DEBATE: Why emergency physicians should use etomidate as the preferred induction agent in patients with sepsis

No Comments

altThose concerned about the use of etomidate in septic patients seem to focus on two primary issues. First, that etomidate results in adrenocortical suppression. Second, that suppression is associated with increased mortality.



Those concerned about the use of etomidate in septic patients seem to focus on two primary issues. First, that etomidate results in adrenocortical suppression. Second, that suppression is associated with increased mortality.

These issues have been settled for three reasons. One, adrenal suppression occurs with etomidate, but is clinically insignificant. Two, etomidate is not associated with mortality any more than other induction agents. And finally, treating adrenocortical suppression in patients who received etomidate, doesn’t change outcomes.

Prompting renewal of this debate is the release of an article in November, 2012. “New article,” doesn’t mean “new data.” In fact, this article is a meta-analysis, a new look at old data. However, I challenge anyone to produce an agent for RSI, which rivals etomidate. I recall that Al Sacchetti made that very argument in the Annals of Emergency Medicine, July 2008. However, Ron Walls countered in that same issue. Sorry Al, I think Ron won that round. Please review his article for additional citations, which prove etomidate’s safety.


Several articles have been published refuting the assertion that the adrenal suppression associated with etomidate has any clinical relevance. I’ll mention just a few. First, The Effect of a Bolus Dose of Etomidate on Cortisol Levels, Mortality, and Health Services Utilization1. I’ll see your meta-analysis and raise you one systematic review. 3,083 titles were reviewed, which identified 20 applicable studies. One study showed an increase in ventilator days, ICU LOS and hospital LOS associated with etomidate use; no other studies found this.  None of the studies (including two in critically ill patients) nor in the author’s pooled estimate was any statistically significant difference in mortality reported.

Our second source is Single-Dose Etomidate Is Not Associated With Increased Mortality in ICU Patients With Sepsis2. In this retrospective review of 2014 patients (Dx: sepsis, severe sepsis or septic shock and intubated), 1,102 received etomidate, while 912 received some other induction agent. Although they report those treated with etomidate were more likely to receive corticosteroids, vasopressors were not used more in that group. Most importantly, hospital and ICU mortality rates were identical for both groups (37.2% v. 37.8% and 30.1% and 30.2%, respectively).

Next up is Etomidate use in severe sepsis and septic shock patients does not contribute to mortality3. In this multi-center study, a retrospective review was performed on intubated patients with severe sepsis or septic shock. Of the 230 patients, 173 received etomidate and 57 received an alternative or no induction agent at all. APACHE II scores were the same for both groups. The etomidate group mortality was 43.9%, with 45.6% in the other group.

Tekwani gives us etomidate lovers the 1-2 punch. In the first study4, 106 patients, requiring intubation due to sepsis, were evaluated prospectively.


Seventy-four patients received etomidate and the rest received propofol, a benzodiazepine, ketamine or no induction agent. The mortality rate was 38% (etomidate) v. 44%. The confidence intervals are wide, but we can safely conclude that etomidate was associated with at least a similar mortality as other induction agents.

Tekwani also brings us a prospective, DBRCT of ED patients with suspected sepsis, requiring intubation5. 122 patients were randomized to receive etomidate or midazolam. No statistical difference was reported for LOS or mortality.

Treating with corticosteroids doesn’t matter. In Payen’s Corticosteroid after etomidate in critically ill patients6, 99 ICU patients without septic shock were treated with a single dose of etomidate and randomized for a 42-hour hydrocortisone infusion versus placebo. Although norepinephrine requirements decreased more quickly in the hydrocortisone group, there were no differences in ICU LOS, ventilator time and 28-day mortality.

Does the Cochrane review by Chan change anything? It’s up to you. However, when looking at mortality, they narrowed their review to five randomized, controlled trials. Only one originally reported an association between etomidate and mortality. Their pooled data for reanalysis reflected a relative risk (RR) for mortality of 1.20 when etomidate was given to sepsis patients. By the time the number crunching ended to account for differences in the studies, the pooled relative risk for 28-day mortality was only 1.08, 8 hundredths above the line that would say the 28-day mortality is just as likely with or without etomidate.

Until a definitive trial says otherwise, etomidate is great. It’s time to end this debate.

Kevin Klauer, DO, EJD is Editor-in-chief of Emergency Physicians Monthly, CMO of Emergency Medicine Physicians, Vice Speaker of the ACEP Council. 

Leave A Reply