Is It Safe to Use Etomidate When Intubating Patients With Sepsis or Septic Shock?

By Andrew M. Luks, MD, Pulmonary and Critical Care Medicine, University of Washington, Seattle, is Associate Editor for Critical Care Alert.

Dr. Luks reports no financial relationship to this field of study.

Synopsis: This retrospective cohort study demonstrated that single-dose etomidate administration during rapid-sequence intubation was not associated with adverse outcomes in patients with sepsis or septic shock.

Source: Dmello D, et al. Outcomes of etomidate in severe sepsis and septic shock. Chest 2010;138:1327-1332.

Due to its lack of cardiovascular side effects, the short-acting non-barbiturate sedative, etomidate, has been one of the primary agents used to sedate hypotensive patients during rapid-sequence intubation (RSI). While use of this valuable agent is associated with transient, reversible suppression of adrenal function following bolus administration, the effect on patient outcomes resulting from this problem is not clear. Dmello and colleagues sought to examine this question further and determine whether use of etomidate during RSI was associated with increased mortality and other adverse outcomes in a subset of critically ill patients.

To investigate this question, the authors conducted a retrospective cohort study of patients in a multidisciplinary ICU of a single institution. They identified 224 consecutive patients with sepsis or septic shock — defined using consensus criteria developed by the American College of Chest Physicians and Society of Critical Care Medicine — requiring invasive mechanical ventilation within 48 hours of admission. Selected patients were then divided into two groups, those who underwent RSI with etomidate plus a paralytic agent and those who did not undergo RSI and typically received midazolam and fentanyl. The original decision as to whether to use RSI or another technique for intubation was not based on pre-specified criteria and was, instead, at the discretion of the intubating physician. Among data gathered on all patients, they examined whether patients had relative adrenal insufficiency, defined as a random baseline cortisol level < 15 µg/dL or a less than 9 µg/dL response to 250 µg of adrenocorticotropic hormone (ACTH). Treatment protocols following intubation appeared to be at the discretion of the attending physician. Multivariate logistic regression was used to adjust for the effects of age, sex, APACHE II scores, and corticosteroid use, and compare differences in mortality, vasopressor use, ICU length of stay (LOS), and ventilator days between the two groups.

Etomidate was used as part of RSI in 113 patients of the 224 patients included in the study. The average age and APACHE II scores were similar between the two groups of patients. Of those patients who underwent testing, 24% of those in the etomidate group and 22% of those in the non-etomidate group met criteria for relative adrenal insufficiency. Corticosteroids were used in 42% of the etomidate patients and 22% of the non-etomidate patients, although the authors did not comment as to whether there were explicit criteria used to determine who received corticosteroids in this particular ICU. The relative risk of steroid use was 1.34 (95% confidence interval [CI], 1.11-1.61; P = 0.003). The relative risk of mortality in the etomidate group was 0.92 (95% CI, 0.74-1.14; P = 0.51), while the relative risk of vasopressor use was 1.16 (95% CI, 0.9-1.51; P = 0.31). There were no significant differences in ICU-LOS (14 vs 12 days; P = 0.31) or ventilator days (11 vs 8 days; P = 0.13).


Intubation and initiation of invasive mechanical ventilation is an all too common event in the ICU and a significant number of patients must undergo this procedure while in the midst of septic shock. Given how frequently we use etomidate instead of propofol in such situations because of differences in their hemodynamic profiles, it would be important to know whether etomidate's transient effects on the adrenal axis are associated with worse patient outcomes. Were mortality to be increased as a result of adrenal suppression, for example, it would behoove us to consider alternative means for inducing patients during rapid sequence intubation.

Unfortunately, the trial by Dmello and colleagues does not provide a sufficient answer to this question. While the trial did not reveal any significant differences in mortality or vasopressor use, there were significant methodological issues in the study that make it difficult to draw any strong conclusions. Critically ill patients are often so complex that it is difficult to control for enough factors and isolate the effect of a particular intervention on big outcomes, even with a prospective randomized design. Dmello and colleagues used a retrospective design and it does not appear that any specific protocols were used to guide care of the patients. Most decisions were left to the treating physician and not based on pre-specified criteria or guidelines, including the decision whether or not to use rapid-sequence intubation — the key factor separating the two groups in the study. Given these problems, there are likely too many factors that varied in the care of the patients as to make it difficult to attribute the observed outcomes, or lack thereof, to the use of one particular medication at the time of intubation.

Had this trial revealed that mortality was significantly increased with etomidate, we might be in the slightly tricky situation of having to decide whether to change our current practice based on results from a study with methodological problems. With no glaring differences in outcomes, however, it is likely acceptable to continue using etomidate until adequately controlled prospective studies provide further insight into this question.