of 47 trauma victims who had been hypotensive (systolic BP b90 mm Hg) in the field but who were normotensive by the time they arrived at the ED. Thirty-seven were victims of blunt trauma, and 10 had penetrating trauma; the mean ISS was 16, and 9 had ISS of 25 or higher (”critical”). More than half required surgery, and 90% required ICU admis- sion. Overall mortality was 10%. All 5 who died had significant head and neck injuries. The authors concluded that hypotension in the field, even if it spontaneously resolves by the time the patient gets to the hospital, is indeed a marker for potentially serious injury and should prompt thorough evaluation.

to severe injury. Recent studies have shown that the inflammatory response

is affected by genetics: different people react with different degrees of activation of inflammatory cascades, including production of inflammatory mediators, to the same insult, be it necrotizing pancreatitis, bacteremia, or mechanical trauma. This good review from Johns Hopkins describes the current state of our understanding of genetic influences on injury response, and how progress is being made by trying to untangle genetic factors with clinical studies, as well as by using various animal strains inbred with certain polymorphisms at specific genetic loci.

Almogy G, Luria T, Richter E, et al. Can external signs of trauma guide management? Arch Surg 2005;140:390-393.

The title is a bit imprecise: the purpose of this article was to help on-scene medical crews at the sites of mass-casualty terrorist bombings to recognize blast injury and, more effectively, triage salvageable vs nonsalvageable victims. Explosive blasts cause injury by 4 mechanisms. Primary blast injury is caused by rapid outward spread of the overpressure shock wave, which

injures gas-containing organs including the middle ear, lungs, and intestines; the energy of the blast overpressure wave and, thus, the magnitude of resulting injuries, is inversely proportional to the square of the distance from the explosion’s epicenter. Secondary blast injury is caused by penetrating missiles propelled outward by the explosion, composed of bomb-based material (shrapnel and debris). Tertiary blast injuries result from a victim’s body being thrown violently back by the expanding gases and high winds; injuries are incurred from tumbling and impacting other objects. Burns and injuries secondary to collapse of buildings are defined as quaternary injuries and are common among survivors of explosions. The extent of bomb injuries also depends on whether the explosion was in a confined space, such as in a bus or a building, vs in an open space: blast lung injury is much more commonly seen in closed-space explosions. Signs and symptoms of blast lung injury may be subtle and are often underestimated, but patients may deteriorate quickly and often require sophisticated modes of mechanical ventilation. The principal finding of the present article was that evaluating the degree of external trauma can anticipate the risk of blast lung injury. The authors found that victims who suffer from Skull fractures, burns covering more than 10% of the body surface area, and/or penetrating injuries to the head are more likely to suffer from blast lung injury: it is likely that these victims were very close to the explosion. The distance of the victim from the explosion determines the amount of energy and penetrating missiles that are absorbed by the victim and are the most important factors influencing survival. Full recognition of these identifiable injuries can help emergency crews at the scene in triaging patients to the appropriate receiving facility.

Salamone JP, Ustin JS, McSwain ME, et al. Opinions of trauma practitioners regarding Prehospital interventions for critically injured patients. J Trauma 2005;58:509-517.

Just how much should be done by field rescue teams at the scene of a trauma remains controversial: is ”scoop and run” the correct approach, or should responders proceed as required with intravenous (IV) placement, endotracheal intubation, cricothyrotomy? This article contains the results of a survey of US trauma surgeons concerning the appropriateness of selected basic and advanced life support therapies for the management of gravely injured patients in the prehospital setting. Of the 858 members of the American Association for the Surgery of Trauma, 345 were randomly selected and mailed a questionnaire that presented case scenarios of critically injured patients, followed by questions concerning elements of prehospital care, particularly airway management, IV volume resuscitation, the use of the pneumatic antishock garment, spinal immobilization, and pulse oximetry. Of the 345, 182 questionnaires were returned. Results included the following: all of the respondents believed that monitoring pulse oximetry should be mandatory, and half felt that capnometry is essential as well. About half of the surgeons supported the use by paramedics of chemical paralysis for intubation; 30% supported surgical cricothyrotomy by paramedics for patients with severe facial trauma who required intubation. intravenous fluid administration to hypotensive patients with gunshot wounds of the abdomen was controversial: some respondents felt that IV fluid would be appropriate but should be limited in amount; an urban setting with a short transport time made all respondents more interested in rapid transport than IV fluid administration. All agreed that either normal saline or Ringer’s lactate could be used. For patients with pelvic fracture and hypotension, most of the respondents felt the use of pneumatic antishock garment was indicated if the transport times were greater than 20 to 40 minutes. Endotracheal intubation was highly recommended for patients with traumatic brain injury, particularly if they were combative.

American College of Emergency Physicians, American College of Surgeons Committee on Trauma, and The National Association of EMS Physicians, Drug-assisted intubation in the prehospital setting. J Am Coll Surg 2005; 201:585.

All emergency medical service (EMS) directors need to read this article. The establishment of an Adequate airway is a top priority in any and

all resuscitations. Controversy still reigns concerning whether placement of an endotracheal tube in the prehospital setting should be attempted. Endotracheal intubation can be both difficult and dangerous even in experienced hands. This reference, the first of several articles published this year concerning prehospital intubation, represents a joint position statement from the American College of Emergency Physicians, the American College of Surgeons Committee on Trauma, and the National Association of EMS Physicians. The recommendation is that drug-assisted endotracheal intubation should be performed only under the following conditions: (1) medical direction with concurrent and retrospective oversight supervision;

(2) proper patient selection, to include training and continuing education designed to demonstrate initial and ongoing competence in the procedure;

(3) training in airway management of patients who cannot be intubated and the availability and competence in the use of backup rescue airway methods; (4) standardized drug-assisted intubation protocols; (6) resources for drug storage and delivery; (7) resources for continued monitoring and recording of vital signs; oxygen saturation and end-tidal carbon dioxide;

(8) appropriate training and equipment to confirm initial and verify ongoing tube placement; (9) continued quality assurance, performance review, and when necessary supplemental training; and (10) research to clarify the role of drug-assisted intubation on improved patient outcomes within EMS systems.

Wang HE, Kupas DF, Hostler D, et al. Procedural experience with out- of-hospital endotracheal intubation. Crit Care Med 2005;33:1718-1721. Gerbeaux P. Should emergency medical service rescuers be trained to prac- tice endotracheal intubation [Editorial]? Care Med 2005;33:1864-1865.

This article follows nicely the joint position article above. Out-of- hospital emergency endotracheal intubation can be a difficult procedure, and errors and adverse events during performance of intubation can have dire consequences. It is a truism that success rates in performing difficult procedures correlate with experience in performing the procedures. Emergency Medicine specialists from Pennsylvania queried their state’s database of all the more than 1.5 million EMS Patient care reports in Pennsylvania for the year 2003 and examined the number of prehospital endotracheal intubations performed. Almost 11 500 intubations were performed by 5245 providers. The median number of intubations performed by individual providers was one, and more than two thirds of rescuers performed 2 or fewer intubations per year. The number of intubations was higher, of course, for those providers working in high patient volume areas, for urban rescuers, and for air medical transport teams, but even these higher volume rescuers performed only 2 to 3 intubations per year. These findings highlight a challenge to the EMS community: how to optimize prehospital airway management. Options, none of which are perfect, include increasing refresher training, allowing only certain providers to be specialists in intubation and using alternative airways such as the laryngeal mask airway or Combitube. The accompanying editorial by Dr Gerbeaux expands nicely on all these alternatives.

Bulger EM, Compass MA, Sabath DR, et al. The Use of Neuromuscular blocking agents to facilitate prehospital intubation does not impair outcome after traumatic brain injury. J Trauma 2005;58:718-724.

The authors of this study attempted to determine the impact of the use of Neuromuscular blocking agents (NMBAs) on outcome of patients with traumatic brain injury requiring prehospital intubation. They studied 2012 patients who had had head injury and who had a recorded prehospital GCS score. There were 920 patients with a GCS score of 14 or 15, 293 patients with a GCS score of 9 to 13, and 799 patients with a GCS score of less than

9. Prehospital intubations were done in 17.4%, 57.7%, and 94.9%, respectively, of the above groups. Overall, 54% of the patients underwent prehospital intubation, with three fourths of these using NMBA. Patients not receiving NMBA were more likely to be hypotensive or to have ongoing cardiopulmonary resuscitation (CPR). For the entire group requiring intubation, the mortality was 25% in the NMBA group vs 37%

in the non-NMBA group. Among patients with a GCS score of 8 or lower, the mortality was 32% in the NMBA group vs 46% in the non-NMBA group. The authors concluded that the use of paralyzing agents in the prehospital environment to facilitate endotracheal intubation did not worsen the outcome.

Bozeman WP, Idris AH. Intracranial pressure changes during rapid sequence intubation: a swine model. J Trauma 2005;58:278-283.

This is an experimental study on pigs to determine the effects of laryngoscopy and intubation on BP, heart rate, and intracranial pressure . The use of rapid sequence intervention with barbiturates, etomidate, and/or propofol is known to decrease ICP, but controversy surrounds the use of succinylcholine, a depolarizing agent which is known to increase ICP. Some experts feel that patients with head trauma would benefit from pretreatment with both IV lidocaine as well as a nondepolarizing Neuromuscular blocker before rapid sequence intubation. The authors performed this study as a pilot to develop a model so that the effects of intubation on ICP could be studied. Four thiopental-based induction regimens reflecting common Intubation practices were compared: (1) sedation-only intubation, (2) standard rapid sequence with sedation and paralysis, (3) lidocaine pretreatment before rapid sequence intubation, and

(4) both lidocaine and pancuronium pretreatment before rapid sequence induction. All regimens showed an increase in ICP during intubation. Perhaps, surprisingly, the sedation-only regimen showed the smallest mean peak in ICP. All 3 of the regimens using succinylcholine showed approximately 3-fold higher increases in mean peak ICP during rapid sequence induction.

Davis DP, Peay J, Sise MJ, et al. The impact of prehospital endotracheal intubation on outcome in moderate to severe traumatic brain injury. J Trauma 2005;58:933-939.

For patients with head injury, prevention of hypoxia and aspiration are primary goals. But as noted above, the beneficial effects of invasive airway management on eventual outcome remain unproven. In this study, the authors used a large database to explore this relationship. Logistic regression was used to adjust for multiple variables that impact traumatic brain injury outcome, and neural network analysis was performed to identify patients who might benefit from paramedic endotracheal intuba- tion. A total of 13625 patients from 5 trauma centers with moderate to severe traumatic brain injury were identified from the county trauma registry and included in the analysis. The authors found an increase in mortality with prehospital intubation, but despite their conclusion that prehospital intubation was associated with a decrease in survival among patients with severe traumatic brain injury, the authors recommend that readers should not conclude that endotracheal intubation is detrimental to patients with traumatic brain injury. I am still uncertain of the validity of the authors’ statistical methods, particularly controlling for the obvious association between a worse head injury to start with and the consequent increased likelihood or requiring an emergent airway. Even the authors say their article is inconclusive.

DiRusso SM, Sullivan T, Risucci D, et al. Intubation of Pediatric trauma patients in the field: predictor of negative outcome despite risk stratifica- tion. J Trauma 2005;59:84-91.

The authors of this study were more sure of themselves and more vehement with their recommendation against the use of prehospital intubation: ”Field intubation is a strong negative predictor of survival or good functional outcome despite adjustment for severity of injury…except for specific circumstances where bag-valve-mask ventilation is ineffective, field intubation in pediatric trauma patients should be abandoned.” The authors queried the National Pediatric trauma registry, gathering data on a total of 50199 pediatric patients with head trauma. The overall morality was 2.9%. Five thousand four hundred sixty patients required intubation.

Unadjusted mortality was lowest in those patients intubated at a trauma center (13.2%), higher in those intubated at nontrauma center hospitals (16.7%), and highest in patients intubated in the field (38.5%). Importantly, higher mortality rates were seen not only for the more severely injured or those with severe head injury but also for those with moderate or mild injuries. From an outcome perspective, patients intubated in the field were more likely to go to a more extended care facility than those intubated in the hospital. The authors found that transport times were almost 30 minutes longer in field intubated patients.

1. Resuscitation

Pickens JJ, Copass MK, Bulger EM. Trauma patients receiving CPR: predictors of survival. J Trauma 2005;58:951-958.

The authors of this article wanted to validate the guidelines on the ED triage of trauma victims in cardiopulmonary arrest that were published in a consensus statement in 2003. That consensus statement, representing both the National Association of the Emergency Medical Services Physicians and the American College of Surgeons Committee on Trauma, was entitled Guidelines for Withholding or Termination of Resuscitation in Prehospital Traumatic Cardiopulmonary Arrest and appeared in the Journal of the American College of Surgeons, 2003;196:475-481. The consensus statement was the result of a review of the existing literature and presented guidelines for withholding or terminating CPR for patients in traumatic arrest, which included patients with blunt trauma who were apneic, pulseless, and without an organized rhythm on electrocardiogram (ECG) on presentation to the ED; victims of penetrating trauma found apneic and pulseless in the field; and all patients with trauma with CPR ongoing for more than 15 minutes before hospital arrival. The authors of the present article attempted to examine the relationship between prehospital clinical assessment and survival following cardiac arrest in their patient population. The central issue, which the authors evaluated, was whether the survival of patients with trauma can be accurately predicted by EMS personnel using these clinical criteria. This study was performed in a retrospective fashion reviewing all cardiopulmonary arrest patients who were transported to the authors’ level I trauma center. One of the key findings was that pulselessness was subjective: there were several discrepancies between the initial assessment by basic life support providers and that of advanced life support personnel. Advanced life support personnel were able to restore a pulse in many of the patients who were considered by the First responders to be pulseless. The authors concluded that either these patients responded to treatment or that they had not been appropriately assessed in the field. Because of difficulties with prehospital assessments, the authors suggest that the ED, rather than the prehospital setting, is the more appropriate setting for patient triage. In particular, the authors did not support the 15-minute resuscitation time limit: 3 of the authors’ survivors had CPR times greater than 15 minutes.

Daniel SR, Morita SY, Yu M, et al. Uncompensated metabolic acidosis: an unrecognized risk factor for subsequent intubation requirement. J Trauma 2004;57:993-997.

Patients with metabolic acidosis because of trauma and hypovolemia are often able to at least partially compensate for the acidosis using their respiratory system. The authors hypothesized that patients with trauma with elevated Paco₂ levels relative to their degree of acidosis have a higher rate of respiratory failure and a need for intubation. In a retrospective study, patients admitted over a 1-year period were analyzed with specific interest in the arterial blood gases. Of 140 patients with metabolic acidosis, 45 ultimately required intubation. The authors concluded that the inability to mount a respiratory response to metabolic acidosis induced by traumatic shock is an excellent predictor of the need for future intubation. In retrospect, this conclusion is obvious. The value of the authors’ article is to remind us to think ”? ET tube” whenever we see an uncompensated acidosis, if we have not already done so.

Sung J, Bochicchio GV, Joshi M, et al. Admission hyperglycemia is predictive of outcome in critically ill trauma patients. J Trauma 2005;59:80-82.

Several studies over the past decade and a half have highlighted the deleterious effects of hyperglycemia on patients with head injury and patients with burns. More recent have been studies from ICUs showing that tight control of Serum glucose with titrated Insulin infusion can improve outcome in many different types of critically ill patients. In this, the first of a series of 3 articles by the same group on much the same subject (a practice I strongly disapprove of), the authors prospectively evaluated whether admission hyperglycemia was predictive of outcome in a general population of critically ill patients with trauma. All nonDiabetic patients with trauma admitted to the R Adams Cowley Shock Trauma Center for longer than 48 hours or who were admitted to the intensive or intermediate care unit, over a 2-year period, were studied. Admission serum glucose levels were obtained on each patient and correlated with eventual outcome. A total of 1003 patients were studied, 73% of whom were male. Of interest, after matching for injury severity, women were more likely to be hyperglycemic on admission. Patients who were hyperglycemic on admission had a significantly greater number of ventilator days, ICU days and hospital days, greater risk of infection, and 2.2 times greater risk of mortality. Because severe trauma stimulates the stress response resulting in an increase in cortisol, glucagons, and epinephrine, the question now becomes: is hyperglycemia the cause of worse outcome or rather simply a marker of a more critically injured patient?

Bochicchio GV, Salzano L, Joshi M et al. Admission preoperative glucose is predictive of morbidity and mortality in trauma patients who require

evidently routinely use a combination of crystalloids and colloids for volume resuscitation. The colloid used in this study was either gelatin or hydroxyethyl starch. The authors prospectively analyzed a cohort of 212 consecutive patients with trauma, whose mean ISS was 34, and found that, on admission to the ED, almost two thirds had mild ionized hypocalcemia, and 10% had Severe hypocalcemia. Hypocalcemia was correlated with the amount of colloid used in prehospital resuscitation but not with the amount of crystalloid used. Because we in the United States do not routinely use colloid in ED resuscitations, this information is not of paramount relevance to us. But it is a tidbit worth keeping stored away–and it is one more reason not to use colloids in immediate resuscitation.

Wang HE, Callaway CW, Peitzman AB, et al. Admission hypothermia and outcome after major trauma. Crit Care Med 2005;33:1296-1301.

This is a retrospective analysis of the Pennsylvania state trauma registry, seeking a correlation of hypothermia with outcome after serious injury. The authors, from the busy University of Pittsburgh, found that, of 38520 adult patients seen from 2000 through 2002, 5% had an admission temperature of 358C or less. multivariate regression analysis adjusted for age, severity, and mechanism of injury, revealed that these hypothermic patients had increased incidence of mortality. This finding held true for the group as a whole as well as for all subgroups, including those with isolated severe head injury (take notice, all those recommending controlled hypothermia for head injury–cf Section 6). As the authors note, the findings may have been confounded by longer prehospital times and/or more IV fluids in the hypothermic group: the authors could not get that information from the data available. They also note that, like hyperglycemia in the above articles, no causal relationship has yet been established.

immediate operative intervention. Am Surg 2005;71:171-174.

Veale WN, Morgan JH, Beatty JS, et al. Hemodynamic and pulmonary fluid

This second report from the same group at the Maryland Shock Trauma Center separated the subgroup of 252 consecutive patients with trauma who required operation, none of whom had a previous diagnosis of diabetes mellitus, into 2 groups according to whether the admission serum glucose was greater or less than 200 mg/dL. The authors found that the group with elevated glucose, otherwise matched for age and ISS, had a significantly higher incidence of postoperative infection, longer ICU and hospital length of stay, and twice the mortality–basically the same information as in the article above on the group as a whole. Now, it may be that the patients with elevated glucose were indeed just undiagnosed persons with diabetes, but it does not matter: if the glucose is high, if the body is unable to use and process glucose properly, the outcome is worse. The important question remains: is the high glucose just a marker or is it a cause of the poor outcome? Furthermore, will treating this high glucose with insulin improve things? The answers to these questions await further studies.

Bochicchio GV, Sung J, Joshi M, et al. Persistent hyperglycemia is predictive of outcome in critically ill trauma patients. J Trauma 2005;58:921-924.

The same group then extended their study to prospectively evaluate whether persistent hyperglycemia is associated with increased morbidity and mortality in patients with trauma. A total of 942 patients were in this part of the study. The authors found that persistent hyperglycemia had a significant impact on morbidity and mortality. Again, it is not clear whether the hyperglycemia is a cause of the worsening of the prognosis or whether it is a reliable marker of critically injured patients.

Vivien B, Langeron O, Morell E, et al. Early hypocalcemia in severe trauma. Crit Care Med 2005;33:1946-1952.

The authors of this study from this level I trauma center in Paris, France, sought to determine the incidence and reasons for hypocalcemia in seriously injured patients. It should be noted at the outset that Europeans

status in the trauma patient: are we slipping? Am Surg 2005;71:621-626.

A perennial question in critical care: is the patient wet or dry? The authors of this study found that even experienced clinicians could accurately assess the volume and hemodynamic status of patients with acute trauma using clinical parameters only half the time. Especially important were the following: (a) several patients whom the clinicians thought had normal or high cardiac index who actually had low cardiac index; and (b) the clinicians were most often wrong with the more severely injured patients. Use of the invasive pulmonary artery catheter has fallen into disfavor in the past decade in much of the country, for good or for ill. The authors of this article suggest that impedance cardiography (ICG) can and should be widely available for use to noninvasively assess volume and hemodynamic status. I have no personal experience with ICG, but it sounds like it should be very easy to learn to use: 4 ECG pad-like electrodes are placed on the neck and lower anterior chest, and the examination is painless and takes less than 5 minutes and gives bedside readouts of cardiac index and thoracic fluid content. But as David Feliciano asked the authors in the discussion at the end of the article, ”please explain.. .the lack of widespread usage of the [sic] ICG in the surgical ICU.. .why don’t we all have these? Is the equipment too expensive? Too sensitive? Too operator dependent?…sell it to me.” We need more convincing that ICG is as accurate, as the authors of this article allege but do not prove.

Murrell Z, Haukoos JS, Putnam B, et al. The effect of older blood on mortality, need for ICU care, and the length of ICU stay after major trauma. Am Surg 2005;71:781-785.

The development of the modern blood bank was a tremendous step forward in medical care. But surgeons have known all along that fresh blood is somehow better than stored blood. Stored blood contains the waste products of dead white blood cells, which survive in vitro only a few days, and the red cells themselves, although alive, are less deformable and less efficient at transporting oxygen than fresh cells. The authors of this article retrospectively reviewed the records of

275 consecutive patients with trauma admitted over a period of 16 months, who received 1 or more U of Packed red blood cells (PRBCs). They found that patients who received older blood had a significantly longer ICU length of stay, which they thought might be a surrogate marker for a higher incidence of multiorgan failure, although there was no significantly higher Inhospital mortality. I do not think this study adds much to our practice. We all eagerly await the development of synthetic blood substitutes but have no good proof that our blood banks need to change their protocols yet.

Masuno T, Moore EE, Cheng AM, et al. Prehospital hemoglobin-based oxygen carrier resuscitation attenuates postinjury acute lung injury. Surgery 2005;138:335-341.

The team at the University of Colorado Health Sciences Center have been investigating hemoglobin-based oxygen carriers as a substitute for standard crystalloid solutions for more than a decade; some of their work has been included in previous years’ annotated bibliographies. This rat study focused on the demonstration that using their polymerized human hemoglobin solution to resuscitate rats subjected to hemorrhagic shock led to faster recovery of Tissue oxygenation and correction of Base deficit, as well as decreased neutrophil infiltration and inflammatory injury in the lungs, compared with using crystalloid. This group continues to pioneer the use of hemoglobin-based oxygen carriers; they feel it is ready for use in patients, and we should see early clinical trials soon. It is almost, but not quite, ready for general prime time.

The Miami Trauma Clinical Trials Group. Splanchnic hypoperfusion- directed therapies in trauma: a prospective, randomized trial. Am Surg 2005;71:252-260.

(b) vasopressin plus fluid, (c) phenylephrine plus fluid, (d) vasopressin alone, and (e) phenylephrine alone.

The authors found that the control group received more fluid and required significantly more blood and mannitol than any of the pressor groups. There was no difference between the vasopressin plus fluid and phenylephrine plus fluid groups. The authors concluded that resuscitation strategies, which combine either phenylephrine or vasopressin plus crystalloid are superior to either fluid alone or pressors alone to correct Vasodilatory shock after traumatic brain injury.

Rodas EB, Malhotra AK, Chhitwal R, et al. Hyperacute abdominal compartment syndrome: an unrecognized complication of massive intra- operative resuscitation for extra-abdominal injuries. Am Surg 2005;71: 977-981.

The abdominal compartment syndrome (ACS), in which intra-abdom- inal pressure rises to the point at which various organ systems start to fail, can be caused by the sudden appearance of a mass lesion–often a large hematoma–in the abdomen, or by diffuse swelling of the normally present organs. This article describes 5 patients who developed ACS quite rapidly while they were under the drapes being operated on for extraabdominal injuries. Of the patients, 4 had massive hemorrhage and 1 had a 70% bovine serum albumin burn and developed ACS during burn wound excision. All had ongoing massive fluid resuscitation, receiving a mean of 15 L of crystalloid and 11 U of blood during their operations. Abdominal compartment syndrome was manifest as hypotension, increased airway pressures, and acidosis; it was recognized by measurement of bladder pressure, and all parameters promptly improved with abdominal decom- pression. All patients were found to have large amounts of intraperitoneal serous fluid as well as massive edema of the abdominal viscera. All 5 patients’ abdomens were able to be closed in 2 to 5 days.

This is an excellent study from several points of view, although it has

significant weaknesses. The first strength is that the authors designed a protocol based on all available laboratory and clinical work that is directed at improving splanchnic blood flow in critically ill patients with trauma. Research over the past decades has gradually taken us in the direction of believing that the mesenteric circulation is a primary player in the development of delayed septic and organ failure complications. The busy trauma/critical care service at the University of Miami/Jackson Memorial Hospital tried to use all the tools currently available in clinical medicine, including monitoring of gastric intramucosal pH, neutralization of gastric acid, aggressive hemodynamic support using invasive monitor- ing, using ”splanchnic sparing” vasopressors when needed, and admin- istering antioxidant agents, to manage their critical patients. It was a prospective, randomized trial of 151 patients, controlled by using standard techniques of management only in a similar group of patients. A final strength of the report is that the authors had the honesty to conclude that their well-thought-out, well-meaning, and laborious efforts were all for naught: there was no significant effect on any clinical outcomes. The main weakness of the article is that the authors bit off more than they could chew and tried to do too many things at once. There were too many variables being manipulated in an inherently heterogeneous population. But I applaud them for their efforts, even if all they could conclude is that we need go back to the drawing board and the laboratory bench before we are ready to apply all our reasonable and plausible current manipulations to our patients.

Feinstein AJ, Patel MB, Sanui M, et al. Resuscitation with pressors after traumatic brain injury. J Am Coll Surg 2005;201:536-545.

This laboratory study in rats compared 5 different resuscitation strategies for patients with trauma with brain injury and hemorrhage. The authors used standard management–including blood, crystalloid, and mannitol–combined with pressors titrated to the same clinically relevant end points. The experiment consisted of a 30-minute hemorrhage, causing hypotension followed by resuscitation, including (a) control fluid,

Britt RC, Gannon T, Collins JN, et al. Secondary abdominal compartment syndrome: risk factors and outcome. Am Surg 2005;71:982-985.

This second article dealing with ACS after massive volume resuscitation describes 10 patients seen at the authors’ hospital over a period of 6 years. Four had burns greater than 40% bovine serum albumin, 3 had penetrating extremity injuries, 1 had blunt abdominal trauma, 1 was struck by lightning, and 1 had a spontaneous retroperitoneal bleed while on heparin. The average volume given these patients was 33 L (range, 12-69 L) in 24 hours. Abdominal compartment syndrome was manifest as hypotension, decreased tidal volumes on pressure control ventilation (the same as increased airway pressure on volume ventilation), low urine output, and base deficit. Overall mortality was 60%, and 43% for those who had abdominal decompression. These 2 articles highlight an important complication of massive volume resuscitation, one which we should all not only keep in mind but actively look for as we treat these patients. The authors conclude that monitoring of bladder pressure should be routine in all patients who are receiving IV fluids at more than 500 mL/h. I would add that perhaps this complication could be avoided, at least in some patients (especially those with known vasodilatory conditions such as large burns), with the use of low-dose vasopressin infusion to decrease the volume of fluid needed for resuscitation.

1. Activated factor VII

There was a minor flood of articles in 2005 dealing with a very hot topic in trauma care: the use of recombinant human activated factor VII (rFVIIa). rFVIIa was first introduced for clinical use in the mid 1980s, and the US Food and Drug Administration approved its use in March, 1999, for patients with hemophilia A and B, especially those with inhibitors to Factor VIII or Factor IX. Since then, its use has undergone explosive growth in the United States and Europe, by far, mostly for ”off-label” indications. It has been described for trauma, cerebral bleeds, cirrhosis, gastrointestinal bleed, sepsis-induced disseminated intravascular coagulation , necrotizing pancreatitis, pulmonary alveolar hemor-

rhage, and reversal of coumadin anticoagulation. For trauma, it is being used not only by trauma surgeons in the ED, OR, and ICU, but is being strongly considered for use by paramedics in the prehospital arena, as well as in rural areas and forward-deployed military units without access to blood bank or definitive surgery. The first use in a trauma patient was reported in Lancet in 1999; the first US trauma patient use was reported in 2002. Most reports of its use in surgical and trauma patients are dramatic. For example, the first US use in a trauma victim came after the patient had had 3 operations, 2 interventional angiographic procedures, and 105 U of PRBC, and was still bleeding at 45 hours after injury. A single dose (100 lg/kg) of rFVIIa stopped all bleeding immediately.

The mechanism of action of rFVIIa is at least 3-fold. First, factor VIIa binds to exposed tissue factor at sites of endothelial injury to form factor X-activating complex (as well as factor IX-activating complex); thus, clot formation is thought to be limited to the site of injury. Second, factor VIIa can also activate factor X on the surface of activated platelets even in the absence of tissue factor. Third, it also inhibits local fibrinolysis by activating thrombin-activatable fibrinolysis inhibitor. The dose for hemo-

Levi M, Peters M, Buller HR. Efficacy and safety of recombinant factor VIIa for treatment of severe bleeding: a systematic review. Crit Care Med 2005;33(April):883-890.

Levi et al reviewed all the literature for use of rFVIIa, finding 156 articles/408 patients on its use for hemophilia; 109 articles/242 patients, effective for a wide variety of platelet disorders; 37 articles/684 patients with liver disease, in which rFVIIa was quite effective for variceal bleeders, hepatectomy, and liver transplant; 47 articles/84 general surgery patients; and a randomized controlled trial (RCT) performed on patients undergoing prostatectomy, which found rFVIIa to be very effective. rFVIIa works for reversal of heparin anticoagulation therapy, but it only lasts 2 to 3 hours, vs

8 hours after prothrombin complex concentrates are used. rFVIIa also blocks/reverses the new pentasaccharide anticoagulants fondaparinux and idraparinux and, perhaps, even ximelagatran (a new direct thrombin inhibitor). The adverse effects were few: serious adverse events in all patients with hemophilia were less than 1%, and in all other patients, DVT occurred in 1.4%, probably many unrelated to the factor VIIa.

philia is 90 lg/kg every 2 hours until hemostasis is achieved; for patients

with trauma, most centers are using 90 to 120 lg/kg. Its half life is

2.3 hours, and a second dose can be given after 2 hours if needed. Its trade name is NovoSeven (Novo Nordisk, Inc). Its cost is huge: it comes in 1.2-mg ($731), 2.4-mg ($1427), and 4.8-mg ($2826) vials; a dose of 90 lg/kg for an 80-kg person thus costs about $4500.

Here are the most important articles, all from 2005:

Holcomb JB. Use of recombinant activated factor VII to treat the acquired coagulopathy of trauma. J Trauma 2005;58(June):1298-1303 [Commander USA ISR/Brooke].

This is the best review article. Get it and read it.

Haan J, Scalea T. A Jehovah’s Witness with complex abdominal trauma and coagulopathy: use of factor VII and a review of the literature. Am Surg 2005;71:414-415 [Maryland Shock Trauma].

This is a case report of a 23-year-old Jehovah’s witness patient with blunt splenic and left diaphragmatic injury, and left hepatic vein injury. The patient was coagulopathic and bleeding diffusely after surgical control. After a single dose (90 lg/kg) of FVIIa, the bleeding resolved almost immediately.

Chino J, Paolini D, Tran A et al. Recombinant activated factor VII as an adjunct to packing for liver injury with hepatic venous disruption. Am Surg 2005;71:595-597 [UNM].

This report describes 2 children with blunt hepatic vein injury. A single 90 lg/kg dose caused rapid, dramatic cessation of coagulopathic bleeding.

Benharash P, Bongard, F, Putnam B. Use of recombinant factor VIIa for adjunctive Hemorrhage control in trauma and surgical patients. Am Surg 2005;71:776-780 [Harbor-UCLA].

This is a retrospective review of 15 patients (9 trauma, 5 cardiac or vascular, and 1 general surgical patients) who developed coagulopathy after major procedures, with Massive bleeding despite surgical control.

Boffard KD, Riou B, Warren B et al. Recombinant factor VIIa as adjunctive therapy for Bleeding control in severely injured trauma patients: two parallel randomized, placebo-controlled, double-blind clinical trials. J Trauma 2005;59(July):8-18.

This is the most important article regarding the use of rFVIIa in patients with trauma. It was a multicenter (32 centers in 8 countries–South Africa, Israel, France, Germany, Canada, Australia, Singapore, UK) prospective double-blinded RCT. It was actually 2 simultaneous double-blinded RCT: 1 with 143 patients with blunt trauma, and 1 with 134 patients with penetrating trauma. In both, patients were between the ages of 16 and 65 years; exclusions included those in cardiopulmonary arrest, those with gun shot wound of the head or with a pH lower than 7.0, or an injury more than 12 hours old. Of interest, informed consent was waived–this study could not be done in the United States. Patients were randomized after requiring 6 U of PRBC in the first 4 hours. The first dose of rFVIIa was 200 lg/kg, followed by 100 lg/kg 1 hour later, and 100 lg/kg in another 2 hours. The findings included a significant reduction in the number of transfusions and a trend to decreased mortality and critical complications (sepsis, adult respiratory distress syndrome, MODS). Again, adverse events were few: thromboembolism in 3% to 5% (total of 12: 6 in rFVIIa patients and 6 in placebo patients).

Stein DM, Dutton RP, O’Connor J, et al. Determinants of futility of administration of recombinant factor VIIa in trauma. J Trauma 2005;59(Sept):609-615 [Maryland Shock Trauma Center].

This is a retrospective review from the years 2001 to 2003: of 81 patients at Maryland Shock Trauma Center who received rFVIIa, this article studied the 46 who were given rFVIIa for acute hemorrhagic shock within the first 4 hours. The dose was 100 lg/kg; the average number of units before giving rFVIIa was 25 (!!). Twenty-six patients responded with clinical resolution, only 2 requiring a second dose. Twenty patients did not respond, kept bleeding, and went on to die of hemorrhagic shock. Independent predictors of nonresponse were a higher serum lactate (12 vs 7), higher prothrombin time (21 vs 15) at the time of administration, and lower revised trauma score (4 vs 6) at the time of admission. Number of units transfused was not a predictor.

These patients were in dire straits: the mean international normalized ratio

was 4.3; partial thromboplastin time , 125; temperature, 35.48C; base deficit, 5.2; and lactate, 8.9 mg/dL. The mean number of transfusions in 24 hours before the administration of factor VIIa was 17 U of PRBC, 8 fresh frozen plasma, 23 platelets, and 20 cryoprecipitate. Of the 15, 12 had partial or complete hemostatic response to 90 to 120 lg/kg (average initial dose, 7.3 mg); 3 patients received a second dose of 60 to 90 lg/kg. These 12, all of whom had been expected to die, stopped bleeding and lived at least 48 hours; 7 survived to hospital discharge. There were no thrombotic adverse effects.

MacLaren R, Weber LA, Brake H et al. A multicenter assessment of recombinant factor VIIa off-label usage: clinical experiences and associ- ated outcomes. Transfusion 2005;45:1434-1442.

These authors reviewed all 701 patients getting rFVIIa at 21 US academic medical centers during the 28-month period (January 2002 through May 04). Ninety-two percent were treated with the drug for an off- label use: 38% to prevent bleeding in coagulopathic patients, and 62% to treat bleeding: surgical in 37%, gastrointestinal in 30%, intracerebral

hemorrhage in 13%, and pulmonary in 10%. Two thirds received a single dose, a median of 75 lg/kg when used for prevention and 90 lg/kg when used for bleeding. Of note, the PT and international normalized ratio were always found to normalize after rFVIIa, whereas the activated PTT did not. In the prevention group, 14% bled, 5 of which were major bleeds. In the bleeding group, 53% stopped bleeding within 6 hours, but 25% rebled within 48 hours. arterial pH below 7.2 was the only predictor of lack of response to rFVIIa. There was a 10% incidence of adverse events within 24 hours: 8 VTE, 3 myocardial infarction (MI), 3 DIC, 2 cerebrovascular accident (CVA), possibly although not definitely related to the rFVIIa.

O’Connell KA, Wood JJ, Wise RP et al. Thromboembolic adverse events after use of recombinant human coagulation factor VIIa. JAMA 2006;295:293-298 [FDA].

This report comes from the FDA’s Adverse Event Reporting System (”MedWatch”), reviewing all AE’s between licensure of rFVIIa in March 1999 and the end of 2004, both from US and foreign/overseas hospitals. Of note, the authors state that the number of Adverse Event Reporting System reports usually significantly underestimates the actual number of occurrences. The authors found 168 reports describing 185 Thromboembolic events; 151 of these were from use for off-label indications. Of the 185 thromboembolic events, 39 were embolic CVA, 34 were acute MI, 26 were other arterial thrombosis, 32 were pulmonary embolism, and 42 were DVT. Half of the adverse events occurred in the first 24 hours after drug administration. Of the 50 reported deaths, 36 were due to the thromboembolic event.

The following is a summary of where we are vis-a`-vis rFVIIa:

rFVIIa is an exciting new agent for use in coagulopathic bleeding patients

It is extremely expensive, and its use at present is restricted to ongoing clinical trials

The following are some of the main questions:

1. What is the optimal dose for the various indications?
2. Should it be used only for patients who are coagulopathic? And coagulopathic by what criteria?
3. When should it be used in actively bleeding patients?
   • Too earlyYwasted money, possible side-effects
   • Too lateYthe horse is out of the barn
4. There is ongoing concern about causing diffuse systemic micro- thrombi, causing not only DVT/pulmonary embolism, CVAs, MIs, and others but also later MODS
5. Laboratory studies/future directions

Powers KA, Zurawska J, Szazi K, et al. Hypertonic resuscitation of hemorrhagic shock prevents alveolar macrophage activation by preventing systemic oxidative stress due to gut ischemia/reperfusion. Surgery 2005;137:66-74.

The problem of reperfusion injury after ischemia has remained central in many fields, ranging from stroke and MI to resuscitation from shock to transplantation, and considerable progress in unraveling the cellular and molecular events occurring during reperfusion has been made in the past 2 decades. Briefly, ischemia leads to generation of Reactive oxygen species, which then activate various nuclear transcription factors in macrophages and neutrophils, leading to production and release of inflammatory

oxidant stress and the consequent production of inflammatory mediators. The present study is one in a series of studies in rats of hemorrhagic shock. The authors demonstrated that the use of small-volume hypertonic saline during resuscitation prevented both gut injury as reflected histologically, as well as lung leukosequestration and priming of alveolar macrophages.

Hoen S, Mazoit J-X, Asehnoune K, et al. Hydrocortisone increases the sensitivity to alpha-one adrenoreceptor stimulation in humans following hemorrhagic shock. Crit Care Med 2005;33:2737-2743.

Deitch EA. The swinging pendulum of corticosteroid use in the intensive care unit: has it swung too far or not far enough? Editorial, Crit Care Med 2005;33:2842-2843.

This is an interesting study for those of us who have been following the resurgence of interest in using steroids for patients in septic shock in the ICU. The authors studied 23 moderately injured patients with trauma, ages 20 to 42, after volume resuscitation had restored normal hemodynamics. The patients were first given an adrenocorticotropic hormone stimulation test with an IV bolus of 250 lg corticotrophin; a surprising 43% were ”nonresponders”: their total plasma cortisol remained less than 9 lg/dL. Phenylephrine infusion was given 12 to 24 hours later, and a dose-response curve vs mean BP was constructed for each patient before and after a bolus dose of 50-mg hydrocortisone. All patients, responders and nonresponders alike, demonstrated an increase in sensitivity to a₁-adrenoreceptor stimulation. Now, there are previous experimental/animal studies, which have shown decreased vascular responsiveness to adrenergic agents after acute hemorrhage, thought to be somehow related to tumor necrosis factor (TNF), but this is the first human study with such a finding. It has been known since the first studies of patients with Addison’s disease that glucocorticoids maintain vascular tone and potentiate the actions of catecholamines. But this present (and other) studies suggest that almost physiologic doses of glucocorticoid were effective even in patients with normal or even stress-elevated cortisol levels. Does this represent relative cortisol resistance? The mechanism of the effect is unknown; among the questions to be answered is, what are the roles of cytokines and nitric oxide? Does this represent catecholamine receptor resistance and/or down- regulation? This is an interesting material for future work.

Wu R, Dong W, Zhou M, et al. A novel approach to maintaining cardiovascular stability after hemorrhagic shock: beneficial effects of adrenomedullin and its binding protein. Surgery 2005;137:200-208.

The advent of molecular biology has ushered in a revolution no less exciting and dramatic than the change brought to life sciences with the discovery of the microscope and the beginnings of cellular biology. From nucleic acids as genes to proteins as enzymes and receptors to smaller molecules as messengers, it is truly amazing to watch as the molecular mechanisms underlying macroscopic clinical pathology are unraveled.

Adrenomedullin (AM) is one of the most recently discovered (1993) molecular species that is being found to have a prominent role in human physiology. It is a potent vasodilator peptide, found most especially in the intestinal wall, which acts in an autocrine (same cell) or paracrine (local regional) fashion. Adrenomedullin has a single specific circulation plasma binding protein, AM-binding protein 1, which greatly potentiates AM’s vascular relaxation effects. The present study demonstrated that giving AM/ AM-binding protein 1 to rats in hemorrhagic shock improved cardiac function and splanchnic organ blood flow and reduced elevated TNF-a to normal levels. This is preliminary work, but–especially since the gut does appear to be at least partially the ”motor of Multiorgan dysfunction”–it may herald important advances to come.

mediators. During conditions of global ischemia (shock), the intestinal

mucosa is a, if not the, primary source of these inflammatory mediators. Hypertonic saline appears to have more beneficial effects other than just allowing smaller volumes of resuscitation fluid and, thus, decreasing edema. It has direct vasodilating properties and prevents capillary narrowing, reducing microcirculatory dysfunction and, thereby, reducing

Earle SA, Proctor KG, Patel MB, et al. Ubiquitin reduces Fluid shifts after traumatic brain injury. Surgery 2005;138:431-438.

Here is another candidate for future use in acute trauma resuscitations. Ubiquitin is a small peptide found in all eukaryotic cells, which appears to

have multiple roles, some of which are involved in immunomodulation. Previous work by these authors, from the Miami Ryder Trauma Center, noticed that administration of ubiquitin significantly reduced fluid require- ments after experimental trauma. The present study, in swine, showed that a bolus of ubiquitin given to animals 35 minutes after sustaining a standardized brain injury and subsequent hemorrhage, again significantly decreased the amount of fluid resuscitation needed to stabilize hemody- namics. It also attenuated the rise in ICP and increase in lung water seen after resuscitation of controls. The mechanism of ubiquitin’s action are not yet understood, but it appears that the endothelium is the primary target.

1. Head

Vincent J-L, Berre J. Primer on medical management of severe brain injury. Crit Care Med 2005;33:1392-1399.

This is an excellent ”concise definitive review” of the current state of the art of the management of severe (GCS score V8 ) traumatic brain injury. The highlights for EM physicians include the following: (1) using repeated GCS and pupillary evaluation to monitor neurologic status; (2) classification of CT findings of diffuse injury grades 1 to 4; (3) discussion of indications for ICP and jugular venous oxygen saturation monitoring; (4) the importance of early resuscitation based on the VIP rule (ventilate, infuse, pump) and of avoiding deleterious systemic influences including hypotension, hypoxemia, fever, and hyperglycemia; (5) the role, if any, for hyperventilation; (6) control of ICP and cerebral perfusion pressure; and (7) current controversies, including the role of hypothermia, steroids, and other experimental therapies. For anyone who cares for patients with major head trauma, this article is worth getting, reading, and circulating among your colleagues.

Esposito TJ, Reed L, Gamelli RL, et al. Neurosurgical Coverage, Essential,

Utilization Project’s Nationwide Inpatient Sample database, collecting outcome information on patients 21 years old or younger who had traumatic brain injury and required endotracheal intubation (indicating serious injury) from 1988 through 1999. The database approximates a 20% sample of US community hospitals, and provided a national estimate of hospitalizations over the 12-year period; the estimate totaled almost 100000 patients. The main findings were the following:

Mortality rate decreased by 8% and was paralleled by an 11% increase in the use of ICP monitoring over the period; an estimated almost 6500 children survived because of improved treatment. For bean-counters, improved technology generated $17 billion in benefits, assuming a Life expectancy of 50 additional years, a value of $100000 per life-year and a discount rate of 3% (?!).

Lack of health insurance was associated with lower use of ICP monitoring and much worse outcome: an estimated 1418 children died because of being uninsured; which translated to a $3.76-billion loss in economic benefits.

The incidence of hospitalization for pediatric head injury rose by more than 200%, or more than 20% per year, between 1988 and 1999. The editorialist asserts that 95% of these injuries are preventable with lower highway speed limits, use of helmets for bicycles and motorbikes, car seats, driver’s education for teenagers, and the like.

This article certainly does not prove that ICP monitoring is, in fact, the reason for the improved outcomes, but as an Epidemiological study, it does show at least that we are doing something right. Likewise, it does not explain at all exactly why uninsured children have such a worse fate. But insofar as its limitations go, it does tell us in which directions we need to go from here.

Desired or Irrelevant for Good Patient Care and Trauma Center Status.

Ann Surg 2005;242:364-374.

This is a ”must-read” for all who care for patients who have had neurologic injury. As most are aware, the American College of Surgeons mandates immediate neurosurgical availability for verification as a level I or II trauma center. In today’s staffing and manpower milieu, however, this may not be achievable, feasible, or maintainable in many areas of the country. The purpose of the present study was to examine the national profile of head- injured patients with trauma and determine the actual need for the specific expertise of a neurosurgeon. The National Trauma Data Bank was queried, yielding information on all head-injured patients with trauma seen at 268 trauma centers from 1994 through 2003. Craniotomy was performed in 3% of all head-injured patients. The mean GCS score of patients requiring craniotomy was 9. Subdural hematoma occurred in 18% of all head-injured patients, and 13% of these underwent craniotomy. epidural hematoma occurred in 10% of all head-injured patients, with 17% of these undergoing craniotomy. Median time to the OR for all craniotomies was 195 minutes. The authors suggest that the immediate care of head-injured patients rarely requires the immediate presence of a neurosurgeon. More than 95% of the patients were managed nonoperatively, with only 1% of all patients with trauma, and 2% to 4% of all head-injured patients, requiring craniotomy or even ICP monitoring. Most head-injured patients can be managed without craniotomy, and the authors suggest that nonneurosurgeons can judge who needs operative intervention. This conclusion is clearly controversial, and there is an excellent discussion following the article itself.

Tilford JM, Aitken ME, Anand KJS, et al. Hospitalizations for critically ill children with Traumatic brain injuries: a longitudinal analysis. Crit Care Med 2005;33:2074-2081.

Goldstein B. Severe traumatic brain injury in children: the good, the bad, and the ugly (editorial). Crit Care Med 2005;33:2140-2141.

This is a well-done and important article. The authors, mostly from the University of Arkansas, queried the national Health Care Cost and

Neidlinger NA, Pal JD, Victorino GP. Head computed tomography scans in trauma patients with seizure disorder. Arch Surg 2005;140:858-864.

The purpose of this retrospective consecutive case review was to study the set of patients with acute trauma who exhibit Seizure activity after trauma, seeking to differentiate, if possible, those with an underlying seizure disorder and those with trauma-induced seizures. Of the 12870 patients with trauma seen during the study period of 7 years, 220 (1.7%) had a preexisting seizure disorder, and 136 (1%) had new onset seizure activity.

Of the 220 patients with a prior seizure disorder, in 161 or 73%, the seizure was implicated as the initiating factor in the injury; noncompliance with antiepilecptic medication was 75% of these. Of 220 patients with a history of epilepsy, 59 had no seizure activity identified during that admission. One hundred thirty-six closed-head injury patients who had no prior history of seizure disorder had a seizure either in the field or in the ED. All patients with First-time seizure activity underwent head CT, and the head CT was significantly more likely to have intracranial abnormalities when the seizure resulted from injury (27%) rather than when it preceded injury (11%). The authors stated that their bias before the study had been that for patients with a known history of seizures and the neurologic examination would be helpful in predicting who needed a CT scan, but they instead found that neurologic evaluation was not in fact reliable in predicting new intracranial lesions in patients with preexisting seizure disorder. The main conclusion can be summed up as: all trauma patients with seizure activity, regardless of prior history of a seizure disorder, need to undergo cranial CT scanning.

Schuster R, Waxman K. Is repeated head computed tomography necessary for traumatic intracranial hemorrhage? Am Surg 2005;71:701-704.

It’s hard to remember what it was like taking care of head-injured patients before the advent of CT scanning. The GCS, in fact, was first created as a tool to help clinicians determine which patients in the United Kingdom should be transferred to 1 of the 2 CT scanners they had in the country at the time. Computed tomographic scans now are, quite rightly, the

cornerstone of management of anyone with intracranial pathology, but we probably do have too low a threshold to get repeat scans. The authors of this study sought to determine which patients who had traumatic intracranial bleeds need repeat scans. They retrospectively reviewed the records of 184 patients with CT-proven intracranial blunt traumatic hemorrhage who were initially treated nonoperatively and then who had a repeat CT scan, usually within 12 to 24 hours. Forty percent of these had subdural hemorrhage (SDH), 36% had Intraparenchymal hemorrhage, 22% had subarachnoid hemorrhage, and 4 patients (2%) had an epidural hemorrhage; 30% actually had multiple hemorrhages. Only 5 patients, all with SDH, eventually required craniotomy, and 4 of those had deteriorating neurologic status. The one patient who had a stable neurologic examination but who did require surgery was a 67-year-old female who was on long-term coumadin anticoagulation for chronic atrial fibrillation, who had an initially small SDH, which progressed to become large enough to warrant drainage. Multivariate analysis of admission risk factors associated with a need for repeat CT indicating the need for craniotomy, which include treatment with anticoagulant or antiplatelet medications (80% in those requiring craniot- omy vs 26% in those not needing surgery), elevated PT, and age older than 70 years, was performed. The findings of this study need to be verified by studies of larger numbers of patients (especially epidural bleeds), but it does seem that traumatic Subarachnoid hemorrhages are not like spontaneous subarachnoid hemorrhages, and are unlikely to rebleed; SDHs are the ones that most need revisualization.

Smith MJ, Stiefel MF, Magge S, et al. Packed red blood cell transfusion increases local cerebral oxygenation. Crit Care Med 2005;33:1104-1108.

We are gradually moving away from transfusing blood, not only because of the risk of transmission of serious infectious (mostly viral) diseases and transfusion reactions but also because of its depressant effects on the immune system. But we do need to have more precise information concerning what exactly are some of the necessary beneficial effects of transfusion, so that we don’t avoid it when we could do some good. The authors of this report studied 35 consecutive volume-resuscitated patients seen over a 2-year period with acute central nervous system (CNS) injury: 12 with subarachnoid hemorrhage, 7 with SDH, and 16 with traumatic brain injury. They measured transcutaneous brain Po₂ (Pbto₂), which reflects local Po₂ in white matter, before and after transfusion of PRBCs–a mean volume of 2 U, enough to produce a mean increase in hematocrit from 25 to

29. Transfusion increased Pbto₂ in 26 (74%) of the 35. The mean PbtO₂ increase was 15%, although it was 49% in the 26 in whom it did increase. Intracranial pressure, cerebral perfusion pressure, Sao₂, and Fio₂ were all unchanged. Of note, older blood worked just as well as fresher units. This study does not, of course, prove that blood transfusion is actually beneficial to the injured brain. For one thing, whether cellular oxygenation is improved remains unknown. But it is suggestive, and follow-on work should be done.

Polderman KH, Rijnsburger ER, Peerdeman SM, et al. Induction of hypothermia in patients with various types of neurologic injury with use of large volumes of ice-cold intravenous fluid. Crit Care Med 2005; 33:2744-2751 [Netherlands].

It has long been known that moderate hypothermia (338C) is protective of the CNS; it is used daily all over the world in elective cardiac and neurosurgery. Some recent studies have documented good results of using hypothermia for comatose survivors of cardiac arrest, although no benefit has yet been shown when used in victims of acute stroke or trauma. The benefit shown in these studies was present, although the cooling was slow, using air convection only. The present article reports the results of using rapid cooling of 134 consecutive patients after Anoxic brain injury–61 after CPR for cardiac arrest, 34 after subarachnoid hemorrhage, 16 after traumatic brain injury, and 9 others (postasphyxia without CPR, increased ICP from stroke or hepatic coma, hypoxia from severe ARDS). The hypothermia was induced in the ICU using at least a central venous pressure and an indwelling Arterial line as monitors. Cooling was accomplished with 48C IV saline (average amount

required 1500 mL) and cooling blankets for 1 hour to a achieve a target core temperature of 328C to 338C. Findings included a decrease in pressor requirements and an increase in mean BP by 15 mm Hg, especially in patients who were hemodynamically unstable at the outset. There were no adverse effects on hemodynamics, electrolytes, glucose, cell counts, bleeding, and no pulmonary edema. The authors conclude that rapid induction of moderate hypothermia is safe in acutely brain-injured patients. Note, however, that this was not a study of neurologic outcome: it demonstrated only that rapid cooling can be done safely.

Carric MM, Tyroch AH, Youens CA. Subsequent development of thrombo- cytopenia and coagulopathy in moderate and severe head injury: support for serial laboratory examination. J Trauma 2005;58:725-730.

The primary objective in the care of patients with traumatic CNS injury is to prevent secondary neurologic injury, which can result from hypotension, hypoxia, hyperthermia, hyperglycemia, hypocapnia, hypercapnea, acidosis, anemia, coagulopathy, and/or thrombocytopenia. This article addresses the last 2 conditions. Thrombocytopenia and coagulopathy are 2 secondary insults that are recognizable and treatable. The authors sought to determine the incidence of thrombocytopenia and coagulopathy in moderate and severe head-injured patients both at the time of admission and subsequently. Coagulopathy was defined as a prothrombin time greater than 14.2 or PTT time greater than 38.4 and thrombocytopenia as a platelet count less than 150000/lL. All patients with blunt trauma with an initial GCS score of less than 14 or Abbreviated Injury Score less than 3 were categorized into moderate (GCS score of 9-15) and severe (GCS score b9) traumatic brain. Patients with cirrhosis, renal failure, or chronic anticoagulants use were excluded. One hundred eighty-four patients with blunt trauma with moderate trauma brain injury and 118 with severe trauma brain injury were evaluated. Thrombo- cytopenia was present in 14% and coagulopathy in 21% of these patients on initial evaluation. By the third postinjury day, the percentage of patients with thrombocytopenia increased to 46%, and the percentage with coagulopathy increased to 41%. In their discussion, the authors note that the brain contains a high concentration of tissue thromboplastin, which is thought to be released into the circulation after traumatic brain injury, activating the extrinsic clotting cascade. Damaged cerebral endothelium also activates platelets and the intrinsic clotting cascade to produce intravascular thrombosis with resultant depletion of platelets and coagulation factors. Thus, a combination of DIC and consumption coagulopathy ensues. It is important that these conditions be anticipated, sought for, and treated with fresh frozen plasma and platelets as needed.

Cohan P, Wang C, McArthur DL, et al. Acute secondary Adrenal insufficiency after traumatic brain injury: a prospective study. Crit Care Med 2005;33:2358-2366.

Relative and absolute adrenal insufficiency has been recognized over the past decade as being a not infrequent occurrence in critically ill patients (cf Section 1). The authors of this article did a prospective cohort study of

80 patients with moderate to severe (GCS score 3-13) traumatic brain injury, compared with 41 patients with trauma with ISS greater than 15 but without brain injury. They obtained twice daily Serum cortisol and adrenocorticotropic hormone levels on all patients for up to 9 days during the patients’ ICU stays. None of the patients developed sepsis during the study period. Adrenal insufficiency was defined as (a) 2 consecutive cortisols of 15 lg/dL or less, which was the 25th percentile or less for non- brain-injured patients or (b) any cortisol less than 5 lg/dL. At least transient adrenal insufficiency occurred in more than 50% of brain-injured patients and was associated with younger age, greater ISS, early Ischemic events (hypotension, hypoxia, or Severe anemia), and the use of etomidate or high- dose propofol or pentobarbitol. Lower cortisol levels were associated with lower BP and greater vasopressor use. Of importance for ED physicians, patients who developed adrenal insufficiency did so at a mean of 2.4 days post injury, with a range of 12 hours to 6 days; 75% did so in the first 4 days. As most ED physicians who use etomidate know, this drug inhibits adrenal

steroidogenesis, but this effect is unlikely to persist for more than 24 hours; the reasons for the association with propofol and pentobarb are unclear. The authors are now conducting a randomized trial of stress-dose hydrocorti- sone for brain-injured patients.

1. Neck

Brown CVR, Antevil JL, Sise MJ, et al. Spiral computed tomography for the diagnosis of cervical, thoracic and lumbar spine fractures: it’s time has come. J Trauma 2005;58:890-896.

with trauma underwent helical CT to evaluate the cervical spine over a 2-year period; 437 consecutive patients were thus studied. Sixty-one patients (14%) had a cervical spine injury, of which 31 (7%) were unstable. CT scan had a sensitivity of 98.1%, a specificity of 98.8%, and a negative predictive value of 99.7%. The only injury missed was an undisplaced fracture of the anterior-inferior body of C3, which was visible on cross- table lateral plain film and which was confirmed to be stable on subsequent magnetic resonance imaging. The authors conclude that helical CT scanning currently represents the best modality for assessment of cervical spine injury in the unconscious patient with trauma, a conclusion with which everyone working in the trauma arena agrees.

Prompt diagnosis and treatment of spinal injuries remains one of the

most important priorities in the care of the injured patient, and all agree we must aim for zero tolerance of missed injuries. The ”gold” standard for examination of the cervical spine for several decades has been Plain radiography with 3 views: lateral, anteroposterior, and open-mouth odontoid. Similarly, for the thoracic and lumbar spine, plain radiography with anteroposterior and lateral views is standard. As we all know, on the other hand, plain films of the spine are not only often difficult to obtain but also difficult to read. In these days of the near ubiquity of CT scanners and the extremely liberal use of scanning to evaluate both the head and the chest and abdomen, perhaps it is time to formally anoint a new gold standard. Older generation scanners were, in fact, not as good as plain films for detecting certain spinal fractures, but the later generation of high-resolution helical scanners do not appear to have these limitations. The purpose of the present study was to determine whether spiral CT should be so designated as the primary imaging modality in the diagnosis of cervical, thoracic, and lumbar spinal fracture. The authors queried their trauma registry for all blunt trauma evaluations at their community trauma center over a 2-year period, including all patients with neck pain, back pain, spine tenderness, or neurologic symptoms, all of whom underwent spiral CT of the symptomatic region; patients who were intoxicated or unconscious underwent scan of the entire spine. Scans used 5-mm axial cuts, and both sagittal and coronal reconstructions were made; all studies were interpreted by an attending radiologist. Of 3537 patients with blunt trauma evaluated, 236 (7%) sustained a cervical, thoracic, or lumbar spinal fracture. The most common location for a spinal fracture was the lumbar spine (48%), followed by the cervical spine (43%) and the thoracic spine (28%). Forty-five patients, or 19%, had a spinal fracture in more than 1 anatomical region. Computed tomography missed only 2 fractures. First, a mild compression fracture of C7 in a patient with degenerative disease of the spine was detected by magnetic resonance imaging performed to evaluate persistent neck pain with a negative CT; the patient was Neurologically intact and was successfully treated with a rigid collar. The second missed injury was another mild compression fracture of T8 in a patient with degenerative disease, which was likewise causing no neurologic deficit and required no specific treatment. The authors found that spiral CT was extremely sensitive in all areas of the spine, had an overall sensitivity of 99.3% for all axial fractures, and a sensitivity of 100% for all significant fractures. Their review of the literature indicated that using plain radiography for the evaluation of the spine in patients with trauma requires greater staff resources, higher cost, and longer time to obtain a significantly less accurate test. This article does not prove any of these assertions because it did not compare CT and plain films head to head. It does show that CT is indeed extremely sensitive: only 2 minor fractures missed in more than 3500 patients. But without a direct comparison, it is not quite yet possible to conclude that ”zero tolerance” has been attained, and that there is a new gold standard in town.

Brohi K, Healy M, Fotheringham T, et al. Helical computed tomographic scanning for the Evaluation of the Cervical Spine in the Unconscious, Intubated Patient. J Trauma 2005;58:897-901.

This study, such as the one above, evaluated the use of CT for imaging the spine in patients with blunt trauma. This study looked only at Cervical spine injuries and Unconscious patients. All unconscious, intubated patients

Holmes JF, Akkinepalli R. Computed tomography versus plain radiography to screen for cervical spine injury: a meta-analysis. J Trauma 2005;58: 902-905.

This article’s goal was similar to that of the above 2, namely, to compare the performance of plain radiography with CT for identifying patients with cervical spine injuries after blunt trauma. After an extensive search of the literature, the authors found 7 high-quality studies published between 1995 and 2004, inclusive, which directly compared plain films and CT. They then pooled the published raw data from each study and calculated performance statistics. Sensitivity for fracture of plain radiography of the cervical spine according to the pooled data was only 52%, whereas it was 98% for CT; using only the 5 level III studies, the figures were 54% and 99%, respectively. Because of the heterogeneity among the 7 studies, a pooled specificity could not be calculated. The authors concluded that CT should be the initial screening test to rule out cervical spine fracture in all patients with trauma with a high risk of spine injury, namely, those with significantly depressed mental status. However, there is insufficient data at present, they added, to recommend that CT replaces plain film as the initial screening test for less-injured patients who are at low risk for fracture but who still need radiographic examination. These 2 conclusions do indeed describe the current state of affairs as of 2006.

Grogan EL, Morris JA, Dittus RS, et al. Cervical spine evaluation in urban trauma centers: lowering institutional costs and complications through helical CT scan. J Am Coll Surg 2005;200:160-165.

This article basically echoes the findings of the several above. It compared helical CT scan with plain radiography in the initial radiographic evaluation of the cervical spine in moderate to high-risk patients with trauma. The conclusion of the study was that helical CT scan is the preferred initial screening test for the detection for cervical spine fractures among moderate- to high-risk patients seen in urban trauma centers. The authors found that the use of CT to evaluate the cervical spine was more sensitive, more accurate, and more cost-effective. The initial evaluation of the cervical spine in patients of high to moderate risk of cervical spine fracture should be performed by helical CT rather than by plain films.

Woo K, Magner DP, Wilson MT et al. CT angiography in penetrating Neck trauma reduces the need for operative Neck exploration. Am Surg 2005;71:754-758.

The last 70 years has seen the pendulum swing back and forth between mandatory operative exploration and Nonoperative management several times. Until the late 1980s, it was standard practice to explore all penetrating neck injuries that penetrated the platysma (a very superficial structure) that were in zone II of the neck (between the cricothyroid and the mandible), to perform angiography for injuries above the angle of the mandible (zone III), and to do angiography, bronchoscopy, esophagoscopy, and esophagography for zone I injuries (below the cricoid). Currently accepted practice is to explore only those zone II injuries with signs of arterial injury (hemodynamic instability, pulsatile bleeding, expanding

hematoma, bruit, or thrill), definite airway injury (bubbling from the wound, etc), or extensive soft tissue disruption requiring repair; the rest are either ”nonoperatively explored” with radiologic studies or simply observed. The present article delineates the utility of CT angiography in the management of penetrating neck wounds. It was both sensitive and specific: no injuries were missed and negative operative explorations were eliminated. computed tomographic angiography also obviated the need for both esophagography and regular angiography. Computed tomographic angiography is definitely the way to go in 2006.

Munera F, Cohn S, Rivas LA. Penetrating injuries of the neck: use

concerns what they learned about unsuspected pneumothorax, but it is applicable to the ED evaluation of patients with trauma, especially in the days of Focused Abdominal (or Assessment) Sonogram for Trauma (FAST) exams. The authors describe in excellent detail the 3 principal ultrasono- graphic signs of pneumothorax (Lung sliding, the A-line sign, and the Lung point)–this section, on pages 1232 to 1233, must be read in its entirety because it is not amenable to condensing. With a portable bedside US, 43 Occult pneumothoraces were found (of 345 hemithoraces) in 197 patients. The examinations took less than 2 minutes and had excellent sensitivity and specificity when compared even with CT. This is a technique that will become standard in the ED in the near future.

of helical computed tomographic angiography. J Trauma 2005;58:

413-418.

This article echoes the results of the preceding one. It reviews the authors’ current protocol for managing penetrating neck injuries, with specific evaluation of the use of helical CT angiography. The technique used was administration of nonionic contrast delivered at a rate of 3 to

4.5 mL/s to a total administered volume of 100 mL. Three-dimensional volume-rendered images were then recreated, which gave the surgeon an excellent perspective of any lesion viewed from multiple lines of sight. The authors found that helical CT angiography was 100% sensitive, 98% specific, and had a predictive value of 93%.

Bhojani RA, Rosenbaum DH, Dikman E et al. Contemporary assessment of laryngotracheal trauma. J Thor Cardiovasc Surg 2005;130:426-432.

This article is a retrospective analysis of the charts of all 71 patients who had had serious laryngotracheal trauma, admitted to 2 busy Texan trauma centers over the 8-year period 1996 through 2004. Seventy-three percent had penetrating injury, and these were both less likely to require an emergency airway and less likely to die. Of the 19 patients with blunt injury, 11 presented with cardiac arrest, 10 of which had had collision of the neck with the automobile steering wheel; only 1 survived. Of the 19, 8 had associated Major injuries. Only 4 of the 19 required operation on the airway. Among the 52 patients with penetrating trauma, half had gunshot and half had Stab wounds. Of the 52, 40 required operative repair of the airway wound. Of particular interest to EM physicians, oral endotracheal intubation was successful in 14 of 15 with blunt trauma and 20 of 24 with penetrating wounds who required an emergency airway. The authors found that flexible Fiberoptic laryngoscopy and CT scanning were the most helpful diagnostic modalities, with esophagoscopy and Barium swallow added if necessary.

Weeks C, Moore FD, Ferzoco SJ, et al. Blunt trauma to the thyroid: a case report. Am Surg 2005;71:518-521.

This is just an Interesting case report of a fracture of, and hematoma formation within, the thyroid gland of a driver in an auto crash whose neck hit the steering wheel. She presented with neck pain and swelling, dysphagia, and hoarseness. The injury was diagnosed by CT, and conventional angiography ruled out a major Vascular injury (they should have used CT angiography). She required ICU observation for 3 days, but did not need either intubation or operative intervention, and did well.

1. Chest

Lichtenstein DA, Meziere G, Lascols N, et al. ultrasound diagnosis of Occult pneumothorax. Crit Care Med 2005;33:1231-1238.

Eversince CT scans began to be used, we have known that plain radiographs of the chest miss many pneumothoraces, both small- and even medium-sized, especially when a formal upright anterior/posterior film is not available. The intensivists from Paris, France, who wrote this article decided to perform whole-body US and CT on 200 consecutive ”undifferentiated” ICU patients upon admission to the ICU. This article

Flagel BT, Luchette FA, Reed L, et al. Half-a-dozen ribs: the breakpoint for mortality. Surgery 2005;138:717-725.

Rib fractures are the most common chest injury. Young bull riders walk around with a several ribs broken as if nothing is wrong, although 2 fractured ribs can be deadly for an elderly patient. The authors of this article queried the National Trauma Data Bank, finding information on almost 65000 patients who had 1 or more rib fractures from 1994 through 2003. Complications developed in 13%, half of which were related to the chest injury. The overall mortality was 10%, and the mortality and chest morbidity, including pneumonia, pneumothorax, empyema, and ARDS, increased for each additional rib fracture. The mortality for those with 5 or fewer ribs broken was less than 10%, whereas the mortality for those with 6 ribs broken was 11%; 7 ribs, 15%; and 8 or more, 34%. The incidence of flail chest was low: 4% to 6% among patients with 6 or more ribs broken. The use of epidural analgesia was associated with lower mortality although not with lower pulmonary complications. This article presents epidemiological numbers only, and no causal relationships are demonstrat- ed, much less, proven. Rib fractures are associated with other injuries, and it is, of course, true that the more ribs fractured, the greater must have been, on average, the energy of the trauma. For example, 60% of patients with rib fractures required mechanical ventilation, but such a high number was most certainly not due to the ribs fractures alone. The title is misleading; the information is of interest, if not of great use.

Perl M, Gebhard F, Bruckner UB, et al. pulmonary contusion causes impairment of macrophage and lymphocyte immune functions and increases mortality associated with a subsequent septic challenge. Crit Care Med 2005;33:1351-1358.

How do rib fractures, and the underlying lung contusion, produce systemic derangements? In this animal study, mice were subjected to a standardized blunt Chest blow and then studied for serum markers of inflammation and histologic changes in various tissues. The authors found that at 2 hours post injury, serum TNF-a and IL-6 levels were markedly increased, as were peripheral blood mononuclear cell cytokine production, lung myeloperoxidase activity, and lung chemokine concentrations. At

24 hours post injury, cytokine release from peritonjeal and splenic macrophages was suppressed. Additional experiments in which pulmonary contusion was followed by sepsis (induced by a standard laboratory model, cecal ligation and puncture) showed that survival was decreased when compared with either chest trauma or sepsis alone.

Sharma OP, Hagler S, Oswanski MF. Prevalence of Delayed hemothorax in blunt Thoracic trauma. Am Surg 2005;71:481-486.

Missed injuries are an anathema to both EM physicians and trauma surgeons. But there are also Delayed presentations of injuries, as well as delayed complications, that, by definition, cannot be picked up during the first evaluation. This article is a retrospective report of the authors’ experience with 8 years worth of trauma in both adults and children, focusing on hemothorax and, in particular, delayed hemothorax. Of a total of 167 patients with traumatic hemothorax, presentation was delayed in

7 (5%). The delay ranged from 22 hours to 16 days. Of the 7 delayed hemothoraces, 6 occurred within 4 days. All were males, all had multiple rib fractures, and all were secondary to blunt trauma. Most had displaced fractures, and all had bleeding from an injury to an Intercostal artery. Many of the delayed hemathoraces started as an extrapleural hematoma which then ruptured into the pleural space.

Pacini D, Angeli E, Fattori R, et al. Traumatic rupture of the thoracic aorta: ten years of delayED management. J Thor Cardiovasc Surg 2005; 129:880-884.

The cardiac surgeons of the Massachusetts General Hospital pioneered the immediate medical and delayed surgical management of patients with traumatic rupture of the thoracic aorta starting in 1971. Since their initial report in 1981, several additional series have been published attesting not only to the feasibility of such management but to the actually improved outcomes it makes possible. This article from Bologna, Italy, adds to that collection and contributes a new dimension: a series of patients treated not only with delayed repair but with endovascular stenting rather than via open thoracotomy. The charts of 69 patients with traumatic thoracic aortic rupture seen over 23 years were reviewed. Twenty-one, seen between 1980 and 1993, were managed with immediate operative repair. Four of these died, and 3 developed paraplegia. Beginning in 1993, patients were treated with initial medical management, followed by delayed repair. Of these 48, 5 required urgent repair, 3 via surgery and 2 via endovascular stent placement. Indications for urgent repair were massive hemothorax and contrast extravasation on CT, rapid enlargement of the pseudoaneurysm, and pseudocoarctation. The remaining 43 had delayed repair, 13 with an endovascular stent and 30 via open surgery. Only 2 of these patients died, and only 1 developed paraplegia. All 15 stent placements were successful and uncomplicated. This is real progress and is certainly the direction of the future.

Rousseau H, Dambrin C, Marcheix B, et al. Acute traumatic aortic rupture: a comparison of surgical and stent-graft repair. J Thor Cardiovasc Surg 2005;129:1050-1055.

This report from Toulouse, France, is similar to the article above by Pacini et al. It is a review of their hospital’s experience with traumatic rupture of the thoracic aorta from 1981 through 2003. They saw 76 patients, 6 of whom died early of other severe injuries. Of the 70 who lived to be treated, 35 underwent Surgical repair: 28 immediately, with 21% mortality and 7% postoperative paraplegia. Seven had delayed open surgical repair, with no deaths and no paraplegia. Twenty-nine had endovascular stent repair, all of which were successful with no major morbidity during a mean follow-up of 46 months (range, 13-90 months). Six had small lesions and were treated only medically, all with good results. There are limitations to using the stent for traumatic aortic tears, most importantly, the exact anatomy–how far the tear extends proximally toward the origin of the left subclavian artery or even past that into the aortic arch. But as these 2 series attest, most traumatic thoracic aortic ruptures can indeed be repaired with endovascular stents. The pathology of traumatic aortic tears is, unlike aortic dissection and degenerative thoracic aortic aneurysms, very localized to a single segment of aorta and, thus, quite amenable to placement of a short stent. The risk of paraplegia, due to obstruction of blood flow to the artery of Adamkiewicz, which originates from the descending thoracic aorta most often between T8 and T12, is, for the same reason, much decreased because the stent will not block blood flow to that area of the aorta.

Dignon P, Martaille J-F, Francois B, et al. Transesophageal echocardiog- raphy in therapeutic management of patients sustaining blunt aortic injury. J Trauma 2005;58:1150-1158.

transesophageal echocardiography for the diagnosis of traumatic thoracic aortic disruption was introduced in the early 1990s but largely

discarded in most US trauma centers in favor of spiral CT. The authors of this report describe (1) their continuing experience with TEE in the diagnosis of aortic tear and (2) their series of nonoperative management of patients with minor aortic tear. The authors have more faith in the ability of TEE to have a high sensitivity for aortic tear than do most American experts; they do admit that, compared with angiographic techniques, TEE usually fails to identify traumatic lesions of aortic branches, but they maintain that it is more sensitive to minor lesions such as intramural hematomas and small intimal flaps. They reviewed 31 consecutive patients with thoracic aortic injury after blunt chest trauma. All 7 patients with minor blunt aortic injuries, defined as intramural hematoma (1), limited intimal flap (4), or mural thrombus (3), were managed nonoperatively with Antihypertensive treatment. All did well, with follow-up TEE at a mean of 15 months, later showing resolution in 3 and no change in 4. The 4 patients with major rupture, defined as complete transection with either active bleeding or aortic obstruction, were all operated on immediately. The remaining 20 patients all had moderate injuries, defined as subadventitial rupture or modification of the geometric shape of the aorta. Of these, 9 had urgent surgical repair, and 11 had delayed repair. Five had surgery delayed because of coexisting severe head trauma, 3 had ongoing nonaortic bleeding, 2 had severe hypoxemia, and 1 had severe sepsis. Of these 11, 9 underwent delayed repair and 2 refused any surgery; no deaths in this group were attributable to aortic rupture;. Follow-up TEE in the 11 showed stability of the false aneurysm until repair in those who had surgery and complete healing in one of the patients who refused surgery.

The authors conclude that TEE successfully and reliably demonstrates significant blunt aortic injury and can provide accurate management information. They also conclude that patients with minor aortic injuries can be successfully managed nonoperatively with BP control. This last conclusion, I think, is premature, because their follow-up is quite short. If conservative management is selected, serial follow-up studies are necessary.

Brown CVR, Velmahos G, Wang D, et al. Association of scapular fractures and blunt thoracic aortic injury: fact or fiction? Am Surg 2005;71:54-57.

In the 1980s, standard trauma dogma was that a fracture of the first rib or scapula was an absolute indication for angiography of the great vessels. The reasoning was that it takes so much energy to fracture a first rib or a scapula and that it was very likely that there was a concomitant injury of the thoracic aorta or one of its main branches. Studies in the 1990s showed that the association of first rib fracture and aortic injury is not close enough to justify routine angiography solely because there is a first rib fracture. Probably more important is that newer CT scans are good enough and are performed liberally enough that we can see the great vessels anyway. The present article shows that what is true for the lack of close correlation between first rib fractures and great vessel injury is also true for scapular fractures. The authors reviewed the records of all patients with either scapular fracture or blunt thoracic aortic injury seen at 2 busy trauma centers over a 10-year period. There were 392 patients with scapular fracture and 224 with thoracic aortic injury. Ninety-nine percent of patients with scapular fracture had other injuries, but only 4 (1%) of these had thoracic aortic injury. The authors rightly conclude that scapular fracture should not be used as an indicator of possible thoracic aortic injury.

Wall MJ, Mattox KL, Wolf DA. The cardiac pendulum: blunt rupture of the pericardium with strangulation of the heart. J Trauma 2005;59:136-142.

Pericardial tear caused by blunt chest trauma and subsequent cardiac herniation is an uncommon, difficult-to-diagnose, but interesting ”zebra” to keep in the back of your mind. Portions of the atria, ventricles, or almost the entire heart can herniate, although a pericardial tear, more commonly on the left, causes shock due to simple mechanical interference with either filling or emptying the heart. The authors of this report found, based on forensic data, that 40 to 50 cases per year may be encountered in a large metropolitan area! Important clues suggesting pericardial tear and cardiac herniation include an empty pericardium on chest x-ray, FAST, or chest CT, abnormal lateral

displacement of the heart and pericardium on anterior/posterior chest x-ray, loss of pulse when changing position, and/or intermittent loss of BP especially with a change in position. On a side note, Dr Mattox states that Princess Diana of England may have had just such an injury.

Dauphine C, McKay C, De Virgilio C, et al. selective use of cardiopul- monary bypass in trauma patients. Am Surg 2005;71:46-50.

To be certified as a level I trauma center by the American College of Surgeons, cardiopulmonary bypass (CPB) must be available around the clock. Not all designated level I centers, however, have this capability. The authors of this article sought to find out how often CPB is needed. Review of their own experience at the very busy Harbor-UCLA Medical Center from 1994 through 2002 revealed that perfusionists were called in for 24 patients; 16 of these were actually placed on CPB. Indications for CPB were penetrating heart or great vessel injury in 10, blunt aortic or great vessel rupture in 3, hypothermia in 2, and bullet embolus to the pulmonary artery in 1. Of the 16, 11 survived. The authors note that 1800 patients with trauma are seen at Harbor-UCLA per year, and 10 to 15 patients per year require operative intervention for chest injuries; they ”urge that CPB capability be routine at level I trauma centers.” We all agree that would be optimum. On the other hand, no hospital is going to develop CPB capability unless they do elective cardiac surgery, nor should they.

Kaya A, Dekkers P, Loforte A, et al. Traumatic aorto-right ventricular fistula with aortic insufficiency. Ann Thorac Surg 2005;80:2362-2364.

We always include a few interesting case reports in the annual annotated bibliography just to help keep our indices of suspicion appropriately high. This report describes a 19-year-old man with a stab wound from a thin stiletto blade just to left of the sternum in the third intercostal space. On presentation, he was awake and alert with no distress, normal vital signs, and no evidence of disease or injury on examination (distended neck veins, thrill, bruit, murmur, etc) other than the small stab wound. Echocardiog- raphy–which is the key diagnostic tool to use in ruling out penetrating injury of the heart–was positive for pericardial fluid. He turned out to have a stab of the root of the aorta, the aortic valve, and ventricular septum. Keep your guard up!

DuBose RA, Karmy-Jones R. Delayed diagnosis and management of ”occult” stab wound to the heart. Am Surg 2005;71:879-881.

In that vein, here is another case to keep us alert. A 29-year old with 2 small Stab wounds just above and medial to the left nipple presented with tachycardia but with otherwise normal vital signs. Two Chest tubes placed for hemopneumothorax-drained 900-mL blood. Both portable ED US and then formal transthoracic echo were negative for both pericardial fluid and wall motion abnormalities. However, serial Troponin I levels were markedly elevated, and the ECG revealed diffuse ST elevations. Thoracic surgery was recommended, and then performed using thoracoscopy, followed by sternotomy for repair of 2 myocardial lacerations involving coronary artery branches. Noted at repair was an intense acute pericarditis.

What are the lessons learned here? (1) Echocardiography, and even subxiphoid pericardial window, can be falsely negative, especially if there is a pericardial laceration through which pericardial blood can escape into the pleural space. (2) Although cardiac enzymes, even troponin I, are not of value in making a diagnosis of blunt cardiac injury, they can be helpful in the setting of penetrating injury. (3) Thoracoscopy may well be preferable to subxiphoid pericardial window, especially for left-sided stab wounds.

Molina JE. Evaluation and operative technique to repair isolated sternal fractures. J Thor Cardiovasc Surg 2005;130:445-448.

This article is included in our Bibliography only to make EM physicians aware of the fact that some–albeit few–Sternal fractures will eventually

need to be operatively repaired. Acute sternal fractures are rarely, if ever, a problem in themselves: morbidity and mortality are due to concomitant injuries. They are therefore almost always managed by observation alone. However, a minority, especially those in which the manubrium is displaced posterior to upper body of the sternum, cause symptoms such as pain, restriction of neck movement, and kyphosis, in addition to deformity; these can and should be surgically repaired.

1. Abdomen

Blackbourne LH, Soffer D, McKenney M, et al. Secondary ultrasound examination increases the sensitivity of the FAST exam in blunt trauma. J Trauma 2004;57:934-938.

Care of the trauma patient is dynamic, and caregivers must make sure to frequently reevaluate everything, from vital signs to GCS to pulses to laboratory tests. The authors of this article performed a prospective observational study on all patients with trauma admitted to their level I trauma facility over a 9-month period to assess the value of performing a follow-up FAST examination at some point between 30 minutes and 24 hours after an initial US. Upon examining 547 patients, intra-Abdominal injury and/or fluid was found in 30% on the first examination and in more than 70% on the second exam. Specificity did not fall with this increase in sensitivity–it was 99.8% for both. Thus, both the negative predictive value and the accuracy of FAST were improved by repeating the exam. The average time that the second US was done was 250 minutes (about 6 hours) after the first study. Sequential observation–”serial exams”–and second- ary and tertiary surveys are key in caring for victims of trauma.

Choi KC, Peek-Asa C, Lowell M, et al. Complications after therapeutic trauma laparotomy. J Am Coll Surg 2005;201:546-553.

The title of this article is a little misleading. The actual purpose of this study was to analyze the effect of the amount of time elapsed from arrival at the ED to arrival at the OR, on the incidence of complications and death in patients with trauma undergoing therapeutic laparotomy in a level I trauma center. Indications for trauma laparotomy included hemodynamic instabil- ity, signs of peritoneal irritation on physical examination, positive CT or ultrasonography, positive Diagnostic peritoneal lavage, and an unexplained falling hematocrit after acute trauma. The authors retrospectively reviewed the records of 175 patients with abdominal trauma who underwent therapeutic laparotomy, with primary end points of mortality and complications, and time from arrival in the ED to start of operation as the independent variable. They found that a length of time from the ED to OR exceeding 1 hour was associated with significantly increased mortality and morbidity: patients whose operations began more than 1 hour after ED arrival were 11 times more likely to die and 3 times more likely to have complications. The authors conclude that both EM physicians and trauma surgeons need to maintain a strong sense of urgency and facilitate rapid evaluation and transfer to the OR if needed. Languishing in the ED has demonstrable deleterious effects.

Richardson JD. Changes in the management of injuries to the liver and spleen. J Am Coll Surg 2005;200:648-669.

This was the annual Scudder Oration on Trauma given at the 2005 Congress of the American College of Surgeons. Dr Richardson, one of the real deans of trauma care, reviewed the changes in the management of injuries of the liver and spleen that have occurred in the past half-century. We no longer rely upon physical examination to detect intra-abdominal blood; that function has been replaced by US, which is much more accurate. Even better and foremost among all the changes because it is the basis upon which all nonoperative management rests, is the use of CT. Nonoperative management of both liver and spleen injuries is now the accepted, even ”politically correct,” default position. The only real indications for operation are

hemodynamic instability and/or concomitant hollow viscus injury. Angiog- raphy is now widely used for control of hemorrhage, and laparoscopic and US- or CT-guided drainage of collections have proven quite useful. Nonoperative management of liver injuries has been especially successful. Nonoperative management of splenic injuries is more controversial and remains not as well proven in Dr Richardson’s view. First of all, it is unclear what the real incidence of overwhelming postsplenectomy infection in the adult is. Moreover, there is a real difference in what happens to injured spleens during the first week after injury, as compared with the injured liver; true delayed bleeding and even free rupture is a real concern for injured spleens, much more so than for injured livers. Nonoperative management of the ruptured spleen should be appropriately used but not at the risk of threatening the patient’s life. This excellent article, which reviews what has happened in the last 40 years, is both educational and enjoyable to read.

Wong C, Taubman K, Hoehne F, et al. The quality of surgical resident interpretation of after-hour CT scans of the head and abdomen in trauma. Am Surg 2005;71:772-775.

The American College of Surgeons has been trying to mandate that all trauma CTs be read by attending radiologists in real time, and of course, most EM physicians and surgeons would prefer that to happen. Although it has not yet been done, in many hospitals, CTs done at night are read by radiology residents, and then reviewed the next morning by the attendants. This study from a busy level 2 trauma center compared head and abdominal CT readings by surgical residents to the later reading by the radiology attendant, and found that only 16 (4.8%) of the 335 scans done at night over a 6-month period had discrepancies in the readings. In no instance was management or outcome affected.

This is a great performance by this group of residents, but unfortunately, it does not prove much: it is too small a study, done at only 1 institution, over too short a time.

Sharma OP, Oswanski MF, Singer D, et al. Assessment of nonoperative management of blunt spleen and Liver trauma. Am Surg 2005;71:379-386.

Planned nonoperative management of blunt splenic injuries was first introduced by the pediatric surgical community in the late 1960s and then was extended to include both blunt liver injuries as well as spleen and liver injuries in adult patients in the 1980s. Since then, significant improvements in CT imaging, interventional radiology techniques, and laparoscopy have been made. Where do we stand in 2005? The trauma group in Toledo, Ohio, retrospectively analyzed the records of all the patients, both adult and pediatric, with blunt splenic and liver injuries who they managed over the 8-year period from 1995 through 2003. They treated 279 patients, 93 children (b18 years old), 137 younger adults (18-54 years old), and 49 older adults (N55 years old). Nineteen patients died within 60 minutes of arrival and were excluded. Fifteen percent needed urgent operative intervention because of hemodynamic instability. Of the 221 (85%) managed non- operatively, 11 (5%) failed and required operation: 14% of those with combined spleen and liver injury, 5.6% of those with isolated spleen injury, and 1.5% of those with isolated liver injury. Nonoperative management was done in 99% of children, in 81% of the younger adults, and 68% of the older adults, with 0, 8%, and 10% Failure rates, respectively. Although higher injury grades were, of course, associated with more frequent necessity of operation, neither age nor CT injury grade nor CT fluid grade were contraindications to nonoperative management. The authors did not try to identify factors that may be associated with failure of nonoperative management or with need for angiography, such as active contrast

Sharma OP, Oswanski MF, Singer D. Role of repeat computerized tomography in nonoperative management of solid organ trauma. Am Surg 2005;71:244-249.

This article is an extension of the one above. The Toledo group went on to analyze further their 8-year experience with 221 patients with blunt splenic and/or liver injury managed nonoperatively to try to identify which patients should undergo repeat CT scanning. Sixty-five patients (29%) had 1 or more repeat CT scans. Forty-four of these had unchanged or improved scans. Twenty-one patients had worse scans, but only 4 of these ended up being operated on. As noted above, 11 of the 221 patients failed nonoperative management, but only 4 of these had repeat CT before surgery; all 4 of these scans showed deterioration of the CT abnormalities, but all 4 also had significant clinical deterioration–peritonitis and/or hemodynamic instability–as well. The other 7 patients who were operated on had hemodynamic instability and did not have repeat CT. Thus, repeat CT was not helpful in any of the patients described in this study. The authors conclude that the indications for repeat CT scanning are (1) suspicion of blunt intestinal and/or mesenteric injury, especially in patients with diminished sensorium or chemical paralysis; (2) occult hemorrhage in the face of a normal initial CT; (3) grade 4 or 5 liver or spleen injury, as these patients often develop complications such as biloma, pseudoaneur- ysm, hemobilia, and others; (4) infrequent healing assessment–only in those who desire to return to potentially dangerous lifestyles such as contact sports. These are reasonable and fairly widely accepted guidelines.

Christmas AB, Wilson AK, Manning B, et al. Selective management of blunt hepatic injuries including nonoperative management is a safe and effective strategy. Surgery 2005;138:606-611.

Five hundred sixty-one patients with blunt hepatic injuries seen over the 10-year period (1993 through 2003) were reviewed in this article from the trauma group at the University of Louisville (Louisville, Ky). Fifty-nine percent had grade 1 or 2 injuries. Most of these were managed nonoperatively; the ones who were operated on went to surgery for other abdominal injuries, and there were no Liver-related deaths. Two thirds of patients with grade 3 or 4 liver injuries and one third of patients with grade 5 injuries were managed nonoperatively, with a mortality of only 2.2%. Patients who required operation because of hemodynamic instability or concomitant injuries had a mortality of 30%. Eleven percent of the 378 patients managed nonoperatively required adjunctive procedures. Twelve patients underwent angioembolization for bleeding, all within the first 24 hours after injury. All of these patients had grade 3 injuries or higher, and 11 of the 12 survived. Laparoscopic or percutaneous drainage of bilomas or blood collections was performed in 18 patients. Endoscopic retrograde cholangiopancreatography (ERCP) was needed in 12 patients, 8 of whom required a stent, and 1 required operative repair of a completely transected left hepatic duct.

Among the important points emphasized in this article are the following:

hemodynamic stability is a sine qua non for nonoperative manage- ment; it is actually a sine qua non for even getting a CTscan which, of course, is the basis for any nonoperative management. The authors of this study defined as hemodynamically stable those patients who maintained, or regained after 2 L of IV crystalloid, a systolic BP greater than 90 mm Hg and a heart rate less than 100.

Patients who lose hemodynamic stability need to be taken to angiography for embolization, if it is known from the CT that liver bleeding is the source of the problem, or to the OR if the source is unknown or if there is a suspicion of other Intra-abdominal injuries.

extravasation on CT; they had only a few such patients (see articles

below). The overall mortality was 12.8% in the operated group, and 2.7% in the group were managed nonoperatively; these patients were judged to have died of their other injuries. The authors concluded, as have most trauma groups, that nonoperative management of blunt splenic and liver injuries should be attempted in all Hemodynamically stable patients, regardless of age and CT findings.

Wahl WL, Brandt M-M, Hemmila MR, et al. Diagnosis and management of bile leaks after blunt liver injury. Surgery 2005;138:742-748.

We all agree that nonoperative management is the preferred approach for all patients with blunt liver injuries who are hemodynamically stable and who do not have other operative intra-abdominal injuries. The principal

sequelae of liver injury are (1) bleeding, which can be managed by operation or angioembolization; (2) bile leaks, which can be managed by operation for major ductal injury or by a combination or laparoscopic or percutaneous drainage and ERCP with or without stenting; and (3) sepsis, caused by infection of a hematoma or necrotic tissue, which is rarely seen if the abdomen has not been opened. This article describes the incidence of bile leaks after blunt liver injury and advocates the use of hepatobiliary iminodiAcetic acid (HIDA) scanning as the best method for diagnosing major bile leaks. The authors reviewed the records of all 241 patients with blunt liver injury they treated over an 8-year period. Seventy-one percent of those treated operatively, 50% of those treated with angiographic embolization for bleeding, and 17% of those managed nonoperatively developed bile leaks; the lower percentage in the nonoperative group was simply a reflection of the fact that this group had less severe injuries. The American Association for the Surgery of Trauma Abbreviated Organ Injury Scale score was at least 4 (high-grade) for all patients who developed bile leaks. HIDA scanning was used to detect bile leaks in all nonoperated patients, and had a sensitivity of 100%: no patient with a negative HIDA developed a subsequent bile leak. The authors advocate early HIDA scanning–within the first 5 days after injury–for all patients with high- grade Liver laceration being treated nonoperatively because earlier recognition and treatment (by laparoscopic or percutaneous drainage plus ERCP) led to better outcomes. Other reports have also demonstrated a very high incidence of bile leaks in patients with grade 4 and 5 liver injuries. It is not an unreasonable recommendation.

Haan JM, Bochicchio G, Scalea TM, et al. Nonoperative management of blunt splenic injury: a 5-year experience. J Trauma 2005;58:492-498.

Series of patients with traumatic splenic injury treated nonoperatively have been published for the past 15 years. This article, from the R Adams Cowley Shock Trauma Center in Maryland, is one of the most recent updates on the current ”state of the art.” The most distinctive aspect of care at Shock Trauma is the aggressive use of angiography in patients with splenic injury. From the beginning of the study in October 1997 through June 2000, angiography was used for all patients not immediately taken to the OR. Because of ”poor therapeutic yield” in patients with AAST grades 1 and 2, they subsequently performed angiography only on patients with injury grades 3 to 5 who were being managed nonoperatively. A total of 648 patients with blunt splenic injury seen over a 5-year period were retrospectively evaluated. Two hundred eighty were unstable and were immediately operated upon. Three hundred sixty-eight underwent nonop- erative management. Seventy of these, all with injury grade 1 or 2, were managed by serial exam, hematocrit, and follow-up CT alone, and all remained stable with a 100% success rate. Of the 298 patients undergoing angiography, 166 had a negative angiogram, and 94% of these had successful nonoperative management. One hundred thirty-two patients had a positive angiogram and underwent embolization, with a nonoperative salvage rate of 90%. Salvage rate decreased with the severity of splenic injury; even so, more than 80% of patients with injury grades 4 and 5 were successfully managed nonoperatively. The CT finding of an arteriovenous fistula had a particularly high (40%) nonoperative failure rate, even after embolization. The authors’ results with an aggressive use of angiography are indeed impressive and must be taken seriously by all trauma centers. It is obviously quite expensive in terms of both personnel and facilities, and it is unclear how many trauma systems would be able to financially support such a program.

Ekeh AP, McCarthy MC, Woods RJ, et al. Complications arising from splenic embolization after blunt trauma. Am J Surg 2005;189:335-339.

Here is another perspective on angiographic embolization for blunt splenic injuries. This study was designed to assess complications that have arisen from the increasing use of Splenic artery embolization at the authors’ level I trauma facility. The study encompassed a 26-month period, during which 284 patients were admitted with blunt splenic injury, 15 of whom

underwent splenic artery embolization. Indications for angiography with embolization included the presence of a contrast blush on CT in 8 patients, a grade 4 splenic injury in 6 patients, and grade 3 splenic injury in 1 patient. major complications occurred in 4 patients (27%): postprocedure bleeding in 1 patient, splenic infarction in 2 patients, splenic abscess in 1 patient, and contrast-induced renal insufficiency in 1 patient. The only patient who required splenectomy after embolization was the 1 patient with a splenic abscess. Minor complications, including fever, bloody pleural effusion, and distal coil migration occurred in 8 patients.

McIntyre LK, Schiff M, Jurkovich GJ. Failure of nonoperative management of splenic injuries. Arch Surg 2005;140:563-568.

The management of patients who have had blunt trauma to the spleen is still controversial (cf Richardson JD, above). The spleen is not the same as the liver, and the occurrence of delayed bleeding and rupture of the spleen is a real possibility. The authors of this article, interested in determining the characteristics of patients who are likely to fail nonoperative management, reviewed the records of all patients admitted to Harborview Medical Center in Seattle with splenic trauma over a 7-year period. Almost 2250 patients were studied, and of these, 1633 (73%) were admitted to planned nonoperative management. Of note, the authors of this article consider angiography to be operative–anyone who underwent angiographic embolization was considered not to have been treated nonoperatively. This is in direct contrast with many, if not most, experts who view angiography as a nonoperative intervention, and a reasonable adjunct to nonoperative management. Fifteen percent (252) of the 1633 failed nonoperative management–they required either operation or angiography. Those who failed were more likely to be older than 55 years and/or to have an ISS higher than 25; neither GCS, Associated injuries, nor presenting hemody- namics were predictive of failure. The authors also found that nonoperative management was more likely to fail if the patients were cared for in a level III or level IV trauma center, which they thought could be due to either a relative lack of comfort with the Management strategy or to a lack of effective resources for close observation of the patients. Counting all patients, including those who were managed operatively from the start, nonoperative management was successful in 62% of all patients with blunt splenic injury. This is an important article with a large number of patients. The 85% success rate of nonoperative management in those for whom it was tried, especially considering that need for angiography was considered a failure, is impressive and consistent with other reports.

Peitzman AB, Harbrecht BG, Rivera L, et al. Failure of observation of blunt splenic injury in adults: variability in practice and adverse consequences. J Am Coll Surg 2005;201:179-185.

The purpose of this study was very much the same as the one above: to determine the common characteristics in adult patients who fail nonoper- ative management of blunt splenic injury. The lessons learned, however, are quite different. The authors of this study reviewed records from 27 different trauma centers (26 level I and 1 level II) for the year 1997, through the Multi-Institutional Trials Committee of the Eastern Association for the Surgery of Trauma. Of the 1488 adult patients with blunt splenic injury, 913 (61%) were managed initially nonoperatively, and 97 (11%) failed. Failure of nonoperative management was defined as any patient who was admitted to the ICU or the floor, with a diagnosis of blunt splenic injury with planned nonoperative management, who later required laparotomy. Analysis of the failures revealed, however, that many of the patients, including most of the 10 who died, were, in fact, hemodynamically unstable and should not have been triaged to nonoperative management in the first place. The most significant point of these findings is that essential criteria for nonoperative management includes hemodynamic stability. Secondly, nonoperative management will fail in some patients; indications for operation should include hemodynamic decompensation, new or worsened abdominal pain, falling hematocrit, persistent tachycardia, and the need to rule out an associated hollow viscus injury. The concern today with nonoperative management, as voiced above by Dr Richardson, is that the pendulum

toward nonoperative management may have swung too far and that indications for laparotomy are being ignored.

Todd SR, Arthur M, Newgard C, et al. hospital factors associated with splenectomy for splenic injury: a national perspective. J Trauma 2004;56:1065-1070.

The article by McIntyre et al above concluded that nonoperative management of patients with splenic injury was more often successful at level I and II trauma centers than at level III or IV centers. The purpose of the present article was to determine whether the odds of laparotomy and splenectomy among patients with spleen injuries differed between urban and Rural hospitals when adjusting for other important variables. The authors used a national inpatient sample compiled in the Healthcare costs and Utilization Project. The authors discovered that by reviewing the records, the incidence of nonoperative management varied remarkably between urban teaching hospitals, urban nonteaching hospitals, and rural hospitals. Splenic salvage was primarily achieved by nonoperative management. They found that 40% of the patients underwent a laparotomy with a 28% incidence of splenectomy in the urban teaching hospitals. In contrast, in the urban nonteaching hospitals, there was a 35% splenectomy rate. In the rural hospital, there was a 36% incidence of splenectomy. In an effort to explain these differences, the authors recognized that in teaching hospitals, emergency surgical capabilities are readily available. Their resources include in-house OR staff and anesthesia personnel and the ready availability of blood products. In addition, surgical house staff members are available to manage patients with nonoperative splenic trauma. In the urban, nonteach- ing, and rural hospitals, different practice patterns may be necessary, resulting in a higher splenectomy rate and a lower rate of attempt at and of success of nonoperative management.

Harbrecht BG, Zenati MS, Alarcon LH, et al. Is outcome after blunt splenic injury in adults better in high-volume trauma centers? Am Surg 2005;71:942-949.

This study is along the same lines as the article above. There has been a big push over the last decade to mandate that various high-risk surgical procedures, such as Coronary artery bypass grafting, pancreaticoduodenec- tomy (Whipple), and esophagectomy, be performed only in those hospitals that do many of the procedures–the ”high-volume” ”centers of excel- lence.” There are good a priori reasons to do such a thing, and there are some fairly good data showing that outcomes for these procedures are indeed better in high-volume centers (HVC). It is more than just that the surgeons alone are more experienced; it involves the whole system, OR organization and equipment, recovery room and ICU facilities and personnel, floor nursing, and even outPatient evaluation and follow-up. The entire level I through IV trauma center system is predicated on the same principles. This article asks if a particular subgroup of patients with trauma, those with blunt splenic injury, need to be triaged to the HVCs (cf McIntyre et al and Todd et al, above). The authors queried the Pennsylvania state trauma database and compared the outcomes of 1829 patients with splenic injury treated at 11 HVCs (N240 patients per year with ISS N15) to that of 1040 patients treated at 16 low-volume centers (LVC). Patients at HVCs were slightly more likely to be managed nonoperatively (70% vs 66%), but the failure rates and the mortality rates for nonoperative treatment were similar. Patients at LVCs were more likely to be admitted to an ICU, despite having a somewhat lower mean ISS, but overall ICU and hospital length of stay were shorter at LVCs, consistent with the lower mean ISS. The authors found basically that, in Pennsylvania, the quality of care given to patients with blunt trauma of the spleen at the LVCs was just as good as at the HVCs.

Statter MB, Liu DC. Nonoperative management of blunt splenic injury in Infectious mononucleosis. Am Surg 2005;71:376-378.

This case report is included just to show that nonoperative management of blunt splenic injury can even be extended on occasion to those with

diseased spleens. The patient reported on was a 14-year old with a grade III injury of a spleen that was at least double- or triple-normal size due to infectious mono. The trauma had been a low-energy blow to the flank, there were no other injuries, and the patient was successfully managed without need for operation or even transfusion.

Tyroch AH, McGuire EL, McLean SF, et al. The association between Chance fractures and intra-abdominal injuries revisited: a multicenter review. Am Surg 2005;71:434-438.

We have all been taught that Chance fractures–transverse fractures of the vertebral bodies of the lower thoracic or lumbar spine, caused by hyperflexion and often associated with seat belt injuries–are commonly associated with injury to abdominal organs. This multicenter review was undertaken to establish just how close this association is. The records and radiographs of all patients with thoracic and lumbar spine fractures seen at

7 level I trauma centers in several states over a 6-year period were reviewed, and 79 Chance fractures were found. More than two thirds were from motor vehicle crashes , and 13% were from falls. Half of the fractures occurred at T12 or L1, and the rest at T10, T11, L2, L3, or L4. Only 17% had accompanying spinal cord injury. One third of the patients had intra-abdominal organ injury: 12, Small intestine; 7, colon; 2, stomach; 10, spleen; 4, kidney; 4, pancreas; 4, mesentery; and 3, liver. Half of the patients who had an intra-abdominal injury had more than 1 Organ injured. Of particular note, 85% of those with Chance fracture plus abdominal wall contusion–a ”seat-belt sign”–had intra-abdominal injury, whereas only 14% of those without an abdominal wall contusion had intra-abdominal organ injury (actually, not all these abdominal wall contusions were from seat belts: a quarter of the patients with intra-abdominal injury had not been wearing seatbelts). The moral of the story: the association between Chance fractures and intra-abdominal organ injury may not be quite as high as the almost 90% previously reported, but it is still high, and hollow viscus injury is especially common. Thus, CT, which is known to be rather poor at detecting hollow viscus injury, cannot be one’s sole Diagnostic modality: one needs to maintain very high index of suspicion. Furthermore, it could be argued that the combination of a Chance fracture and ”seat belt sign” is an indication for operative abdominal exploration.

Huerta S, Bui T, Porrai D, et al. Predictors of morbidity and mortality in patients with traumatic duodenal injuries. Am Surg 2005;71:763-767.

Injuries of the duodenum are fortunately not often seen. The duodenum is buried deep within the body, protected on all sides; if it is injured, adjacent organs are almost always injured as well. This retrospective review of the records of all patients with duodenal injury seen at a very busy urban trauma center (UC-Irvine) over a 7-year period was undertaken to attempt to identify which factors were associated with death and complications. Previous studies from the past 3 decades identified Delay in diagnosis beyond a day as a very important determinant of outcome. The authors of this report found 52 patients, all of whose duodenal injury was diagnosed within 24 hours–an excellent start. Thirty- two had blunt injury (almost all MVC), 14 had gunshot wound injury, and 6 had stab wounds. Mean ISS was 19, there was a mean of 2 associated injuries per patient, and overall mortality was 15%, with no difference in mortality based on mechanism of injury. All patients with penetrating injury had operative repair. By far, most patients with blunt injury had duodenal hematoma, almost all of which were managed nonoperatively. Of the 52, 8 died. Four patients (2 blunt, 2 penetrating) died of associated injuries before any repair was attempted. Two patients with blunt- mechanism duodenal hematoma managed nonoperatively died of other injuries, and the last 2 who died, with penetrating injury, had several associated injuries including inferior vena cava laceration. Univariate and multivariate analysis of this statistically-speaking small number of patients who, on the whole, did very well, identified only age, initial condition (hypotension, systemic pH, base deficit) and associated injuries as predictors of bad outcome.

Lopez PP, Benjamin R, Cockburn M, et al. Recent trends in the management of combined pancreatoduodenal injuries. Am Surg 2005; 71:847-852.

This next article concerns, thankfully, an even more uncommon and more seriously injured group of patients: those with both duodenal and pancreatic trauma. The surgeons at the Miami, Fla, Ryder Trauma Center reviewed the charts of all 33 patients with combined pancreatoduodenal injuries (of 240 patients with either duodenal or pancreatic injury–that is a lot of patients!) they treated over 4 years. As expected from Miami, and as opposed to the UC-Irvine population, 27 of the these 33 had penetrating injury, 24 being gunshot wounds. The mean ISS was a very high 22: 45% had an associated major vascular injury, more than one third presented in extremis, and yet only 6 died–another very impressive record. The authors’ principal findings were that abbreviated ”Damage control” laparotomy most likely contributed to their good results, and that the main cause of death in patients with combined pancreatoduodenal injury is injury to adjacent vital structures, mainly blood vessels. The authors’ discussion of the evolution and current variety of operative approaches to the injured duodenum and pancreas is worthwhile reading for anyone interested in the subject.

Ahmed N, Whelan J, Brownlee J, et al. The contribution of laparoscopy in evaluation of penetrating abdominal wounds. J Am Coll Surg 2005; 201:213-216.

The management of Abdominal stab wounds continues to evolve. Patients presenting with shock, peritonitis, or organ evisceration are all immediately explored. Difficulty arises in patients who have none of the aforementioned signs but clearly have an injury that might have penetrated the peritoneum, especially when the patient’s mental status is abnormal, and thus, the abdominal physical examination is unreliable. It has been shown many times that if all patients with abdominal stab wounds are explored, one third to more than one half will have no significant injury found. On the other hand, delay in diagnosis of a Perforated hollow viscus can lead to increased morbidity and mortality. The authors of this study used selective laparoscopy to decrease the incidence of nontherapeutic laparotomy engendered by a policy of mandatory exploration. All patients with abdominal stab wounds but without indications for immediate exploration underwent Diagnostic laparoscopy with low-pressure pneumoperitoneum. If no peritoneal or diaphragmatic wounds were detected and there was no abnormal fluid– bloody or clear–in the abdomen, then internal injuries were deemed absent and laparoscopy terminated. If defects of the peritoneum or diaphragm were found but there were no obvious signs of visceral injury, systematic laparoscopic exploration was performed. Nonbleeding liver, spleen, omental, or mesenteric hematoma without bowel involvement were not followed by laparotomy. Discovery of visceral injuries (holes in bowel, actively bleeding vessels) or secondary signs of such injury (foreign material, enteric fluid, bleeding without an obvious source) was followed promptly by open laparotomy. Fifty-two patients fulfilled the selection criteria and were included in the study. Seventeen patients were discovered to have no violation of the peritoneum. Fifteen patients had peritoneal penetration but had negative laparoscopic exploration. Of the 20 patients who were found to have visceral injuries, 12 required open operative repair. Patients with minor bleeding from stab wounds to the liver required no laparotomy. There was no mortality. The authors conclude that laparoscopy for the evaluation of abdominal stab wounds can be performed safely and can eliminate many nontherapeutic laparotomies, while at the same time ensuring no Significant injuries are missed. Laparoscopy is definitely the way to go for trauma centers that are not inundated with patients with abdominal stab wounds. Whether the surgeon feels comfortable with being able to rule out a significant injury via laparoscopic exploration alone depends on the surgeon’s laparoscopic experience. But at a minimum, if laparoscopy proves there has been no peritoneal penetration, unnecessary laparotomy will be avoided.

Arikan S, Kocakusak A, Yucel A, et al. A prospective comparison of the selected observation and routine exploration methods of penetrating abdominal stab wounds with organ or omentum evisceration. J Trauma 2005;58:526-532.

This study from Istanbul, Turkey, investigated and compared routine laparotomy vs selective observation management in patients with abdom- inal stab wounds with evisceration of the omentum or bowel. In the current US practice standard, all such patients are operated on. Sixty-one patients with abdominal stab wounds accompanied by evisceration of omentum or other organ were seen over a 6-year period. Nine had other indications (shock, peritonitis, evisceration of bowel with visible bowel injury, etc) for immediate operation and were operated on. Of the remaining 52 patients, 21 were randomized to ”mandatory laparotomy.” One third of these had unnecessary nontherapeutic laparotomy, and one fifth of the 21 developed postoperative complications. Thirty-one were randomized to selective observation. Twenty-five of these had no tenderness on initial exam, 6 of these developed tenderness and underwent laparotomy, and in only 1 of the

6 was the laparotomy nontherapeutic. Of the 6 with suspicious initial physical exam, only 1 worsened and was operated on; he had a nontherapeutic laparotomy; the remainder improved and were discharged home without an operation. The hospital stay was, of course, shorter for the nonoperated patients.

The authors conclude that selective observation of these patients resulted in a lower unnecessary laparotomy rate as well as a lower complication rate and shorter hospital stay. They concluded that selective observation is superior to routine laparotomy and call for a larger study to further investigate this conclusion. Perhaps laparoscopy can be effectively added into the treatment algorithm for these patients as well.

Alzamel HA, Cohn SM. When is it safe to discharge asymptomatic patients with abdominal stab wounds? J Trauma 2005;58:523-525.

If we decide to observe patients with abdominal stab wounds, how long do we have to observe them? The authors of this article reviewed a prospectively collected trauma registry database and reviewed the medical records of 650 hemodynamically stable patients who presented over a 7-year period. Sixty-eight patients (11%) required surgical intervention. One third underwent surgery within the second hour of presentation, one third had surgical intervention between the second and fourth hours, and only 13% had abdominal intervention between the fourth and sixth hours. Only 1 patient showed signs of hemodynamic instability from hemoperitoneum from a stab wound 12 hours after admission, and no patients required abdominal surgery after 12 hours of close observation. The authors conclude that after 12 hours of Close monitoring and serial physical examinations, if no signs of hemoper- itoneum and instability arise, the patient can be discharged from the hospital. I disagree to this conclusion, having personally cared for several patients, and having read about many more, with Hollow viscus injuries whose peritonitis and signs of inflammation/infection did not appear for more than 24 hours. If you discharge people before this time, you better be sure they are reliable and will return to your ED immediately if they develop worsening symptoms.

Dissanaike S, Griswold JA, Frezza EE. Treatment of isolated penetrating flank trauma. Am Surg 2005;71:493-496.

The authors of this article were interested in formulating a new algorithm for evaluation of patients with stab wounds of the flank. They reviewed the records of the 23 patients they had seen over the previous 5 years and found that triple-contrast CT scan was the preferred method, except in those patients in whom the stab wound was above the bottommost rib, in which case, laparoscopy was required to rule out diaphragmatic injury. This is actually correct, but not news. The proper definition of flank is the area between the 12th rib and the Iliac crest and between the anterior and posterior axillary lines. The authors’ definition was between the fifth

intercostal space and the iliac crest; I would call injuries above the lowermost rib here thoracoabdominal stab wounds, not flank wounds, for the same reason the authors advocate laparoscopy: the diaphragm is at risk in the area above the 12th rib. So, whatever one’s definitions are, if the stab wound is in the ribs, one must evaluate the diaphragm. If the stab wound is below the ribs, and posterior to the anterior axillary line, then triple-contrast CT is the way to go.

Benson DA, Stockinger ZT, McSwain NE. Embolization of an acute renal arteriovenous fistula following a stab wound: case report and review of the literature. Am Surg 2005;71:62-65.

Here’s another advance on the part of the interventional radiologists. This is a case report of a stab wound to the flank (2 cm above the iliac crest, below the 12th rib) causing a renal artery and vein injury which was successfully managed by angioembolization, similar to angioembolization of hemorrhage from the liver and spleen. Important points to keep in mind: the patient was hemodynamically stable (”death begins in x-ray”); CT showed a ”massive left perinephric hematoma,” but the there was no extravasation of contrast from the urinary collecting system; there was a contrast blush in the renal parenchyma, suggesting arterial bleeding.

Brown CVR, Kasotakis G, Wilcox A, et al. Does pelvic hematoma on admission computed tomography predict active bleeding at angiography for pelvic fracture? Am Surg 2005;71:759-762.

This is an important article for all of us who deal with Pelvic fractures. The authors reviewed the records of all 37 of their patients, seen over a 3-year period, who had a pelvis CT and then went to angiography. They found that neither the size of the pelvic hematoma seen on CT nor the absence of an arterial blush on CT had any correlation with whether there was active bleeding found at angiography. Yes, large hematomas and contrast blush did most often–but not always–portend active bleeding, but even patients with no hematoma and no contrast blush often had active, embolizable, bleeding. So, what then are good indications for angiography? They are hemodynamic instability (this is one of the few instances in which angiography is the place to go with a hemodynamically unstable patient); need for transfusion; significant Fracture pattern, such as sacroiliac disruption, diastasis of the symphysis pubis greater than 2.5 cm, and bilateral superior and inferior Pubic rami fractures; and finally, a large pelvic hematoma and/or contrast blush seen on pelvic CT.

Ziran BH, Chamberlin E, Shuler FD, et al. Delays and difficulties in the diagnosis of lower urologic injuries in the context of pelvic fractures. J Trauma 2005;58:533-537.

The association of lower urinary tract (bladder and urethra) injury with fracture of the bony pelvis ranges from 7% to 25%. Conversely, 80% of bladder ruptures are associated with pelvic fracture. Bladder ruptures may be intraperitoneal or extraperitoneal. Complications of missed bladder rupture include entrapment of the muscular bladder wall within the pelvic fracture, persistent urinary leakage, Pelvic abscess, peritonitis, respiratory difficulties, and sepsis from contaminated urine. Urethral injuries occur, by

on initial evaluation, including 4 intraperitoneal and 3 extraperitoneal bladder ruptures and 3 urethral tears. All 4 of the patients with a missed intraperitoneal bladder rupture had gross blood in the Foley catheter; cystography was falsely negative in 1 patient, 1 cystogram was read incorrectly as an extraperitoneal rupture, and the remaining 2 cystograms were read as inconclusive. Extraperitoneal ruptures were not diagnosed initially in 3 patients; all of these patients had gross blood with Foley catheter insertion also. In these 3 patients, cystogram was misread as an intraperitoneal rupture in 1, inconclusive in 1, and falsely negative in 1; the last patient developed Urinary retention 17 days after injury and cystoscopy revealed an extraperitoneal rupture. Urethral injuries were missed in 3 patients, all of whom had 4 pubic rami fractures in addition to a sacral ala fracture. Urethral tears in 2 patients were diagnosed at the time of urgent Exploratory laparotomy; neither of these 2 patients had blood at the urethral meatus or abnormal prostate, but there was gross hematuria upon Foley catheter placement. One patient with a urethral tear had a straddle fracture associated with a sacral ala fracture; this patient similarly had no blood at the meatus or abnormal prostate, but there was gross blood at the placement of the Foley. The authors concluded that in patients with multiple injuries, signs of urologic injury may not be obvious. Gross blood with Foley catheter placement is present in 95% of the patients with bladder rupture. Uroradiographic studies are indicated in any patient with gross hematuria and must be carefully performed and carefully read. The authors emphasize the importance of performing a cystogram which fully distends the bladder and then performing a postevacuation pelvic film to demonstrate any obscured extravasation. As to urethral injuries, approximately half of the patients will not manifest any of the classic signs of blood at the meatus, scrotal or perineal hematoma, or high-riding or boggy prostate. The authors emphasize the importance of evaluating the pattern of pelvic fracture, which might suggest urethral disruption: patients with Malgaigne’s fractures have a 3 1/2-fold increase in urethral injuries, and patients with straddle injury associated with ipsilateral sacroiliac involvement have a 24-fold increase.

1. Extremities

Brown CVR, Kasotakis G, Wilcox A, et al. Does pelvic hematoma on admission computed tomography predict active bleeding at angiography for pelvic fracture? Am Surg 2005;71:759-762.

This is an important article for all of us who deal with pelvic fractures. The authors reviewed the records of all 37 of their patients, seen over a 3-year period, who had a pelvis CT and then went to angiography. They found that neither the size of the pelvic hematoma seen on CT nor the absence of an arterial blush on CT had any correlation with whether there was active bleeding found at angiography. Yes, large hematomas and contrast blush did most often–but not always–portend active bleeding, but even patients with no hematoma and no contrast blush often had active, embolizable, bleeding. So, what then are good indications for angiography? They are hemodynamic instability (this is one of the few cases in which angiography is the place to go with a hemodynamically unstable patient); need for transfusion; significant fracture pattern, such as sacroiliac disruption, diastasis of the symphysis pubis greater than 2.5 cm, and bilateral superior and inferior pubic rami fractures; and finally, a large pelvic hematoma and/or contrast blush seen on pelvic CT.

far, primarily in males, and most often are located at bulbomembranous

junction. Complications of unrecognized urethral injuries include inconti- nence, impotence, and stricture formation. This study was initiated to identify missed lower Urinary tract injury in patients with pelvic fractures and the clinical implications of such injuries. The records of 635 patients with pelvic fractures were reviewed. Forty-three of these had lower urinary tract injuries. All patients with urologic injury had some type of significant anterior pelvic ring injury. The mechanism of injury included MVCs, industrial accidents, falls, and pedestrian injury. Fifteen intraperitoneal and 14 extraperitoneal bladder ruptures were identified. Thirteen patients had complete urethral tears, and 1 patient had both a urethral tear and extraperitoneal bladder rupture. In 10 patients, urologic injury was missed

Shapiro M, McDonald AA, Knight B, et al. The role of repeat angiography in the management of pelvic fractures. J Trauma 2005;58:227-232.

Hemorrhage from a pelvic fracture can be a difficult, potentially lethal, problem. Most often, approximately 80% of the time, bleeding from pelvic fractures is from small veins and venous plexus(es), and only 1 in 5 patients will actually have an arterial source. In the 1970s, before the advent of interventional radiology and angiographic embolization, trauma surgeons tried operative ligation of both internal iliac arteries in an attempt to stop ongoing hemorrhage from pelvic fractures, with quite poor results. The reasons for failure were many but include both the fact that most of the

bleeding is indeed venous, as well as the rich collateral blood flow to the pelvic bones and ligaments from arteries other than the internal iliacs. Angiographic embolization, on the other hand, coupled with restoration of the pelvic bones to normal apposition with some method of external compression and fixation, has been very effective in staunching pelvic bleeding. This article asks the question, what do you do if a patient with a bad pelvic fracture continues to bleed even after angiography? The authors’ retrospective review found 678 patients with pelvic fracture in a 2 1/2-year period, 31 of whom went to angiography for bleeding. Arterial hemorrhage was diagnosed and embolized initially in 16 patients; 3 of these required repeat angiography and embolization due to ongoing pelvic hemorrhage. Fifteen patients initially had negative arteriography; 5 of these were returned to angiography for ongoing bleeding, and 4 of the 5 had had arterial bleeders that were successfully embolized. The authors quite rightly emphasize that other sources of ongoing bleeding must be ruled out. But if they are, and if after angiography the patient continues to bleed, as manifest by continued or recurrent hypotension and persistent base deficit greater than 10 for more than 6 hours, return to the angiography suite is the right thing to do.

Garcia-Covarrubias L, McSwain NE, vanMeter K, et al. Adjuvant Hyperbaric oxygen therapy in the management of Crush injury and traumatic ischemia: an Evidence-based approach. Am Surg 2005;71: 144-151.

Emergency medicine physicians use Hyperbaric oxygen therapy perhaps more than any other group, dealing with carbon monoxide poisoning and others. This is a literature search article, which found

9 studies published between 1966 and 2004, describing a total of 150 patients who underwent HBO therapy for Extremity crush injuries and/or compartment syndromes. The articles are all described in some detail and are quite varied as to the type of injuries studied and the other treatments used. Most of the articles did not, in fact, deal with crush injuries but, rather, with vascular injuries, both blunt and penetrating, and reperfusion issues. Only 1 study was a prospective, RCT of HBO for crush injuries, but it had serious design flaws. The only thing the authors were able to conclude was that HBO is ”not likely to be harmful and could be beneficial if administered early.” You don’t need to rush out and buy an HBO chamber for now, but if you happen to already have one and are interested, ”well designed clinical studies are warranted.”

1. Burns

Burke JF. Burn treatment evolution in the twentieth century. J Am Coll Surg 2005;200:152-153.

This is a very interesting historical account written by one of the true pioneers in the management of burns: among other things, he is the coinventor of artificial skin. Dr Burke describes how the early treatment of burn victims was limited to pain relief and prevention or treatment of complications. The mortality from a burn was very high; if more than one third of the body surface area was involved, death was almost a certainty. Modern scientific burn care started during World War II, when research began, directed toward the understanding of burn shock and burn wound

sepsis. The evolution of burn care in the twentieth century has focused on early excision and grafting of the burned skin and the restoration of normal skin physiology. Appropriate topical and systemic antibiotics and skin substitutes have improved our ability to provide early burned skin coverage. The future will hold advancement in the care of the elderly burn patients and improvement in the care of inhalation injuries.

Cone JB. What’s new in general surgery: burns and metabolism. J Am Coll Surg 2005;200:607-615.

This article is an excellent concise review of all significant advances in Burn injury and burn care over the last 12 months. Subjects include initial ED and Burn unit evaluation, resuscitation, and stabilization; aggressive airway management with liberal Flexible bronchoscopy and instillation of various agents to try to reduce airway blockage by mucus and sloughed epithelium; Ventilator management; new skin substitutes; modulation of the hypercatabolism of burns; and chemical burns, with a focus on hydrofluoric acid.

1. Pediatric

Nadler EP, Potoka DA, Schultz BL, et al. The high morbidity associated with handlebar injuries in children. J Trauma 2005;58:1171-1174.

The authors of this article reviewed their experience with bicycle injuries in an effort to determine the frequency, magnitude, and outcomes of handlebar-impact injuries. A primary finding was that most often, during an accident, as the child loses control of the bike, the front wheel rotates to lie in a plane perpendicular to the child’s body, and the abdomen is then struck by the end of the handlebar. This mechanism is underappreciated and the injuries that result are often initially overlooked. The authors found that intra- abdominal injuries occurred in 31% of the children who had had direct impact handlebar injuries, with injuries to the bowel predominating. Retroperitoneal viscera, including the pancreas and duodenum, were also significantly frequent casualties. This information is especially important in light of the fact that injuries to the small bowel and pancreas can both be difficult to detect with CT scan and may manifest clinically in a delayed manner.

Mooney DP, Downard C, Johnson S, et al. Physiology after pediatric splenic injury. J Trauma 2005;58:108-111.

Current standard of care in the management of pediatric patients with a ruptured spleen is nonoperative. The Committee on Trauma of the American Pediatric Surgical Association has developed guidelines for the nonoperative management of hemodynamically stable children with isolated splenic trauma based on the CT grade of splenic injury. An underlying assumption is that hemodynamic stability correlates with CT grade of injury. This study was conducted to test that assumption, reviewing the records of all children seen at a single level I trauma facility from 1993 through 1999. A single experienced pediatric radiologist blinded to patient outcome independently reviewed all CTs. The authors found that the CT grade of splenic injury did indeed correlate directly with the hemodynamic parameters and inversely with hematocrit.