Survival and neurologic outcome after traumatic out-of-hospital cardiopulmonary arrest in a pediatric and adult population: a systematic review

Modern emergency medical services (EMS) systems began in the 1960s to provide emergency medical care in the community in industrialized countries. Historically, the aim of EMS was responding to any life-threatening emergency. Later on some of the goals were also providing rapid medical response and treatment to patients suffering out-of-hospital cardiopulmonary arrest (OHCA) and to victims of vehicular trauma.

A large group of patients do not survive traumatic OHCA despite appropriate, immediate, and extensive efforts by bystanders and EMS personnel. These patients present challenges to the system, because prehospital resuscitative efforts are labor- and resource-intensive, as are these efforts when continued in the emergency department (ED). It is usually assumed that trauma victims with cardiac arrest in the prehospital setting have a dismal prognosis, leading physicians to consider cardiopulmonary resuscitation in traumatic cardiac arrest as futile [1]. The epidemiology of mortality following injury suggests that 34% of traumatic deaths occur before hospital arrival [2]. Despite advances in medical care, survival rates of 0% to 2% have been reported for patients with blunt trauma who arrive at a trauma center with no signs of life [3‐7]. Unfortunately, many of these survivors suffer from severe permanent neurological disability [6, 8].

The National Association of EMS Physicans/American College of Surgery Committees on Trauma produced guidelines in 2003 about withholding or termination of resuscitation efforts in OHCA [7]. According to these guidelines, it is suggested not to resuscitate patients with blunt trauma and OHCA or patients with penetrating trauma who have no detectable pulse, are apneic and without signs of life, and accordingly to pronounce death at the scene of injury. On the other hand it has been proposed in some studies that some traumatic cardiac deaths following motor-vehicle collisions occur because of airway obstruction of the comatose patients [9, 10]. Simple airway maneuvers and intubations could possibly prevent hypoxic cardiac arrest. Other early deaths following trauma might be due to exsanguination and might be prevented in some cases by external pressure on areas with visible, profuse blood loss. Since the publication of these guidelines, at least two articles have described improved survival rates [11, 12] and one has described possible deviation of the guidelines [11]. On the other hand one recent article supported the current guidelines regarding the withholding or termination of resuscitation of patients with prehospital traumatic cardiopulmonary arrest (TCPA) [13].

However, despite the poor survival of pediatric trauma patients receiving cardiopulmonary resuscitation (CPR) (1.5% to 25%), cessation or withholding of CPR after traumatic injury in a child is a difficult and emotionally charged issue [14‐20]. Clear-cut resuscitation guidelines for children requiring CPR after traumatic injury do not exist to our knowledge. A review suggested that almost 22% of cardiac arrests in children are trauma associated, rendering this one of the most common causes of pre-hospital cardiac arrest and death between the ages of 1 to 16 years ([5, 21, 22].

However, no strong evidence is available for the mortality in OHCA. Hence, the aim of this study was to summarize and analyze the best available data on mortality and survivor outcome after traumatic OHCA by performing a systematic literature review.

For this systematic review, an Ovid Medline-based literature search was performed to identify any published clinical studies on mortality after out-of-hospital resuscitation of patients suffering a trauma, including the following databases: MEDLINE, MEDLINE preprints, EMBASE, CINAHL, Life Science Citations, British National Library of Health, Cochrane Central Register of Controlled Trials (CENTRAL). The literature search period was from January 1964 to May 2011. The search was performed on 15 May 2011 using the strategy given in Figure 1. Specific attention was placed on identifying studies that describe the neurological outcome of survivors.

A total of 476 clinical studies were identified and all abstracts were evaluated on screening by two independent reviewers. The only studies included were those reporting patients who received preclinical resuscitation after sustaining trauma, and who were admitted to hospital. The level of evidence was categorized according to the definition given by the Oxford Centre for Evidence-based Medicine published by Hanzlik [23]: 476 studies met these inclusion criteria and full-text articles of these studies were obtained for review. Of these, 429 studies had to be excluded from analysis for reasons demonstrated in Figure 2. The remaining 47 studies were included in the present analysis. Two independent reviewers systemically extracted data that included study characteristics and design, level of evidence, demographic and epidemiological parameters, number of included patients, mechanism of injury, neurological and outcome.

These studies were subdivided into those that were exclusively pediatric studies, and those in a general population in which children were not always explicitly excluded and therefore called a potentially mixed population. The average age in these groups was between 29 and 65 years. Some of the identified studies included children, but the vast majority were based on adults. Specific focus was placed on extracting data describing the trauma mechanism and analyzing the neurological outcome of survivors. To assess for the methodological quality of the neurological outcome the patients were divided into four groups using the Glasgow outcome scale with the following description [24]: I) persistent vegetative state: the patient exhibits no obvious cortical function; II) severe disability: the patient is conscious but disabled and depends upon others for daily support due to mental or physical disability or both; III) moderate disability: the patient is disabled but independent as far as daily life is concerned. The disabilities found include varying degrees of dysphasia, hemiparesis, or ataxia, as well as intellectual and memory deficits and personality changes, and IV) good recovery: the patient has resumed normal activities even though there may be minor neurological or psychological deficits.

The outcome parameters have been used in the studies included in this systematic review. The data were independently cross-checked with the original papers using a standard quality control procedure. Any differences of opinion between the original reviewer and quality control reviewer were resolved by discussion and referenced to the study paper. The data were analyzed using established statistical software.

To analyze the proportion of survivors from different studies, we used binomial response generalized linear models. This allowed analysis of the effect of factors that might influence the proportion. Specifically, we considered the year of publication, where 2010 was taken as a reference, the dimension of the study, and gender proportion. As the studies on children might differ from the studies on adults concerning these factors, we used the adult data to check for their influence, before incorporating factors into joint analysis of child and adult data. For statistical analyses we used statistical environment R (2.13.0) software (R foundation for Statistical Computing, Vienna, Austria).

Injury is the leading cause of death in industrialized countries between the ages of one and forty-four years [7, 25, 26]. Survival after traumatic OHCA is rare, even with maximal resuscitative efforts. There are recommendations in the literature, at least for adults, mandating the emergency team to judge whether to withhold or terminate resuscitation efforts in adult TCPA. Moreover many factors discussed during this decision-making and have been sparsely investigated or are based on marginal or no scientific evidence (such as age, trauma mechanism, time of arrest until start of life support, blood loss and ethical aspects). In order to summarize the available scientific data for some of the above mentioned criteria, the present systematic review was performed.

The aim of the study was to describe the mortality rate after traumatic OHCA in a predominantly adult population and in children taking into account the trauma mechanism and the neurological outcome of survivors on discharge from hospital.

The approach of the present study was also to include case series and studies with a lower level of evidence, because analyses of solely high-quality studies are impossible.

The American College of Surgeons Committee on Trauma suggests withholding resuscitation efforts in any trauma patient found 'apneic, pulseless, and without organized electrocardiographic activity or other sign of life such as spontaneous movements or pupillary reflexes' [7]. It is argued that the number of salvageable patients is too small, considering the inherent costs and risks of resuscitation efforts. A comprehensive review of the literature, on which the 2003 guidelines are based, demonstrated survival rates of 0% to 3.7%, which is similar to our findings except in the pediatric patients [7]. Recently a study of 294 patients, who suffered prehospital TCPA and met the criteria of the guidelines, showed a survival of only one patient (0.3%) with a Glasgow Coma Scale (GCS) of six and violation of the guidelines resulted in six resuscitated patients in a neurologically devastated condition [13]. David et al. found in a prospective study that survival and neurological outcome after OHCA did not differ between trauma and medical patients and they suggested that under the supervision of senior physicians, active resuscitation after traumatic OHCA is as important as in medical patients. Moreover they agreed with the findings of Martin et al. [6], who suggested that the initial rhythm of the heart (pulseless electrical activity) in trauma is probably not a robust predictor of patient outcome and came to the conclusion that trauma patients in cardiac arrest need active resuscitation in the field until they regain cardiac activity instead of quick transportation to a trauma center, especially if the transport will last for > 15 minutes [7].

On the other hand, there are studies that impressively show survivors being resuscitated against the guidelines of the National Association of Emergency Medical Services Physicians and the American College of Surgeons Committee on Trauma (NAEMSP/ACS-COT), although they had met criteria for non-treatment according to the guidelines [11]. Pickens et al. concluded that the prehospital clinical assessments are not reliable for the triage of TCPA patients and that they should be transported to the ED for further evaluation and care [11]. This article described several guideline deviations in traumatic OHCA patients. If EMS personnel had adhered to the guidelines, 111 out of 138 transported patients should not have been transported to the ED and even more striking, 13 out of 14 survivors should not have been resuscitated. Most of the 14 survivors described in that study breached one of the time-related guidelines [11]. The outcome was relatively good with a discharge GCS for the survivors of 13.4 (SD 3.1) with 64.3% (n = 9) having the maximum score of 15. Moreover Pickens et al. suggested that triaging TCPA patients in the ED would not generate large expenses or lead to lengthy hospitalizations and he thereby attributed the cost associated with trauma resuscitations as one of the reasons for the creation of the guidelines.

The mortality with a penetrating trauma mechanism (96.4% in the mixed population and 97.8% in the pediatric group) versus a blunt trauma mechanism (96.7% in the mixed population and 73.8% in the pediatric group) was significantly different in the pediatric group in our systematic review.

The literature is split on this issue, as several authors have not found any significant difference [3, 5, 8, 11, 27], while others have found higher survival among victims of penetrating trauma [28]. Data from Stockinger and McSwain showed a significant survival benefit for victims of blunt trauma [28] and in a study of TCPA patients in the ED, Cera et al. identified 13 out of 15 survivors with blunt trauma (P = 0.06) [12]. However the penetrating injuries themselves are associated with different outcomes. Durham identified a survival rate of 15.2% in patients sustaining stab wounds compared to 7.3% in patients with gunshot wounds [29].

The guidelines determined that 'penetrating trauma, particularly if isolated to the thorax, has a better prognosis than blunt or multisystem penetrating trauma. Survival from cardiopulmonary arrest due to blunt trauma is grave indeed, likely due to the multisystem nature of the injuries sustained.' [7]. In our analysis the overall survival rate in the mixed population was nearly the same at 3.3% in victims of blunt trauma and 3.65% in victims of penetrating trauma.

The mentioned guideline should not be applied to children since the literature suggests improved survival for children, and moreover, supporting evidence is lacking [30]. There are no specific guidelines for the management of children in cardiac arrest secondary to trauma.

In the evaluated results for children a significant difference was identified between the survival rates of 26.2% in blunt trauma and 2.2% in penetrating trauma. With the knowledge of these data this conclusion of the guidelines probably must be revised in the future.

No evident reason could be found for the relatively high survival rate in children with blunt trauma. Perron concluded in his study that because of the potential that cardiopulmonary arrest in a pediatric patient may not be a result of exsanguination but of a primary respiratory cause, the injured patient having arrested in the prehospital setting may be a different entity to an adult patient in a similar circumstance and warrants a more aggressive initial resuscitative effort [20]. Among children, respiratory compromise is the most common cause of cardiac arrest [31, 32]. Previous work has suggested that cardiac arrest in children with trauma is often caused by bleeding or prolonged hypoxemia [16]. Hypovolemic cardiac arrest is often associated with a poor outcome [16, 33‐35]. The high rate of bleeding associated with hypovolemic arrest in the pre-hospital phase is unlikely to respond to conventional CPR. In contrast, when apnea precedes cardiac arrest, provision of adequate oxygenation may restore spontaneous circulation.

After applying inclusion and exclusion criteria, a total of 47 studies reporting mortality of 5,391 patients after traumatic OHCA were included in the present analysis. The patients' characteristics appear representative for what has been described concerning gender, population and age. Male patients are more likely to be involved in trauma as is the case in the general population. These observations correlated with what has been reported by several review articles earlier.

The vast majorities of the studies identified were case series without a control group, and therefore represent a low level of evidence. No single study with a level of evidence I, II or III was included, so all studies had level IV evidence, which displays insufficient overall evidence. Furthermore, the quality of the investigated published data was relatively poor. Most of the studies found had to be excluded because the reason for resuscitation was non-traumatic. Even though there is so little evidence present in the literature it was surprising that 59 studies had to be excluded because they comprised suggestions on therapeutics and comments, 54 were review articles and 10 were different guidelines. Unfortunately, in 37 studies, patients were not analyzed separately according to whether resuscitation was performed due to traumatic or non-traumatic causes. In the extracted 47 studies important information was missing. For example, the distribution between the sexes was listed in 20 studies only, and the average age of the population was given in 18 studies only.

Against the aim of the authors, some points could not be addressed by the present review. Further limitations are that almost no information was available about relevant parameters in the analyzed studies, such as the period of time before basic or advanced life support was performed; the length of time that CPR was performed until the return of spontaneous circulation; whether CPR was performed at the scene or during the transport; the frequency of the need for CPR until reaching the ED; the length of time taken for transport to hospital and the time period between occurrence of the accident and arrival at the ED. The authors believe that these are also important variables which could have a more or less important impact on survival.

Significant lower mortality was identified by separating the results of the mixed population in European trials, where patients were mainly attended by physicians at the trauma scene compared to the mainly American studies, where EMS performed the trauma response. The biggest European study identified a mortality of 92.5% in 909 evaluated patients [35]. On the other hand there are also huge American studies with mortality between only 91% and 92% [11, 29]. In the pediatric group mortality was significantly less in the American studies with one study including 642 pediatric patients with an evaluated mortality of only 74.8% [20]. The biggest European study included 80 patients with an investigated mortality of 91.3% [33]. The potential reasons for the lower mortality could be multiple and should be investigated by further studies. Whether the presence of a physician potentially affects mortality cannot not be proved using the available investigated populations and data.

With regard to the fact that in this field scientific evidence is very low, further high quality and multi-center studies should be demanded. Results could help in the decision-making of medical teams caring for traumatized patients in the initial out-of-hospital phase when TCPA occurs. Large national registers like the German Trauma Register have played an important role in answering such questions [36] and can continue to do so in the future.

Nevertheless, since the aim of the authors was to describe the literature available on survival and clinical outcome following OHCA caused by trauma, and to draw the best possible conclusion based upon the evidence available, even small studies have not been excluded, but have been analyzed for the most important findings.

However, medical personnel caring for patients with a traumatic OHCA are, and always will be, facing an ethical conflict, and must come to a decision within only a few minutes about when and if to stop resuscitative efforts, knowing about a high percentage of mortality and probably a poor neurological outcome for survivors. But as shown above, more than half of the survivors (adults and pediatric patients) have only moderate disabilities or even gain full recovery of neurological function.

Survival rates can help in the creation of guidelines and standardized decision-making, but after all, the decision is still dependent of the ethical human factor and from persons who have to face every situation based on its own individual criteria, which could influence the decision- making. Therefore, this meta-analysis could support recommendations for future guidelines.

However, evidence based support for the important decision about whether or not to resuscitate a patient in preclinical TCPA is sparse. Thus, the authors hope to encourage all medical disciplines involved in the treatment of such patients to perform further clinical research in this field to obtain a higher level of evidence.