To the best of our knowledge, this study is the first attempt to evaluate the outcome of trauma patients who have received etomidate during emergent intubation prior to hospital admission at such a large scale. The ETO and NON-ETO group were comparable in demographic and pre-hospital characteristics. We found no influence of a single dose of etomidate on mortality in trauma patients. We did not find a remarkable difference in morbidity between the two groups, but healthcare resource utilization parameters were prolonged in the ETO group. However, data for organ failures and sepsis were incomplete.
Due to a lack of interfaces between pre-hospital admission and in-hospital patient data documentation, it is difficult to obtain and to merge data from both treatment phases for a large number of patients. The consolidation of the ADAC Air Rescue and the TR-DGU databases gave us the opportunity to generate outcome data for interventions performed in the field prior to hospital admission in a large multicentre approach.
Mortality
To our knowledge, only four publications have analyzed the mortality after induction of anesthesia in trauma patients using etomidate [
21‐
24]. The largest study we found was a recent retrospective investigation in 968 adult trauma patients, whereof 526 patients were induced in the Emergency Department with etomidate and 442 with ketamine [
24]. Hospital mortality was 20.4% for ketamine compared with 17.3% for etomidate (OR: 1.41; 95% CI: 0.92–2.16). In the 4-year study period the use of ketamine compared with etomidate was not associated with an improvement in hospital mortality.
Hinkewich and colleagues [
21] investigated 308 trauma patients, in whom induction of anesthesia was performed at a single institution. They observed a remarkable difference in the 28-day mortality of 18.7% for patients receiving etomidate (
n = 107) and 11.1% for all other anesthetic agents (
n = 201). However, after multivariate analysis, etomidate could not be identified as an independent predictor of mortality (
p = 0.11).
A retrospective chart review [
22] compared a period of liberal versus a period of limited use of etomidate. Four hundred forty patients were included in the liberal period (58.9% received etomidate) and 882 in the limited period (23.2% received etomidate). The inpatient mortality was similar in the both periods (30% versus 29%;
p = 0.848).
Jabre and colleagues [
23] investigated 469 patients, who received etomidate or ketamine for emergent intubation during a randomized, controlled, single-blind trial. The 28-day mortality as a secondary endpoint was 35% for the etomidate group and 31% for the ketamine group (
p = 0.36). One hundred four out of the 469 patients had a history of trauma. The 28-day mortality in this subgroup was 30% in the ketamine and 26% in the etomidate group, without revealing a statistical significance. The interpretation of these findings is limited as the cohort is small and there is no further information on ISS, gender or age available.
Morbidity and healthcare resource utilization
ICU days, ventilator days and hospital stay decreased in general over the past few years. This was also observed in our study population. Etomidate was given predominantly in the first half of the study period and healthcare resource utilization parameters were prolonged in the ETO group. However, after adjusting HOS-LOS was still significantly prolonged in the ETO group.
A trial with a small sample size (
n = 30) found no significant differences in ICU-LOS, HOS-LOS and ventilator-free days [
21]. Equally results without remarkable differences in ICU-LOS and HOS-LOS were reported by comparing a liberal versus a limited period of etomidate application [
22]. Upchurch and colleagues did not reveal any distinctions in ICU-LOS and ventilator-free days between patients induced with etomidate or ketamine [
24].
In a prospective trial investigating the administration of hypertonic saline administration prior to hospital admission 35 of the 94 patients (37%) received etomidate for emergent intubation in the field [
25]. The development of Acute Respiratory Distress Syndrome (ARDS) was defined as primary outcome and was close to reveal statistical significance (
p = 0.06). In a further stepwise multivariate regression analysis the most severely injured patients were selected. In this small subgroup etomidate was an independent risk factor for the development of ARDS (
p = 0.02) and Multi Organ Dysfunction Syndrome (MODS) (p = 0.02). In addition, ICU-LOS (p = 0.02), HOS-LOS (p = 0.02), and ICU-LOV (
p = 0.04) were significant longer in the etomidate group.
Another subgroup analysis out of a previously randomized, controlled trial revealed a greater risk for hospital-acquired pneumonia (HAP) in trauma patients who had received etomidate [
26]. One hundred forty nine patients were originally included in the HYPOLITE trial investigating the use of hydrocortisone in patients with severe trauma. All of these patients were part in the following analysis and had received mechanical ventilation for at least 48 h. In a multivariate analysis, etomidate was an independent risk factor for hospital-acquired pneumonia (
p = 0.016). ICU-LOV did not show a difference between the groups.
Even a lower incidence of cardiovascular failure, defined as any administration of epinephrine/norepinephrine or dopamine use in the clinical course including Emergency Department, was observed in our study for the ETO group. This may reflect the minimal hemodynamic effects of etomidate compared to other induction agents. Even the catecholamine-mediated stabilizing effect of ketamine on the cardiovascular system can fail in patients who are catecholamine depleted [
27].
A recent Cochrane review, investigating the effects of a single-dose of etomidate in critically ill patients on mortality, morbidity or healthcare resource utilization, could not reveal any clear answers [
8]. However, an increased risk of adrenal gland dysfunction and multi-organ system dysfunction by a small amount was observed. This review was able to include seven studies for the analysis, of which only two studies included traumatized patients. This again demonstrates how limited the available data in this patient population is.
A main limitation for our study is its retrospective nature and the choice of anesthetics at discretion of the emergency physician, not according to a standardized protocol. Data is further limited as exact dosages of induction agents have not been documented in our electronical HEMS database. The NON-ETO group includes patients receiving several different anesthetic agents (thiopental, propofol, ketamine, midazolam). This again reflects the selection of anesthetics at discretion of the emergency physician. All of these drugs are routinely used in pre-hospital trauma anesthesia in Germany. Furthermore, when the two respective databases were merged, out of 5301 patients in the ADAC Air Rescue Service database only 1910 could be assigned to a case in the TR-DGU. A main reason for this is the number of hospitals participating in the TR-DGU over the study period. As shown in Fig.
2 we were able to increase the patient recruitment in each year over the study period due to more hospitals participating in TR-DGU. The documentation of data was incomplete and inconsistent with respect to morbidity (organ failure, sepsis). Also, the cause of death (e.g. bleeding, brain death, organ failure) is not documented in the TR-DGU.