Preventable deaths from hemorrhage at a level I Canadian trauma center

BACKGROUND: Studies of trauma deaths have had a tremendous impact on the quality of contemporary trauma care. We studied causes of trauma death at a Level I Canadian trauma center, and tabulated preventable deaths from hemorrhage using explicit criteria. METHODS: Trauma registry data were used to identify all trauma deaths at our institution from January 1, 1999 to December 31, 2003. Demographics, mechanism, and time or location of death were recorded. Registry data analysis and selective chart or autopsy review were then performed to assign causes of death. RESULTS: A total of 558 consecutive trauma deaths were reviewed. Mean age was 48.7 (46.7-50.6), and mean Injury Severity Score was 38.8 (37.6-40.0); 29% were females. Blunt trauma represented 87% of all cases; penetrating injuries were only 13%. Central nervous system (CNS) injuries were the most frequent cause of death (60%), followed by hemorrhage (15%), and then combination CNS and hemorrhagic injuries (11%). Multiple organ failure caused 5% of deaths and 9% of deaths were from other causes. Of hemorrhagic deaths, 48% (n = 41) were from blunt injury, and 52% (n = 45) were from a penetrating mechanism. Of these hemorrhagic deaths, 16% were judged to be preventable because of significant delays in identifying the major source of hemorrhage. Hemorrhage from blunt pelvic injury was the major cause of exsanguination in 12 of 14 of these preventable deaths. CONCLUSIONS: Blunt injury is the major mechanism leading to trauma deaths. Massive bleeding from blunt pelvic injury is the major cause of preventable hemorrhagic deaths in our study.     

Trauma deaths in an Italian urban area: an audit of pre-hospital and in-hospital trauma care

In Italy, a comprehensive regional study of trauma deaths has never been performed. We examined the organization and delivery of trauma care in the city area of Milan, using panel review of trauma deaths. Two panels evaluated the appropriateness of care of all trauma victims occurred during 1 year, applying predefined criteria and judging deaths as not preventable (NP), possible preventable (PP), and definitely preventable (DP). Two hundred and fifty-five deaths were reviewed. Blunt trauma were 78.04% and motor vehicle crashes accounted for over 50%. Most victims (73.72%) died during pre-hospital settings and 91.1% died within the first 6h, principally because of central nervous system injuries in blunt and hemorrhage in penetrating trauma. Panels judged 57% of deaths NP, 32% PP, 11% DP (inter-panel K-test 0.88). Preventable deaths were higher after in-hospital admission. Main failures of treatment were lack in airway control or intravenous infusions in pre-hospital and mismanagement with missed injuries in emergency department.The high rate of avoidable deaths in Milan supports the need of trained pre-hospital personnel and of well equipped referring hospitals for trauma.     

Thoracotomy for blunt trauma: traditional indications may not apply

The indications for performing as urgent thoractomy after trauma are based on the criteria used for penetrating injuries. However, few data are available on the use of these indications for patients with blunt injuries. In a retrospective study (June 1996 to July 2001), we compared the indications of urgent thoracotomy after blunt injury and penetrating injury in patients who underwent thoracotomy within 24 hours of hospital admission at our institution. Patients with blunt aortic injuries or emergency department thoracotomies were excluded from evaluation. Fifty-nine patients were identified (37 penetrating injuries, 22 blunt injuries). Blunt trauma victims had a higher mortality rate than penetrating trauma victims (73% vs. 22%). Chest tube output was the indication for nontherapuetic thoracotomy in 5 patients with blunt injuries whereas this occurred in only 1 penetrating injury victim (P = 0.04). All 5 blunt injury patients underwent a prior procedure and were coagulopathic when thoracotomy was performed. In conclusion, thoracotomy following blunt trauma is associated with a high rate of mortality. The rate of nontherapeutic exploration is increased when chest tube output is the indication for thoracotomy after blunt trauma. Since the majority of such patients have multicavitary injuries that require prior operation and are commonly coagulopathic, caution should be exercised when deciding whether to proceed with thoracotomy based solely on chest tube output.     

Diaphragmatic trauma in southern Saskatchewan--an 11-year review

An 11-year retrospective review of 51 patients with diaphragmatic injuries revealed 33 blunt and 18 penetrating injuries. In the blunt trauma group 24 were left-sided and nine were on the right side. Preoperative diagnosis was made in 24 patients. Delayed diagnosis (greater than 24 hours) during life occurred in four patients and two injuries were found at autopsy. Chest X-ray was diagnostic or highly suggestive in 23 patients. All patients in this group had associated extra-abdominal injuries; 23 patients had concomitant intraabdominal injuries. Transabdominal repair was performed in 24 patients. Four deaths occurred in the blunt injury group. The penetrating diaphragmatic wounds consisted of 14 left-sided, three right-sided, and one pericardial wound. Preoperative diagnosis occurred in only three patients. The penetrating wound, hypotension, or peritoneal signs dictated operation in the remaining 15 patients. One death occurred from peritonitis and septic shock. Blunt and penetrating diaphragmatic injuries remain a diagnostic challenge and associated injuries determine the outcome.     

Acute tracheobronchial injury

We reviewed our experience with tracheal and bronchial trauma from 1977 to 1988. There were 22 patients with tracheobronchial injuries treated in this period. Seventeen (77%) of the injuries were due to penetrating trauma and five (23%) were due to blunt trauma. Thirteen patients had major associated injuries, including six esophageal injuries. The most common physical findings were tachypnea (13 patients) and subcutaneous emphysema (nine patients). Eight patients presented with airway obstruction. All patients with penetrating cervical tracheal injuries underwent neck exploration and primary repair. All blunt injuries were diagnosed by bronchoscopy. Three patients with blunt injuries were treated with primary repair. Two patients with blunt chest trauma and small bronchial tears were treated nonoperatively with good results. All three deaths (14% mortality rate) were due to associated injuries. We conclude that patients with penetrating tracheobronchial injuries should be managed by surgical exploration and primary repair, although selected patients with blunt injury may be treated nonoperatively.     

Seven hundred fifty-three consecutive deaths in a level I trauma center: the argument for injury prevention

BACKGROUND: The past century has seen improvement in trauma care, with a resulting decrease in therapeutically preventable deaths. We hypothesize that further major reduction in injury mortality will be obtained through injury prevention, rather than improvements in therapy. METHODS: Seven hundred fifty-three deaths in an American College of Surgeons-verified, Level I trauma center were reviewed as they occurred. Deaths were classified as therapeutically not preventable, possibly preventable, or preventable. These charts were also reviewed for factors that might have prevented or lessened the severity of the injury. RESULTS: Mean age was 43, mean Glasgow Coma Scale score was 5, mean Revised Trauma Score was 4, mean Injury Severity Score was 41, and mean probability of survival was 0.25 (according to TRISS). Forty-six percent underwent cardiopulmonary resuscitation in the field, 52% died within 12 hours, 74% died within 48 hours, and 86% died within 7 days. Primary causes of death included central nervous system injury in 51%, irreversible shock in 21%, multiple injuries (shock plus central nervous system injury) in 9%, multiple organ failure/sepsis and other causes in 3%, and pulmonary embolus in 0.1%. Seven hundred one (93%) were classified as not preventable with a change in therapy, 32 (4.2%) were classified as potentially preventable with a change in therapy, and 20 were classified as preventable with a change in therapy (2.6%). Forty-six percent had cardiopulmonary resuscitation performed before or immediately on arrival to the hospital. Another 23% had vital signs present on arrival, but had a Glasgow Coma Scale score of 相似文献   

Preventability of death from penetrating injury in England and Wales

A total of 1000 deaths from injury in England and Wales have been reviewed to establish the incidence and pattern of penetrating injury and the adequacy of its management. Of the 1000 deaths, 71 (7.1 per cent) were due to penetrating injury. There were 32 knife wounds and 30 firearm injuries. Most of the latter were suicides. Only 17 patients (24 per cent) reached hospital alive. Of these cases, 10 had extracranial injury and all 10 deaths were considered to have been potentially preventable when reviewed by four external assessors. One of seven patients with cranial injury was considered to have been a potentially preventable death. The median age of the 11 cases of potentially preventable death was 37 years (range: 7-61 years). Of these, three did not have any surgery for surgically treatable injuries. Seven patients underwent operation and difficulty was encountered in six of these. It appears from our figures that whilst penetrating injury is an uncommon cause of death, it is poorly managed. The implications of this finding for systems of injury care in the United Kingdom are discussed.     

Management of blunt and penetrating colon injuries.

The records of 28 patients with traumatic colon injuries (TCI) were retrospectively reviewed. Sixteen patients (57%) with 17 TCI had blunt trauma, whereas 8 patients (29%) experienced penetrating trauma. Four TCI were from intraluminal injury. Blunt trauma commonly involved the left colon, whereas penetrating trauma usually involved the right or transverse colon. Fifty-nine percent of the blunt TCI were treated with primary repair, including resection and primary anastomosis, as were 88% of the penetrating TCI. Shock, transfusion requirement of more than 4 units, contamination, and associated injuries did not necessarily preclude primary repair. One of 16 patients (6%) who underwent primary repair developed morbidity related to the colon injury. The morbidity rate for the colostomy group was 13% (1 of 8). The mortality rate was 13% in the patients who experienced external trauma (3 of 24). Two of these deaths were related to severe head injury and chest injury, respectively. These data represent a much higher proportion of blunt injuries to the colon than is reported in the literature. The low rate of morbidity for all patients treated by primary repair tends to support the more liberal trend toward this technique for both blunt and penetrating TCI.     

A review of traumatic airway injuries: potential implications for airway assessment and management

BACKGROUND: Obtaining a patent airway can be difficult in patients with traumatic airway injuries (TAI). There is a paucity of data available about the incidence of airway compromise and techniques used in these patients. METHODS: Charts review of all patients with TAI treated in a Regional Trauma Center from July 1989 to June 2005. RESULTS: One hundred and four patients were identified as TAI in the study period (incidence of 0.4% for blunt and 4.5% for penetrating trauma). Sixty-eighty patients were victims of penetrating trauma (ISS: 24+/-10; mortality: 16%). Thirty-six patients were blunt trauma victims (ISS: 33+/-16; mortality: 36%). Overall, 65% of the patients received a definitive airway (DA) in the pre-hospital setting or at the initial hospital assessment. Alternative techniques for obtaining DA including wound tracheal tube, surgical airway and intubation under fiberoptic bronchoscopy were used in 30% of the patients. Among 24 deaths, 10 were considered primarily due to the airway injury. Twelve patients presented with thoracic TAI with nine deaths in this subgroup. CONCLUSIONS: Overall, the incidence of TAI is low. Blunt trauma TAI is less common, and these patients have a different clinical presentation, higher ISS and mortality than the penetrating TAI group. Early assessment of airways is crucial and DA was required in 2/3 of the patients with TAI. Lower airway injuries have higher mortality than upper airway injuries. Even though most patients died as a result of other injuries, causative factors of death included difficulty in obtaining DA and ventilation/oxygenation problems.     

Correlation of abdominal trauma index and injury severity score with abdominal septic complications in penetrating and blunt trauma.

The Abdominal Trauma Index (ATI) was designed to stratify patients with penetrating injuries, and has been used to classify patients with blunt trauma. The Injury Severity Score (ISS) was originally designed to stratify victims of blunt trauma, and it has also been used for victims of penetrating trauma. We attempted to validate the use of ISS and ATI for both penetrating and blunt trauma. A total of 592 penetrating and 334 blunt trauma patients who underwent laparotomy over a 5-year period were evaluated. The overall rate of abdominal sepsis was 7.5% for blunt trauma and 7.6% for penetrating trauma. Mortality (excluding deaths within 48 hours) was 7% for blunt trauma and 1% for penetrating trauma. In the penetrating injury population, an ATI value greater than 15 and an ATI value greater than 25 were significantly associated with abdominal septic complications (ASCs) (p less than 0.001, both comparisons). An ISS greater than or equal to 16 was also associated with ASCs (p less than 0.001). The ASC rate for gunshots was higher than that for stab wounds (11% vs. 2%; p less than 0.001). In the blunt group, an ATI value greater than 15 and an ATI value greater than 25 were associated with ASCs (p less than 0.01 and p less than 0.001, respectively). The association of ASCs and ISS was linear with increasing ISS in patients with blunt abdominal trauma.(ABSTRACT TRUNCATED AT 250 WORDS)     

Analysis of deaths due to injury in Milano: preliminary results

We reviewed all trauma deaths occurring in the urban area of Milan during one year. Autopsy reports were cross-referenced with pre- and in-hospital records and the Injury Severity Score was calculated by a senior surgeon. Causes of deaths were defined as central nervous system injury (CNS), hemorrhage (HEM), combined central nervous system injury and hemorrhage (CNS + HEM), and burns (BURN). Places of death were considered the scene (DOS), during transportation (DOA), the emergency room (DER), and hospital. Two multidisciplinary commissions reviewed patient reports and deaths were judged non-preventable, possibly preventable or frankly preventable, using the unanimous decision rule. The TRISS method was used to calculate the probability of survival for in-hospital deaths. Overall trauma deaths were 255 with 78.04% blunt and 16.08% penetrating traumas. Burns accounted for 5.88%. CNS and CNS + HEM caused 171 (67.05%) deaths. DOS were 91, DOA 48, DER 34, and in-hospital deaths 33. Victims found dead (49 individuals) were excluded from further analysis. The commissions classified 56.31% of deaths as non-preventable, 32.03% as possibly preventable and 11.65% as frankly preventable. The Injury Severity Score decreased from DOS to in-hospital deaths (p < 0.05). The preventability rate was higher for in-hospital deaths (p < 0.05). The results of this study suggest that the development of a tiered trauma system in Milan is mandatory.     

Preventable deaths in a self-designated trauma system

Organized paramedic care was established in 1974 in Hillsborough County, Florida, with subsequent development of a hospital self-designation system for trauma in 1980. To evaluate the level of trauma care in the county, a review of trauma deaths in 1984 was performed. A total of 452 trauma deaths was identified. Of these, 191 deaths occurred at the scene. The remaining 261 patients were transported to one of six hospitals within the county. One hundred ninety-nine subsequent deaths were attributed to central nervous system (CNS) injury, while 62 deaths were secondary to non-CNS injuries. By the method of group review, 14 (22.6%) preventable non-CNS trauma deaths were identified. Six women died and eight men died; the mean age of the deceased was 44. Ten deaths (71.4%) were secondary to blunt trauma. Mean ISS score was 21.1. Eleven deaths (79%) were due to delay to the OR, 2 deaths (14%) were due to inadequate resuscitation, and 1 death (7%) was due to lack of surgical intervention. This study demonstrates that a self-designation system without regulatory control results in a high percentage of preventable trauma deaths. We conclude that established trauma systems are needed in all areas, including those that have had organized prehospital and hospital levels of care.     

Trauma fatalities: time and location of hospital deaths

BACKGROUND: Analysis of the epidemiology, temporal distribution, and place of traumatic hospital deaths can be a useful tool in identifying areas for research, education, and allocation of resources. STUDY DESIGN: Trauma registry-based study of all traumatic hospital deaths at a Level I urban trauma center during the period 1993 to 2002. The time and hospital location where deaths occurred were analyzed according to mechanism of injury, age, Glasgow Coma Score, and body areas with severe injury (Abbreviated Injury Scale [AIS] >/= 4). Logistic regression analysis was used to identify risk factors associated with death at various times after admission. RESULTS: During the study period there were 2,648 hospital trauma deaths. The most common body area with critical injuries (AIS >/= 4) was the head (43%), followed by the chest (28%) and the abdomen (19%). Overall, 37% of victims had no vital signs present on admission. Chest AIS >/= 4, penetrating trauma, and age greater than 60 years were significant risk factors associated with no vital signs on admission. Patients with severe chest trauma (AIS >/= 4) reaching the hospital alive were significantly more likely to die within the first 60 minutes than were patients with severe abdominal or head injuries (17% versus 11% versus 7%). In patients reaching the hospital alive, the time and place of death varied according to mechanism of injury and injured body area. Deaths caused by severe head trauma peaked at 6 to 24 hours, and deaths caused by severe chest or abdominal trauma peaked at 1 to 6 hours after admission. CONCLUSIONS: The temporal distribution and location of trauma deaths are influenced by the mechanism of injury, age, and the injured body area. These findings may help in focusing research, education, and resource allocation in a more targeted manner to reduce trauma deaths.     

The value of laparoscopy in management of abdominal trauma

The role of laparoscopy (LS) in abdominal trauma is controversial. Concerns remain regarding missed injuries and safety. Our objective for this study was to determine the safety and better define the role of LS in abdominal trauma victims. We performed a retrospective review of all patients who sustained abdominal trauma and underwent LS in a level I trauma center. The main outcome measures were age, gender, mechanism of injury (MOI), indication for laparoscopy, presence of intra-abdominal injury (IA), therapeutic laparoscopy (TxLS), need for laparotomy, length of hospital stay (LOS), missed injuries, complications, and deaths. Forty-eight patients underwent LS (62 per cent male; average age, 28 years; MOI, 35 (85%) penetrating, 7 (15%) blunt; mean ISS, 8). At laparoscopy, 58 per cent of patients had no intra-abdominal injury. IA injury was treated with laparotomy in 14 (29%) and TxLS in 6 (13%). One patient had a negative laparotomy (2%). No injuries were missed. No patients required reoperation. There was one complication: a pneumothorax. There were no deaths. LS was most valuable in penetrating trauma, avoiding laparotomy in more than two-thirds of patients with suspected intra-abdominal injury. LS can serve as a useful adjunct for the evaluation of blunt trauma. In a level I trauma center with LS readily available, the procedure is associated with a low rate of complications and missed injury.     

Chest injuries: a clinical and autopsy profile

The clinical profiles and management of 236 consecutive chest injury patients treated and followed up at All India Institute of Medical Sciences between January 1983 and July 1985 were analyzed prospectively. There were 149 blunt and 87 penetrating injuries; 21 patients (9%) required thoracotomy. Single- or multiple-tube thoracostomy was performed in 141 patients (60%). The remaining 74 patients (31%) required only observation for a period of 24-48 hours. Fifteen patients (6.3%) died, the mortality being related to head injury in four, irreversible hypovolemic shock in four, pulmonary embolism in three, septicemia in two, and respiratory failure in two. Nonfatal complications included residual hemothorax in 18 cases, persistent air leak in 13, pulmonary infection in eight, pulmonary embolism in one, and empyema in one. The average hospital stay was 6.9 days. Evidence of chest injury of various magnitudes was found in 756 of 2,286 autopsies conducted for trauma-deaths during the same study period analyzed retrospectively; however, it was the major cause of death in only 147 (19%). Cardiac injuries accounted for 41% of the deaths resulting primarily from chest trauma. Only 10% of the patients who sustained cardiac injury reached hospital alive.     

The significance of critical care errors in causing preventable death in trauma patients in a trauma system

The regionalization of trauma care has led to a decrease in preventable death after injury. This decrease has been attributed to earlier resuscitation and surgical intervention. Little emphasis, however, has focused on the critical care phase of trauma patient management. This study was undertaken to determine the significance of critical care errors (CCEs) on preventable mortality and morbidity in a regionalized system of trauma care. The records of 12,910 trauma patients admitted to six trauma centers over a three-year period were reviewed. The cause and preventability of death was determined by a panel of trauma experts. Critical care errors occurred in 151 (23%) of all patients with errors occurring during any phase of care. The CCEs were identified in 38 of 125 (30%) deaths of patients with errors in some phase of care. The CCEs were implicated as the cause of death in 30 (48%) of the 62 preventable deaths. The proportion of preventable deaths attributable to CCEs was higher than the proportion of preventable death attributable to errors in the resuscitative and operative phases of care (p less than 0.001, chi-square). These data indicate that CCEs significantly contribute to preventable mortality and morbidity in trauma patients. It is imperative that physicians caring for trauma patients possess expertise in the critical care management of injured patients.     

Preventable trauma death in Tehran: an estimate of trauma care quality in teaching hospitals

OBJECTIVE: The purpose of this study was estimate the number of preventable trauma deaths in teaching hospitals in Tehran. METHODS: We evaluated the complete prehospital, hospital, and postmortem data of 70 trauma patients who had died during a 1-year period in two of the largest university hospitals in Tehran with a multidisciplinary panel of experts. RESULTS: Panel members identified 26% of all trauma deaths as preventable deaths. From 31 non-central nervous system-related deaths, 17 and 6 cases were identified as surely preventable and probably preventable, respectively. In central nervous system-related deaths, 5% of the deaths overall (2 of 38 cases) were identified as surely preventable or probably preventable. Sixty-four cases of medical errors were identified in 31 trauma deaths and 80% of these errors were directly related to the death of the patients. CONCLUSION: The high preventable trauma death rate in our teaching hospitals indicates that a relatively significant percentage of trauma fatalities could have been prevented by improving prehospital and in-hospital trauma care.     

Diagnostic and therapeutic aspects of rectal trauma. Blunt versus penetrating

In the last 6 years, nine patients with blunt and 16 with penetrating rectal injuries were treated at University Hospital, Jacksonville, Florida. Blunt trauma was caused by vehicular accidents in seven patients and crush injuries in two. Penetrating rectal trauma was due to gunshot wounds in ten patients and foreign body insertion in six. All patients with blunt injury had bright red rectal bleeding, which led to diagnostic sigmoidoscopy. Rectal injury was identified at sigmoidoscopy in 12 patients who had penetrating wounds and at laparotomy in four patients. Thirteen patients who had penetrating rectal trauma had injury to only the rectum or to one additional organ. In contrast, all patients who had blunt rectal trauma had at least three associated injuries. In the penetrating group, 13 patients were treated by colostomy and mucus fistula; three patients with mucosal injury were managed nonoperatively. The only death occurred in a patient whose rectal injury was initially missed. Patients who had blunt rectal trauma were managed with colostomy and mucus fistula. Three patients died postoperatively, two of pelvic bleeding and one of head injury. Hemodynamic stabilization, colostomy and mucus fistula, presacral drainage, and rectal washout constitute proper treatment of patients with blunt or penetrating rectal trauma. Because of the greater number and severity of associated injuries, morbidity and mortality are higher after blunt rectal trauma.     

Management of penetrating and blunt diaphragmatic injury

The past 5 years' experience with diaphragmatic injuries at the University of Texas Health Science Center in San Antonio was reviewed to refine the clinical signs and appropriate treatment. During this period 102 patients were treated. Ninety-three patients incurred penetrating trauma to the diaphragm and nine patients suffered blunt trauma. Chest X-rays were normal in 40 patients, a hemo- and/or pneumothorax was present in 57, herniated abdominal viscera in four, and free air in one. Peritoneal lavage was positive in six of seven patients with blunt diaphragmatic injury, but was falsely negative in two of five patients (20%) with penetrating diaphragmatic injury. Eighty-nine patients (87%) experienced 137 associated injuries (excluding hemo- and/or pneumothorax). Nine patients (8.8%) had an isolated diaphragmatic injury. Four patients (4%) had a diaphragmatic injury associated with only a hemo- and/or pneumothorax. All patients, except for three with injuries recognized late, were operated upon immediately. Two patients had a missed diaphragmatic injury at initial laparotomy. There was one death in the series from a consumption coagulopathy. It was concluded that injuries to the diaphragm should be suspected in all patients with severe blunt torso trauma or with penetrating injuries near the diaphragm. Because of the nonspecificity of X-rays and the 20% false negative rate for peritoneal lavage, we believe that missed injuries and morbidity can be minimized by immediate laparotomy for all patients with abdominal and low thoracic penetrating injuries. Care must be taken not to overlook associated injuries.     

THE MAJOR TRAUMA MANAGEMENT STUDY: AN ANALYSIS OF THE EFFICACY OF CURRENT TRAUMA CARE

Background : An audit of the management and outcome of major trauma patients was carried out to determine ways in which the system of care may be improved. Methods : The Major Trauma Management Study (MTMS) collected data prospectively on all consecutively admitted major trauma patients at eight major hospitals during a 12-month period. Outcome was studied using trauma and injury severity score (TRISS) and a severity characterization of trauma (ASCOT) analyses, as well as a preventable outcome analysis, which looked at survivors with complications or with a Glasgow Coma Score < 15 on discharge from hospital, as well as studying deaths. Results : The group of 859 patients was more severely injured than most described previously, with a mortality of 14.8% and a mean injury severity score of 19.8. Formal ASCOT analysis indicated 2.25% fewer survivors than would be predicted by Major Trauma Outcome Study norms. Extrapolating the TRISS and ASCOT process to include those patients with missing data, and then comparing groups of matched severity with the norms, gave no statistically different outcome in the MTMS group of patients. Preventable outcome analysis revealed rates of preventable and potentially preventable (P/PP) outcomes of 32% among deaths and 8% among survivors. The types of management deficiencies responsible for P/PP outcomes are identified. Conclusions : The points of deficiency in a system of care have been identified, and the development of an integrated trauma system in Victoria, based upon these facts, is recommended. Children, the elderly, patients with head injuries and patients being transferred between hospitals would benefit from improvements to the system of care. The calculation of efficacy rate (0.95 for the MTMS patients) is recommended to accurately assess the system of care. Preventable Outcome Analysis is more relevant to auditing a system of trauma care in detail, than is ASCOT or TRISS. The MTMS has refined and defined the process so that it is reproducible in further comparative studies.