According to the World Health Organization, injuries are the leading cause of death for people under the age of 45. In Europe, nearly 800,000 people die from injuries every year 1 . The prevention and management of severe trauma are thus major public health issues in most countries.

Blunt trauma and penetrating trauma present major differences in clinical presentation, management and outcomes 2 3 . In contrast to trauma epidemiology in the United States, blunt trauma is more frequent than penetrating trauma in most European countries, and the leading cause of severe trauma. Medical pre-hospital management (SMUR: Service Mobile d'Urgences et de Réanimation), generally performed by an emergency physician (EP), may take longer than care provided by fire brigades. The benefit for blunt trauma victims of SMUR management on the scene of the accident is controversial 4 .

To our knowledge, no previous studies have investigated the benefit of SMUR versus non-SMUR management on the prognosis of patients with blunt trauma. The FIRST (French Intensive care Recorded Severe Trauma) observational study was initiated to describe the management of patients with severe blunt trauma. The main aim of the present analysis was to examine whether SMUR care reduced 30-day mortality, in comparison with non-SMUR care provided by fire brigades. A secondary aim was to assess the impact of SMUR care on 72-hour mortality.

The 2,703 patients in the study sample comprised 2,063 men (76%) and 640 women (24%) with a mean age of 41.1 years (SD: 18.0 years). The median ISS was 25 (IQR: 18 to 34). Pre-hospital management was performed by SMUR for 2,513 patients (93%) and by fire brigades for 193 patients (7%). In comparison with non-SMUR patients, SMUR patients were significantly younger and more often directly referred to university hospitals (Table 1). SMUR patients were less frequently admitted to the first hospital within the first hour, and more frequently to an ICU within the first three hours after the accident. For the 2,015 patients (1,911 SMUR and 104 non-SMUR patients) with available information, the median time spent on the scene (IQR range) was significantly higher in SMUR patients (46 minutes, IQR: 30 to 68 minutes) than in non-SMUR patients (18 minutes, IQR: 13 to 27 minutes, P < 0.001). The median transport time was also higher in SMUR patients (54 minutes, IQR: 35 to 79 minutes) than in non-SMUR patients (40 minutes, IQR: 25 to 65 minutes, P < 0.001). SMUR patients were more frequently victims of traffic accidents than were non-SMUR patients. The type of pre-hospital management was not significantly related to gender, accident time or accident day.

As indicated in Table 2, an intravenous line was inserted in almost all SMUR patients and a large proportion of them received either crystalloids (71%) or colloids (47%). About half of the patients had tracheal intubation and were given artificial ventilation, whereas a smaller proportion (12%) received vasopressors in the pre-hospital phase. Very few patients received blood-derived products or required chest tube insertion. The GCS score had only a marginal influence on fluid administration. Although tracheal intubation and mechanical ventilation significantly decreased along with GCS score (P < 0.001), their use remained relatively frequent among patients with GCS score >13 (14.1% and 13.5%, respectively).

Among SMUR patients, 74 presented with cardiac arrest during transport and were alive upon arrival at the hospital. Among these, 14 patients (19%) survived to post-trauma Day 30, but only 3 patients made a good neurological recovery. These patients were excluded from subsequent analyses for two reasons. As mentioned above, because of the French law that imposes a systematic call to SMUR in cases of cardiac arrest during transport, we were unable to distinguish between patients initially managed by fire brigades and those initially managed by SMUR. Secondly, we had incomplete information about the clinical status and injury assessment for the majority of these patients.

The initial physiological status, GCS score and ISS according to type of pre-hospital management are compared in Table 3. In comparison with patients transported by fire brigades, SMUR patients had a poorer initial status. They presented a lower GCS score (P < 0.001), SpO₂ (P = 0.052), a higher ISS (P < 0.001) and a higher frequency of abnormal pupils (P < 0.001).

Up until hospital discharge (within 30 days after the accident), the death rate was comparable in patients transported by fire brigades (29 deaths, 15%) and SMUR patients (407 deaths, 17%, P = 0.61) (Table 4). Stratification on the GCS score revealed that the risk of death was systematically lower among SMUR patients than in patients transported by fire brigades, although the between-group difference only reached the level of significance among patients with a GCS score ≥14 (P = 0.025). A significantly better prognosis was also observed among SMUR patients with an ISS <25 (P = 0.002).

Since SMUR patients presented with a more severe initial status, the relationship between the type of pre-hospital management and the risk of death was first adjusted for the GCS score, the ISS and the main initial physiological variables. The risk of death at 30 days was significantly lower (OR: 0.55, 95% CI: 0.32 to 0.94, P = 0.03) in SMUR patients than in non-SMUR patients (Table 5, model 1). Increasing age, high ISS, low GCS score, low initial systolic blood pressure and SpO₂ <90% were all significant risk factors for death, whereas gender and initial respiratory rate were not (model 1). The association between pre-hospital management and 30-day mortality was not modified by further adjustment for the hospital of first admission (OR: 0.55, 95% CI: 0.31 to 0.98, P = 0.043). A short delay to hospital admission (less than one hour) was an independent risk factor for death, and only marginally affected the association with pre-hospital management (Table 5, model 2). In addition, results were not modified by adjustment for delay of ICU admission (OR: 0.52, 95% CI: 0.29 to 0.91, P = 0.022). Patients admitted to hospital after a more than one hour delay were more often intubated (53% vs. 32%, P < 0.001), ventilated (52% vs. 31%, P < 0.001) and more often received catecholamines (12% versus 8%, P = 0.048) than patients admitted more quickly.

The mortality rate at 72 h tended to be higher in SMUR patients than in non-SMUR patients (10.3% and 6.3%, respectively, P = 0.079). After adjustment for other prognostic factors, the impact of SMUR management on the risk of death at 72 h was not significant (OR = 0.77; 95% CI: 0.38 to 1.59; P = 0.49).

To our knowledge, this is the first large prospective study to examine the impact of SMUR vs. non-SMUR management of severe blunt trauma on mortality. This study revealed that SMUR management of severe blunt trauma in France was associated with an almost 50% reduction in 30-day mortality. SMUR management had no apparent benefit on 72-h mortality.

In accordance with previous studies, the present study confirmed the major effect on 30-day mortality of a low initial GCS score (<8) 10 11 , low systolic arterial blood pressure (<110 mmHg) 12 , high injury severity score 13 14 and initial hypoxemia 15 . As previously reported, respiratory rate was not an independent risk factor for death 16 .

In contrast to the beneficial impact of SMUR on 30-day mortality, the risk of death at 72 h was not significantly reduced in our study. The lack of any association may be partly due to the small number of events. The known bimodal distribution of the probability of death in blunt trauma patients may also explain this finding. Patients with very severe injury will die regardless of the type of pre-hospital management 17 . Another explanation may be due to our definition of severe blunt trauma. Indeed, only SMUR patients and patients admitted to a university ICU within 72 h were considered in the study. As a result, the number of early in-hospital deaths for non-SMUR patients may have been under-estimated since 62% of these patients were first admitted to general hospitals.

In France, pre-hospital care provided by fire brigades consists of oxygen administration, immobilization, dressing and cardio-pulmonary resuscitation with a bag valve in cases of cardiac arrest. Only EPs are allowed to perform SMUR management based on a large range of therapeutic strategies. These strategies, used after initial medical assessment, may include careful use of fluid administration, small-volume resuscitation strategies, continuous vasopressor infusion if fluid resuscitation fails to restore arterial pressure, tracheal intubation for mechanical ventilation after rapid sequence induction and continuous infusion of sedatives and analgesic agents, mannitol in cases of suspected clinical intracranial hypertension, and chest tube insertion. As pointed out by some authors, the main disadvantage of intensive pre-hospital management concerns patients with haemorrhagic shock in whom pre-hospital blood administration may delay hospital admission and hospital haemostatic treatment 2 .

Our findings on the beneficial impact of SMUR management on 30-day mortality should be discussed with regard to the results of the OPALS (Ontario Prehospital Advanced Life Support) Major Trauma Study. This study, which compared advanced life support (ALS) and basic life support (BLS) performed by paramedics, found no significant difference between groups 18 . The population characteristics and trauma severity in the two studies were very similar. In contrast, the major outcome in the OPALS study was survival to hospital discharge, which is not strictly comparable to the outcome of the FIRST study. As pre-hospital ALS was performed by trained paramedics in the OPALS study, the spectrum of intensive care therapies was more limited than in the FIRST study.

The FIRST study suggests that the French SMUR system leads to more vigorous on-scene management than ALS provided by trained paramedics. First, the management of haemodynamic status is more intensive. As reported by another recent French study 19 , an intravenous line was inserted in 99% of SMUR patients in the FIRST study compared to 63% of the ALS patients in the OPALS study 18 . According to French pre-hospital guidelines, a saline or colloid solution may be used for all patients on-scene with different blood pressure targets depending on whether patients have neurological injury or not. This early therapy in blunt trauma aims to limit excess fluid loading and later risks of multiple organ failure, acute lung injury and haematological complications 20 . In the FIRST study, around 47% of patients received colloid solution. The paramedics in the OPALS study did not have access to colloids and only 11.7% of the OPALS patients received intravenous fluid bolus therapy 18 .

The influence of broad and early intubation on mortality is controversial 4 . Some studies suggested that it may be beneficial 21 22 , whereas others did not 23 24 . Our study suggests that intensive pre-hospital airway management may explain the survival benefit for SMUR patients. In contrast to other countries, rapid sequence intubation for airway management is usual in France 25 . Nearly 50% of patients managed by EP were ventilated on-scene in the FIRST study, compared to under 7% intubated in the ALS group of the OPALS study 18 . In our study, indications for tracheal intubation and mechanical ventilation on-scene were not limited to patients with GCS scores <8 (97.5% ventilated patients) but extended to some patients with GCS scores between 8 and 13 (53.4% ventilated) and >13 (13.5% ventilated). This strategy, intended to increase the arterial oxygen level, is usually administered together with continuous infusion of sedatives and analgesic agents on-scene in order to decrease oxygen consumption 26 . Intensive airway management by EP possibly contributed to the reduction in acute respiratory distress syndrome (ARDS) and thus, to a decrease in 30-day mortality. On the basis of experimental studies 27 28 29 , the use of continuous infusion of norepinephrine is suggested for sedated patients with hemorrhagic shock in order to avoid excess volume loading. This strategy, in association with frequent use of mechanical ventilation, may contribute to a decreased risk of ARDS 30 and in-hospital mortality 31 .

The present study has the advantage of being prospective, based on a large sample of adult trauma patients consecutively recruited in university hospital ICUs located throughout France. Furthermore, the population was relatively homogenous since only patients with severe blunt trauma were included. However, our study also presents some limitations. The number of patients managed by fire brigades was low, limiting the statistical power of the study. The study was observational and did not allow any causal relationship to be established between the type of pre-hospital management and mortality. Clearly, the initial clinical status was more severe in SMUR than in non-SMUR patients, which reflects the efficiency of the French dispatching system. Differences in initial physiological status and injury severity between the two groups were taken into account in the outcome analysis, as well as the first admission hospital and the delay of first hospital admission or ICU admission. Our adjustment strategy did reveal the beneficial impact of SMUR, although this impact was not apparent in unadjusted analysis. Another limitation lies in our inability to control some potential confounding factors. For example, comorbidities were not recorded and information on time spent on the scene and transport time was available for only 76% and 55% of SMUR and non-SMUR patients, respectively. Furthermore, only patients directly or subsequently admitted to university hospitals were included. Thus, we cannot extrapolate our results to patients managed exclusively in general hospitals.

European pre-hospital management systems, particularly the French system, are controversial 2 . The main subject of debate is the increasing delay to hospital admission, diagnosis and actual salvage haemostasis induced by SAMU/SMUR intervention and on-scene management. In accordance with a recent study 32 , the FIRST study showed that a short delay to hospital admission of less than one hour was an independent risk factor for death among patients with severe blunt trauma. The lower rate of tracheal intubation, ventilation, and vasopressor administration in patients rapidly admitted to hospital strongly suggests that EP involvement in starting resuscitation care early before hospital admission could be beneficial for patients with severe blunt trauma, as reported in other studies 33 34 35 .

This observational study suggests that, despite delayed hospital admission, SMUR management was associated with lower 30-day mortality after blunt trauma. The French SAMU/SMUR emergency system comes at a high cost to society. This cost should be balanced against the number of life years gained for trauma patients who are often in the youngest age range of the population. Clearly, our results need to be confirmed in a randomized trial, but such a trial would be very difficult to organize in France.

• The FIRST study is an epidemiological study designed to prospectively describe the management and care of severe blunt trauma in France.

• Severe blunt trauma patients may have medical pre-hospital management or only management by fire brigades, according to the French pre-hospital health organization.

• Medical pre-hospital management is associated with a significant reduction in 30-day mortality. (OR: 0.55, 95% CI: 0.32 to 0.94, P = 0.03)

• Despite a longer out of hospital time, the organization SAMU/SMUR ameliorates delays to university hospital (trauma centre level 1) admission leading to a positive impact on survival. (OR: 0.62, 95% CI: 0.62 to 1.10, P = 0.1)

AIS: abbreviated injury scale; ALS: advanced life support; ARDS: acute respiratory distress syndrome; BLS: basic life support; CI: confidence interval 95%; EP: emergency physician; FIRST: French Intensive care Recorded Severe Trauma (Study Group); GCS: Glasgow Coma Scale; IQR: interquartile range; ISS: injury severity score; MAP: mean arterial pressure; ml: milliliter; mmHg: mercury millimeter; OR: odds ratio; SAMU: Service d'Aide Médicale Urgente; SD: standard deviation; SMUR: Service Mobile d'Urgences et de Réanimation; SpO_2: pulse oxymetry.

The authors declare that they have no competing interests.

YJM was involved in the study design, in the acquisition, analysis and interpretation of data, and wrote the first draft of the manuscript. GD, JC, DJ, MC and RF participated in the design of the study, in the acquisition of data and in the final revision of the manuscript. RB was involved in the initiation and design of the study, participated in the acquisition of data and contributed to the interpretation of data and final revision of the manuscript. BC participated in the design of the study and performed the statistical analysis. BKC was responsible for the logistic coordination of the study, was involved in the design of the study, in statistical analysis and interpretation of data and helped to draft the manuscript. FM initiated and coordinated the study and was involved at all steps of the study. All authors read and approved the manuscript.