European Journal of Trauma and Emergency Surgery

Focus on Diagnostic and Prognosis of Severely Traumatized Patients

Traumatic Deaths in the Emergency Room: A Retrospective Analysis of 115 Consecutive Cases
Tim Sderlund1, Ilkka Tulikoura1, Mika Niemel2, Lauri Handolin1
Abstract Objective: The aim of the present study was to characterise traumatic deaths occurring in the emergency room (ER) and to assess retrospectively the quality of given emergency care by evaluating whether any of the deaths could be identified as potentially preventable. Methods: All consecutive deaths of trauma patients between 1 January 1998 and 31 December 2006 in the ER of the Tl Hospital Trauma Centre were retrospectively reviewed. The inclusion criterion was death of a trauma patient occurring in the ER. Both the pre- and inhospital medical charts and the autopsy reports of the patients were reviewed. Results: A total of 115 patients fulfilled the inclusion criteria, and the autopsy reports were obtained for all of these cases (100%). The patients were mainly males (n = 84; 73%), and the median age of the patients was 51 years (range 1-93 years). The average injury severity score (ISS) was 34.6. Blunt trauma was the most common type of injury in the study population. A total of 115 injuries in 50 patients were missed in both the clinical and radiological surveys in the ER, i.e., a missed injury was identified in 43% of the cases. Of these patients, 15.7% had a clinically significant missed injury (AIS 4). Based on our review of all available material, we consider that 11 deaths (9.6%) were potentially preventable. Conclusions: Missed injuries did not play a major role in the preventable deaths. Seven potentially preventable deaths were considered to be failures in the surgical decision-making process, resulting in futile non-operative treatment or a delay in surgical bleeding control.

Eur J Trauma Emerg Surg 2009;35:455-62 DOI 10.1007/s00068-009-8179-0

Key Words Preventability of death Trauma benchmarking Traumatic death Trauma patient outcome
1

Introduction Trauma is one of the leading causes of death, accounting approximately for every tenth death worldwide [1, 2]. It is also the leading cause of death in the population under 40 years of age in Western societies [3]. In Finland, there were 4,125 trauma-related deaths in 2003, and trauma was the third most common cause of death on a wholepopulation basis [4]. Approximately 50% of the trauma deaths occur in the early phase - within 24 h of the injury [5, 6]. In Canada, 82% of the trauma patients dying in the emergency room (ER) of a trauma centre had a clinically missed intra-abdominal injury that was observed in autopsy [6]. The incidence of autopsies carried out after traumatic death varies from 6 to 97% [6-8], which may lead to an underestimation of injuries and cause a bias in the estimation of the probability of survival [6, 9]. The Toolo Hospital Trauma Centre (Helsinki University Central Hospital) provides acute trauma care for patients from Helsinki and its surroundings, resulting in a catchment area of about 1.5 million people (approx. 25% of the Finnish population). Helsinki University Central Hospital has three individual units that provide emergency trauma care, with Toolo Hospital providing acute care for all major traumas, with the exception of patients < 16 years without a potential brain injury. Thus, the number of adolescent trauma patients at Toolo Hospital is low and consists only of those with a suspected brain injury. The patients with penetrating torso injuries are mostly admitted to another hospital, which results in a low number of these patients as well at Toolo Hospital. There are approximately 20,000 admissions to Toolo Hospital annually, including approximately 500 patients with an injury severity score (ISS) > 15.

Department of Orthopedics and Traumatology, Tl Hospital, Helsinki University Central Hospital, Helsinki, Finland, 2 Department of Neurosurgery, Tl Hospital, Helsinki University Central Hospital, Helsinki, Finland. Received: September 13, 2008; revision accepted: February 10, 2009; Published Online: May 12, 2009

Eur J Trauma Emerg Surg 2009 No. 5

URBAN & VOGEL

455

Sderlund T, et al. Traumatic Deaths in the Emergency Room

An ambulance doctor service is available in the city of Helsinki, and a helicopter-doctor service is available in the surrounding cities. The coverage of pre-hospital emergency care doctors is considered to be good in the Helsinki University Central Hospital catchment area, and only seldom is a major trauma patient taken to hospital without an accompanying pre-hospital doctor. The aim of the study reported here was to characterise the traumatic deaths occurring in the ER of Toolo Hospital and to assess retrospectively the quality of given emergency care by evaluating whether any of the deaths could be identified as potentially preventable in the light of the medical reports and post-mortem autopsy reports. Finnish law requires an autopsy to be performed on all victims dying in an accident, dying in obscure circumstances and dying or suspected to have died due to violence (both self-inflected and nonself-inflected). Here, we present a unique data set of 115 consecutive deaths with an autopsy rate of 100%.

Materials and Methods All consecutive deaths of trauma patients between 1 January 1998 and 31 December 2006 in the ER of the Toolo Hospital Trauma Centre were retrospectively reviewed. An inclusion criterion was the death of a trauma patient occurring in the ER. Non-traumatic deaths due to a medical condition (such as non-traumatic intracranial bleeding), patients declared dead on arrival (DOA, defined as no active movement, no palpable pulse, no breathing and no electrical cardiac activity) and deaths of patients > 60 years of age with a low-energy proximal femur fracture were excluded from the study. We reviewed the pre- and in-hospital medical charts and the autopsy reports of all patients. The collected data consisted of the mechanism of the injury, recorded vital signs at the scene and in the ER, clinically and radiologically diagnosed injuries, injuries diagnosed in autopsy, results of the laboratory tests, given care and timing in transportation and given care. The given care consisted of diagnostic and therapeutic manoeuvres, including fluid resuscitation and blood transfusions. Vital signs were defined as systolic blood pressure (SBP), heart rate (HR), respiratory rate (RR), Glasgow Coma Scale (GCS) [10] and peripheral oxygen saturation. The injuries were classified using the Abbreviated Injury Scale (AIS) [11] for obtaining ISS [12] and new injury severity score (NISS) [13]. The revised trauma score (RTS) [14] was assessed, and the trauma-injury

severity score (TRISS) methodology [15] was used to calculate the probability of survival, P s, at the time of admission to the ER. According to the AIS philosophy, the AIS scores were downgraded in cases without further specification of injuries. The AIS score was determined by a trained and certified author (LH). The hospital admission data were used for the calculating the RTS and Ps for non-intubated patients. For intubated patients, the pre-hospital data prior to intubation were used to replace the missing values of GCS and RR in order to minimize the bias caused by missing values among the most severely injured patients [16] i.e. those intubated at the scene. The pre- and in-hospital medical charts and the autopsy reports for each patient were formally peerreviewed by all authors. Based on the available data, each individual death was assessed as potentially preventable or non-preventable by two authors experienced in general trauma surgery and one experienced in neurotrauma and neurosurgery. All of the assessments were carried out by the authors working in consensus. Deaths were rated potentially preventable if a treatable life-threatening condition could be retrospectively identified based on the clinical examinations, radiological studies or autopsy report, but was not identified and/or treated. Also, if immediate or earlier procedures to control the bleeding could potentially have changed the outcome, the death was classified as potentially preventable. A potentially preventable death was also considered in cases where a treatable injury, such as tensionpneumothorax or cardiac tamponade, was missed and therefore left untreated, thereby contributing clearly to the death. All deaths were classified by the cause of the death as brain injury (CNS), exsanguination or other reason. A death due to CNS was defined as a brain injury and/ or a spinal cord injury incompatible with life regardless of other injuries. Exsanguination was determined as the cause of death if major bleeding was present, regardless of possible attempts to control it, and when CNS was not applicable. Causes of death other than CNS or exsanguination were defined as another critical injury or condition incompatible with life, such as compromised airway, burn/inhalation injury, air embolism or blunt cardiac injury.

Results A total of 115 patients fulfilled the inclusion criteria, and the autopsy reports were obtained for all of these patients (100%). The patients were mainly male (n = 84, 73%), and the median age of the patients was

456

Eur J Trauma Emerg Surg 2009 No. 5

URBAN & VOGEL

Sderlund T, et al. Traumatic Deaths in the Emergency Room

Table 1. Demographics of the patients. Mechanisms of the injury/clinical characteristics of the patients (n = 115) Dominant injury, n (%) Blunt Penetrating Burn Injury type, n (%) Blunt assault Burn Crushing Explosion GSW Major fall Minor fall Traffic Clinical assessment scores/parameters (average SD) ISS NISS RTS RR SBP GCS Hb (g/l) BE (mmol/l) TT (%)

94 (81.7) 17 (14.8) 4 (3.5) 1 (0.9) 4 (3.5) 1 (0.9) 1 (0.9) 16 (13.9) 21 (18.2) 20 (17.4) 51 (44.3) 35 17 42 15 3.75 1.79 12 7 97 49 4.6 3.4 98 32 -8.47 6.5 60 28

the scene and required extraction. The median time from the accident to the arrival of the pre-hospital unit to the scene was 16 min (IQ25-75 11-22 min), and the median scene time was 27 min (IQ 25-75 20-38 min). Ninety patients (78%) were intubated before arrival at the ER. Eleven blunt trauma patients (9.6%) received cardiopulmonary resuscitation (CPR) at the scene, four of them dying of exsanguination at hospital. Pre-hospital pleural decompression was performed on eight patients (7%), but none of these patients had a cardiac arrest necessitating CPR. Surgical airway was established on one patient at the scene. The median volume of intravenous fluids administered before arrival at the ER was 1,500 ml (IQ25-75 1,000-2,250 ml). Emergency Room Emergency interventions carried out in the ER included intubation (12 patients), chest tube insertion (28 patients), CPR (27 patients), thoraco-laparotomy (one patient), escarotomies (two patients) and external pelvic stabilization (two patients). Intra-cranial pressure monitoring was started in the ER for three patients. The patients were treated in the ER for a median of 98 min (IQ25-75 41-162 min) before death. Nine patients (7.8%) died within 15 min after arrival at the ER, and the majority of the patients (n = 103, 89.6%) died within 4 h after arrival. The median times from ER admission to FAST (focused assessment of sonography for trauma) was 10 min (IQ25-75 615 min), to supine chest X-ray, 15 min (IQ25-75 1020 min), and to computed tomography (CT) scan, 30 min (IQ25-75 20-40 min). The CT scans were performed on 75 patients, and injuries causing the death could be diagnosed from the CT scans in all cases. The median volume of total fluids the patients received in the ER was 3,000 ml (IQ25-75 1,0007,700 ml). Blood products were given to 53 patients in the ER. For these patients the average number of the units of red blood cells (RBC), fresh frozen plasma (FFP) and platelets were 8.6, 1.1, and 0.4 units, respectively. Patients in the exsanguination group received 11.8, 1.1 and 0.5 units of RBC, FFP and platelets, respectively. Trauma Scoring The trauma-scoring characteristics, vital signs and results of the laboratory tests are presented in Table 1. In the CNS group (n = 71), 41 patients were observed to have injuries in other body parts as well, with the most common being the thorax (n = 31) and the bony pelvis (n = 26). The major bleeding sites at the patients dying of exsanguination (n = 30) were the abdomen (n = 6),

Minor fall: < 4 m; major fall: 4 m; GSW: gunshot wound; ISS: injury severity score; NISS: new injury severity score; RTS: revised trauma score; RR: respiratory rate; SBP: systolic blood pressure; GCS: Glasgow Coma Scale; Hb, hemoglobin; BE: base excess; TT: thromboplastin time

51 years (range 1-93 years). Blunt trauma was the most common type of injury. The mechanisms of injuries are shown in Table 1. Eleven patients (9.6%) were referred to from other hospitals, and 104 patients were brought from the scene. Suicide was the reason for injury in 32 cases. Most of the penetrating injuries (13/17) were self-inflicted gunshot wounds in the head. Two patients had a combined cause of injury: one sustaining a major fall combined with hanging and the other with a combination of a minor fall and drowning. Ethanol intoxication (blood ethanol level > 1.00&) was determined post-mortem in 30 patients (26%), and amphetamine use was determined in three patients (2.6%). One patient sustaining a major fall was suspected to have suffered from an acute serotonin syndrome due to an anti-depressive intoxication. Pre-hospital The majority (n = 107, 93%) of the patients were transferred by ambulance, and eight patients (7%) were transferred by helicopter. Eight patients were trapped at

Eur J Trauma Emerg Surg 2009 No. 5

URBAN & VOGEL

457

Sderlund T, et al. Traumatic Deaths in the Emergency Room

Table 2. Cases of preventable deaths (n = 11). Patients whose death were considered to be preventablea 77m 37f 88m 65f 25m Psb Reasons of preventabilityc

0.61 0.13 0.05 0.85 0.36

52f 17f 26m

0.97 0.41 0.02

62f 24m

0.64 0.22

74m

0.67

Injuries were underestimated; haemorrhagic shock was not identified; trauma resuscitation was inadequate Intra-abdominal bleeding was diagnosed, but laparotomy was not performed after pelvic angioembolisation in haemodynamically unstable patient No interventions made to control bleeding from diagnosed pelvic fracture Pre-hospital oesophageal intubation was not identified in ER Bleeding from pelvic fracture and ruptured spleen were not controlled, and patient received only crystalloids and RBCs (18 units) during 3 h of treatment in the ER. Patient developed a non-reversible coagulopathy Patient refused blood transfusions due to religious reasons and died due to bleeding from several closed fractures Proximal control of bleeding was not done in traumatic pelvic amputation Blunt aortic injury was identified, but during the waiting of transfer to department of cardiothoracic surgery the contained mediastinal bleeding turned non-contained bleeding to pleural cavity Pelvic fracture was missed from CT scan and trauma resuscitation was inadequate Liver and spleen ruptures with intra-abdominal bleeding were diagnosed and treated conservatively despite haemodynamic instability. Patient developed non-reversible coagulopathy Patient had profuse bleeding from bilateral open femoral fractures and bleeding from open humeral fractures. Bleeding was ineffectively controlled

RBC: Red blood cell; ER: Emergency Room; CT: computed tomography; aNumber followed by lowercase letter indicates age (years) and sex (m: male; f: female) of the patient; b Ps: probability of survival, calculated using TRISS; cPreventability of the death was evaluated after peer-review of all medical records of the patients

the thorax (n = 6), the bony pelvis (n = 3) and a combination of several body regions (n = 15). The average age of the patients in the CNS group was 51 years, in the exsanguination group, 43 years, and in the other-reason group, 60 years. The average ISS of the patients <50 years and >50 years was 39.7 and 29.5, respectively. The ISS ranged from 4 to 75, and the patients dying of exsanguination had a higher average ISS (45) than those of the CNS and other-reason groups (32 and 29, respectively). The patients in the CNS group had a higher SBP (105 vs. 82 mmHg). The average NISS score was 42.4, and the NISS scores of 65 patients (56.5%) were higher than the ISS score. Preventable Deaths Eleven deaths (9.6%) were considered to be potentially preventable. These cases are presented in Table 2 with the preventable reason that was considered to contribute to the death. One death of these 11 preventable deaths was considered to be due to a missed injury in the ER (missed pelvic fracture with retroperitoneal bleeding, which could be seen in the CT scan). The P s

was compared between the preventable and non-preventable deaths (Table 3). In the non-preventable group, 21% of the deaths were probably preventable (Ps 0.51-0.75) or frankly preventable (Ps 0.76-1) compared to 45% in the preventable group. The average Ps of the preventable death group was 0.45 (IQ25-75 0.05-0.67) compared to 0.31 (IQ 25-75 0.090.45) in the non-preventable group. The preventable death group had a lower ISS (34 vs. 42) and a higher RTS (5.40 vs. 3.58) than the non-preventable group. The patients in the preventable death group received
Table 3. Comparison of the probability of survival between preventable and non-preventable deaths. Ps Preventable (n = 11) 4 (36) 2 (18) 3 (27) 2 (18) Non-preventable (n = 104) 52 (50) 30 (29) 14 (13) 8 (8)

0-0.25 0.26-0.5 0.51-0.75 0.76-1.00

Values are presented as the number of cases, with the percentage in parenthesis

458

Eur J Trauma Emerg Surg 2009 No. 5

URBAN & VOGEL

Sderlund T, et al. Traumatic Deaths in the Emergency Room

Table 4. Clinically significant (AIS > 3) missed injuries (n = 18) in 18 patients. Thorax Rupture of descending aorta Rupture of descending aorta Rupture of descending aorta Air embolism of right ventricle Lung contusion Heart contusion Heart contusion Heart contusion Unilateral (tension) pneumothorax Head and neck Skull fracture Spinal cord contusion Skull fracture CIV fracture with spinal cord contusion AIS 6 6 5 5 4 4 4 4 4 AIS 5 4 4 4 Abdomen Liver rupture Liver rupture AIS 5 4

fractures (3/18, 17%) and abdominal injuries (2/18, 11%). The patients in the preventable death group (n = 11) had 18 missed injuries, which included only two injuries with AIS 4. There were 97 missed injuries in the non-preventable death group (n = 104), including 16 injuries with AIS 4.

Pelvis Pelvic fracture Pelvic fracture Pelvic fracture

AIS 5 4 4

AIS: Abbreviated Injury Scale. Values of each injury are shown

an average of 12.4, 1.5 and 0.5 units of RBC, FFP and platelets; in comparison, the patients in the non-preventable group received 7.2, 0.9 and 0.4 units, respectively. The patients in the preventable group received considerably higher volumes of intravenous fluids than those in the non-preventable group (average 11.5 vs. 4.2 l). The deaths were categorized into three groups according to the time of arrival at the ER: during normal office hours (0800-1600 hours); in the evening hours (1600-2400 hours); during the night (0000-0800 hours). The total number of deaths and the number of potentially preventable deaths were 44/2 (4.5% preventable), 48/4 (8.3% preventable) and 23/5 (21.7% preventable) for office hours, evening and night-time, respectively. In eight of the 11 potentially preventable deaths, the surgeon-in-charge was the in-house resident initially without an attending senior trauma surgeon. Missed Injuries The post-mortem examinations revealed a total of 115 injuries in 50 patients that had been missed both in the clinical and radiological surveys in the ER; this indicates that a missed injury occurred in 43% of the cases. The severity of 97 of these missed injuries ranged from 1 to 3 on the AIS; the severity of the remaining 18 injuries (18 patients) ranged from 4 to 6 on the AIS (Table 4). Thus, 15.7% of the patients had a clinically significant missed injury (AIS 4). These AIS 4 grade injuries were mainly in the chest (9/18, 50%), followed by head and spine injuries (4/18, 22%), pelvic

Discussion The rate of potentially preventable deaths in the ER was determined to be 9.6% in this study. Reported rates of potentially preventable trauma deaths vary from 1% in level 1 trauma centres up to 33% in smaller volume units [17-21]. Most of our patients (89.6%) died within 4 h after arrival to the ER. The majority of clinically significant missed injuries are found in the early-phase deaths [6], and preventable deaths at this phase are mainly due to a failure to stop bleeding or prevent hypoxia or to a delay in surgical treatment [18, 20, 22], as was also seen in this study. In revealed by our analysis, one clearly preventable death occurred due to ethical reasons as the patient refused blood transfusions on religious grounds, and it may be argued whether this death was acceptable or not. One burn victim died of prolonged hypoxia due to unrecognized placement of the endotracheal tube in the oesophagus. Oesophageal placement of the endotracheal tube was also identified in the autopsy of another patient who had suffered a severe head injury. The oesophageal placement of the endotracheal tube was not considered to contribute to the death because the initial traumatic brain injury itself was incompatible with life. In both cases, the intubation was done at the scene by a pre-hospital unit, but the misplacement was not recognized in the primary or secondary survey in the ER. In the literature, the incidence of oesophageal insertion of the endotracheal tube by pre-hospital units has been reported to range from less than 0.5 up to 6% [23], but it is more than reasonable to argue that such a condition should be found in the primary survey (e.g. clinical examination, chest X-ray, and expiratory CO2) upon the patients arrival at the ER. Seven potentially preventable deaths were considered to be failures in surgical decision-making, resulting in futile non-operative treatment or a delay in the surgical control of bleeding that subsequently led to the development of uncontrollable coagulopathy. In one of the seven cases there were, in addition to severe pelvic bleeding, also missed injuries (liver rupture, AIS 3, and bleeding from mesenterial artery, AIS 2). The development of coagulopathy was not sufficiently

Eur J Trauma Emerg Surg 2009 No. 5

URBAN & VOGEL

459

Sderlund T, et al. Traumatic Deaths in the Emergency Room

prevented and/or treated in several cases, as the patients received large volumes of crystalloids and RBC, but relatively low amounts of FFP and platelets. In two cases (77-year-old male and 62-year-old female), the injuries were underestimated and partially missed, and a haemorrhagic shock was not recognized. The initial resuscitation was suboptimal, leading to the development of a non-reversible haemorrhagic shock and the patients not responding to any further resuscitation. The early recognition of a haemorrhagic shock is a common pitfall in the initial resuscitation [24]; it can be especially demanding to recognize this state in elderly patients having arterial hypertension and using betablockers, as the blood pressure and heart rate can be misleading. A higher fraction of patients in the potentially preventable death group than in non-preventable group had Ps > 0.50 (45 vs. 21%, respectively), and the average Ps was higher in the potentially preventable group (0.45 vs. 0.31). A P s > 0.50 has been used to identify patients with unexpected death and thus worthy of peer-review [25]. Five patients in the potentially preventable death group had P s > 0.50. If we had carried out a peer-review only of deaths with P s > 0.50, our rate of potentially preventable deaths would have been 4.3%. For the Ps calculations, we used the TRISS method with coefficients derived from the Major Trauma Outcome Study (MTOS [26]). However, these coefficients may not be appropriate for our study population for several reasons. In MTOS. the patients intubated before admission to ER were excluded (78% of our patients were intubated at the scene), and most of the excluded patients in MTOS were blunt trauma patients (81.7% of our patients had blunt trauma) [26]. The mortality of the excluded patients in MTOS was 16.8% compared to 8.0% of the patients included in the analyses [26]. Consequently, we suggest that in our study population the decision for peer-review should not be based on the TRISS method P s > 0.50 as a cut-off. At Toolo Hospital a senior trauma surgeon is oncall at home during the night-time on most days, and a resident surgeon is in charge of trauma resuscitations. The lack of an in-house senior trauma surgeon during the night-time may partially explain the potentially preventable deaths in the light of delayed surgical decision-making, as four of the five potentially preventable deaths that occurred at night-time were considered to be due to a lack of experience. Two preventable deaths which occurred during the office hours were also considered to result from delayed surgical decision-making, even though a senior trauma surgeon was involved in the trauma resuscitation. Also,

one preventable death during the night-time was treated initially by a senior trauma surgeon. A senior trauma surgeon was present in the initial resuscitation of the preventable deaths (n = 11) in three cases (27%). In the non-preventable deaths (n = 104), the initial treatment was performed by a resident surgeon alone in 84 cases (81%). Thus, the senior trauma surgeon was initially present more often in the preventable deaths than in the non-preventable ones. The effect of the presence of a 24-h in-house trauma surgeon in terms of reducing the mortality of critically injured trauma patients is a controversial question, as the 24-h presence of such a specialist has been reported to both decrease mortality [27] and have no significant effect [28-30]. However, the volume of patients handled by the trauma centre has been shown to have a clear impact on the survival of the patients being treated [29, 31, 32]. The Toolo Hospital trauma protocol was established in 2002 to promote decision-making tools and education for surgeons leading trauma resuscitations. The trauma protocol is also aimed at providing a standard quality of care around the clock. Systematic trauma team simulation training was initiated in 2003 to encourage trauma teams to implement new protocols and to improve overall team performance. The massive transfusion protocol was established at the beginning of 2005 and subsequently incorporated into the trauma protocol. However, the ratios of RBC, FFP and platelets before (9.1:1.2:0.5) and after (7:0.7:0.3) the implementation of the massive transfusion protocol were not improved in patients dying in the ER. The availability of CT scanning was good throughout the study period, and FAST, performed by radiologists, has also become a common routine in the trauma patient survey since 2003. At the beginning of our study period, the technique for pelvic stabilization in the ER was anterior external fixation, which was performed on haemodynamically unstable patients with an unstable fracture of the pelvic ring. Since 2005, we have used the commercially available stabilization belt (T-POD; Pyng Medical, Richmond, BC, Canada) for the initial stabilization of pelvic fractures in the ER. We did not find any potentially preventable deaths during the last 2 years (2005-2006) in our cohort study. However, due to overall low annual numbers of traumatic deaths in the ER (23 deaths during 2005-2006), we are unable to draw any definitive conclusions on the possible effects of the evolved trauma care on the potentially preventable deaths. A longer time survey is needed to identify the possible effects of the changes in our trauma care protocol.

460

Eur J Trauma Emerg Surg 2009 No. 5

URBAN & VOGEL

Sderlund T, et al. Traumatic Deaths in the Emergency Room

Missed injuries did not play a major role in our study. They contributed to three deaths, but only in one case was a missed injury the main reason for preventability. In other studies, unexpected injuries (i.e. missed injuries without CT) of the thorax and abdomen were seen in 42-74% of the patients [33, 34]. The number of missed injuries in the CNS group (21 patients with 41 missed injuries) was mainly due to the fact that in many cases the treatment of the patient was stopped after the head CT scan revealed brain injuries incompatible with life. The post-mortem autopsies revealed 56 missed injuries in 24 patients in the exsanguination group. This result is partly explained by the high proportion of blunt polytraumas in this group (28/ 29 patients) combined with the severely compromised clinical condition of the patients, which prevented CT scanning (body CT scan performed only on 10/29 patients), and partly by the fact that the patients died before possible surgical exposures could have revealed these injuries. Computed tomography scanning is not feasible in critically injured patients, and the decisions for surgical interventions in such cases have to be based on the clinical examination. In treating these patients, the experience of the trauma surgeon can be assumed to have a major impact. Computed tomography scanning is considered to be the golden standard of blunt trauma patient survey, and it is especially important in the early diagnostics of sedated and intubated blunt trauma patients [33, 34]. In our study cohort, the majority of the patients were intubated (89%), but 51 patients were haemodynamically unstable (SBP 90). Thus, the majority of the patients should have been scanned with CT, but the haemodynamic instability prevented it. Computed tomography scanning is also associated with a risk of delaying some critical emergency procedures if the patient in extremis is taken to relatively time-consuming scanning instead of making a prompt decision to operate on the patient. However, there is a possibility of incorporating CT scanning in a routine trauma patient survey in a time-effective way and to gain the benefits of full-body CT scanning without losing time to start possible key-emergency operations [35]. The ER and CT of Toolo hospital have recently been renovated, which has greatly facilitated the performance of fast track scanning as part of the trauma patient survey. The possible effects of this change in our trauma protocol on timing and overall performance will be seen after a few years. We conclude that our rate of preventable deaths is relatively high compared to those reported earlier in Level 1 trauma centres. On the other hand, it can al-

ways be argued that the reported figures may not be comparable as such due to the broad variety of definitions used for the parameters studied. All deaths were peer-reviewed regardless of the estimated probability of survival. More accurate definitions of parameters as well as high-quality trauma registries are needed for reliable benchmarking. The role of the experienced senior trauma surgeon is of vital importance throughout the acute trauma care, and efforts have to be made for obtaining a sufficient number of well-educated and experienced surgeons to cover this task 24/7. A continuous control of the quality of the given care has to be part of the normal routine in a major trauma centre; equally, performance improvement should be an ongoing process in these centres.

Acknowledgments
The authors wish to thank Jouni Backman MD/PhD (Helsinki University Central Hospital, Department of Clinical Pharmacology), for providing insight into the serotonin syndrome.

Conflict of interest statement

The authors declare that there is no actual or potential conflict of interest in relation to this article.

References
1. 2. Murray CL, Lopez AD. Mortality by cause for eight regions of the world: global burden of disease study. Lancet 1997;349:1269-76. Peden M, McGee K, Sharma G. The injury chart book: a graphical overview of the global burden of injuries. World Health Organization, Geneva, 2002. Nijs S, Broos P. Trauma care systems in Belgium. Inj Int J Care Inj 2003;34:652-7. Statistics Finland. Deaths, cause of death. Available at: http:// www.tilastokeskus.fi. 2003. Sreide K, Krger AJ, Vrdal AL, Ellingsen CL, Sreide E, Lossius HM. Epidemiology and contemporary patterns of trauma deaths: changing place, similar pace, older face. World J Surg 2007; doi: 10.1007/s00268-007-9226-9. Hodgson NF, Stewart TC, Girotti MJ. Autopsies and death certification in deaths due to blunt trauma: what are we missing? Can J Surg 2000;43:130-6. Jin PHFK, Klaver JF, Maes A, Ponsen KJ, Das C, Goslings JC. Autopsies following death due to traumatic injuries in Netherlands: an evaluation of current practice. Inj Int J Care Inj 2008;39:83-9. Ong AW, Cohn SM, Cohn KA, Jarmillo DH, Parbhu R, McKenney MG, Barquist ES, Bell MD. Unexpected findings in trauma patients dying in the intensive care unit: results of 153 consecutive autopsies. J Am Coll Surg 2002;194:401-6. Harviel JD, Landsman I, Greenberg A, Copes WS, Flanagan ME, Champion HR. The effect of autopsy on injury severity and survival probability calculations. J Trauma 1989;29:766-73. Teasdale G, Jennett B. Assessment of coma and impaired consciousness: a practical scale. Lancet 1974;304:81-4.

3. 4. 5.

6.

7.

8.

9.

10.

Eur J Trauma Emerg Surg 2009 No. 5

URBAN & VOGEL

461

Sderlund T, et al. Traumatic Deaths in the Emergency Room

11.

12.

13.

14.

15. 16.

17.

18.

19.

20.

21.

22.

23.

24.

25. 26.

Association for the Advancement of Automotive Medicine (AAAM). The abbreviated injury scale 2005. AAAM, Barrington, 2005. Available at: http://www.aaam.org. Baker SP, ONeill B, Haddon W, Long WB. The injury severity score: a method for describing patients with multiple injuries and evaluating emergency care. J Trauma 1974;14:187-96. Osler T, Baker SP, Long W. A modification of the injury severity score that both improves accuracy and simplifies scoring. J Trauma 1997;43:922-5. Champion HR, Sacco WJ, Copes WS, Gann DS, Gennarelli TA, Flanagan ME. A revision of the trauma score. J Trauma 1989;29:623-9. Boyd CR, Tolson MA, Copes WS. Evaluating trauma care: the TRISS method. J Trauma 1987;27:370-8. Skaga NO, Eken T, Steen PA. Assessing quality of care in a trauma referral center: benchmarking performance by TRISSbased statistics or by analysis of stratified ISS data? J Trauma 2006;60:538-47. Acosta JA, Yang JC, Winchell RJ, Simons RK, Fortlage DA, Hollingsworth-Fridlund P, Hoyt DB. Lethal injuries and time to deaths in a level I trauma center. J Am Coll Surg 1988;186: 528-33. Anderson ID, Woodford M, de Dombal FT, Irving M. Retrospective study of 1,000 deaths from injury in England and Wales. Br Med J 1988;296:1305-8. Gilroy D. Deaths from blunt trauma, after arrival at hospital: plus a change, plus cest la mme chose. Inj Int J Care Inj 2005;36:47-50. Gruen RL, Jurkovich GJ, McIntyre LK, Foy HM, Maier RV. Patterns of errors contributing to trauma mortality: lessons learned from 2,594 deaths. Ann Surg 2006;3:371-80. Shackford SR, MacKersie RC, Davis JW, Wolf PL, Hoyt DB. Epidemiology and pathology of traumatic deaths occurring at a level I trauma center in a regionalized system: the importance of secondary brain injury. J Trauma 1989;29:1392-7. Tien HC, Spencer F, Tremblay LN, Rizoli SB, Brenneman FD. Preventable deaths from hemorrhage at a level I Canadian trauma center. J Trauma 2007;62:142-6. Jemmett ME, Kendal KM, Fourre MW, Burton JH. Unrecognized misplacement of endotracheal tube in a mixed urban to rural emergency medical service setting. Acad Emerg Med 2003;10:961-5. MacKersie RC, Dicker RA. Pitfalls in the evaluation and management of the trauma patient. Curr Probl Surg 2007;44: 778-833. Champion HR, Sacco WJ, Hunt TK. Trauma severity scoring to predict mortality. World J Surg 1983;7:4-11. Champion HR, Copes WS, Sacco WJ, Lawnick MM, Keast SL, Bain LW, Flanagan ME, Frey CF. The Major trauma outcome study: establishing national norms for trauma care. J Trauma 1990;30:1356-65.

27.

28.

29.

30.

31.

32.

33.

34.

35.

Rogers FB, Simons R, Hoyt DB, Shackford SR, Holbrook T, Fortlage DA. In-house board-certified surgeons improve outcome for severely injured patients: a comparison of two university centers. J Trauma 1993;34:871-7. Luchette F, Kelly B, Davis K, Johanningman J, Heink N, James L, Ottaway M. Impact of the in-house trauma surgeon on initial patient care, outcome, and cost. J Trauma 1997;42:490-7. Pasquale MD, Peitzman AB, Bednarski J, Wasser TE. Outcome analysis of Pennsylvania trauma centers: factors predictive of nonsurvival in seriously injured patients. J Trauma 2001;50:465-74. Thompson CT, Bickell WH, Siemens RA, Sacra JC. Community hospital level II trauma center outcome. J Trauma 1992;32: 336-43. Smith RF, Frateschi L, Sloan EP, Campbell L, Krieg R, Edwards LC, Barret JA. The impact of volume on outcome in seriously injured trauma patients: two years experience of the Chicago trauma system. J Trauma 1990;30:1066-76. Konvolinka CW, Copes WS, Sacco WJ. Institution and per-surgeon volume versus survival outcome in Pennsylvanias trauma centers. Am J Surg 1995;170:333-40. Deunk J, Dekker HM, Brink M, van Vugt R, Edwards MJ, van Vugt AB. The value of indicated computed tomography scan of the chest and abdomen in addition to the conventional radiologic work-up for blunt trauma. J Trauma 2007;63:757-63. Traub M, Stevenson M, McEvoy S, Briggs G, Lo SK, Leibman S, Joseph T. The use of chest computed tomography versus chest x-ray in patients with major blunt trauma. Inj Int J Care Inj 2007;38:43-7. Weninger P, Mauritz W, Fridrich P, Spitaler R, Figl M, Kern B, Hertz H. Emergency room management of patients with blunt major trauma: evaluation of the multislice computed tomography protocol exemplified by an urban trauma center. J Trauma 2007;62:584-91.

Address for Correspondence Tim Sderlund Department of Orthopedics and Traumatology Tl Hospital Helsinki University Central Hospital Topeliuksenkatu 5 P.O. Box 266 00029 Helsinki Finland e-mail: tim.soderlund@helsinki.fi

462

Eur J Trauma Emerg Surg 2009 No. 5

URBAN & VOGEL