An Expiration Date for Etomidate?

Endotracheal intubation to facilitate invasive mechanical ventilation in critically ill patients is a high-risk procedure. For example, cardiovascular instability occurred in nearly 40% of intubations in critically ill patients included in the recent INTUBE study (International Observational Study to Understand the Impact and Best Practices of Airway Management in Critically Ill Patients), which in turn was associated with higher risks of mortality (1). Given this particularly high-risk nature of endotracheal intubation in critically ill patients, safety and risk mitigation must be prioritized. Consistent with this emphasis, the MACOCHA score (Mallampati score III/IV, Apnea syndrome [obstructive], Cervical spine limitation, Opening mouth 3 cm, Coma, Hypoxia, and Anesthesiologist nontrained) was developed to help identify patients at risk for difficult intubation (2), whereas several randomized controlled trials (RCTs) have assessed strategies implemented across various phases of the periintubation period to support safer airway management.

Critically ill patients pose physiological difficulty when undergoing endotracheal intubation. They are at high risk for cardiorespiratory collapse due to induction agents, conversion to positive-pressure ventilation, and limited physiological reserve rendered by underlying critical illness (3). As such, the pharmacologic agent of choice to induce anesthesia in critically ill patients is highly debated. Recently, ketamine and etomidate have gained popularity because of their more favorable hemodynamic profiles compared with the prototypical induction agent, propofol. However, particularly in patients with sepsis, there has been a longstanding concern of higher mortality rates in patients who receive etomidate versus other induction agents, possibly related to adrenal suppression that can occur after even a single dose of etomidate (4). Individual RCTs comparing the two agents have had mixed results. For example, two moderately sized RCTs comparing ketamine and etomidate found no difference in 28-day mortality (secondary outcome in both trials), yet one trial did show a higher 7-day survival rate in the ketamine arm (5) whereas the other demonstrated higher rates of adrenal suppression in the etomidate arm (6). Overall, pooled data in several meta-analyses point to the potential for increased mortality when etomidate is used to facilitate endotracheal intubation (7, 8).

In this issue of the Journal, Wunsch and colleagues (pp. 1243–1251) help advance our understanding of this clinically important issue with a sophisticated multicenter retrospective cohort study of more than 1.5 million critically ill adults who underwent endotracheal intubation in the United States. Using data from the Premier Healthcare Database, the investigators aimed to characterize rates of etomidate use for intubation and compare it versus ketamine with respect to association with mortality (9). To help address the important confounders that impact any comparison between etomidate and ketamine in an observational study, they adopted a target trial emulation approach with statistical analyses that incorporated hierarchical multivariable regression models and propensity score methods (10).

The study reports several key findings. Etomidate was administered to 44% of patients who received mechanical ventilation, with its use increasing over time. Variation in its use across the 878 hospitals in the cohort was considerable, ranging from approximately 2% to 85% of all intubations. Additionally, when compared against ketamine, etomidate was associated with a clinically and statistically significant increase in in-hospital mortality (adjusted odds ratio, 1.28; 95% confidence interval, 1.21–1.34). This association between etomidate and mortality was unchanged after adjusting for subsequent corticosteroid use but was not present in a subset analysis restricted to patients expected to receive corticosteroids for reasons other than shock (e.g., asthma).

Readers should also consider some limitations when interpreting these study findings. Although sophisticated analytical methods were applied, by and large, they can adjust for only measured confounders. Some important clinical information was not captured by the Premier Healthcare Database, including factors pertaining to drugs (e.g., dose, indication for use, location of administration) and patients (e.g., physiological data). Notably, even after risk adjustment with propensity-score matching, the etomidate and ketamine arms remained unbalanced with respect to some factors such as major surgery admission diagnosis and neuromuscular blocking agent used (rocuronium vs. succinylcholine). This imbalance suggests that patients who received ketamine were more likely to have been intubated in the operating room for surgery, whereas etomidate was more likely to have been used for nonsurgical admissions to critical care units. These two groups plausibly have very different mortality risks.

Several important questions arise from the data in this study and the recent related literature. First, although this study has limitations, the totality of evidence points to a reasonably consistent signal of increased mortality risk with etomidate. Combined with a mechanistic basis for this concern (i.e., adrenal suppression) and evidence that etomidate might not mitigate hypotension exposure compared with other agents like ketamine, the overall balance of risk versus benefit does not favor the use of etomidate (11).

Second, safe alternatives to etomidate exist. This study and other prior research focused on ketamine. Another possibility, which still requires further investigation, is the ultra–short-acting benzodiazepine remimazolam, which might provide the hemodynamic benefits of a benzodiazepine-based induction in a fast-acting and short-duration drug (12).

Third, mitigation of hemodynamic instability following endotracheal intubation in critically ill patients should involve a suite of interventions, with the induction agent (and dose) being just one component. Other key interventions include the proactive use of vasopressors and appropriate fluid loading to counterbalance the anticipated decrease in preload associated with positive-pressure ventilation (13, 14). For example, the Montpellier intubation protocol employs a bundle of interventions intended to reduce the risk of periintubation complications (14). In a pre-versus-post study, implementation of the Montpellier intubation protocol was associated with reduced rates of life-threatening (21% vs. 34%; P = 0.03) and other (9% vs. 21%) complications.

Fourth, the popularity of etomidate as an induction agent in critically ill patients is closely related to its use as a component of rapid sequence intubation (RSI). RSI involves administration of precalculated doses of an induction agent and paralytic drug in rapid succession, followed by immediate placement of an endotracheal tube before assisted ventilation (15). Although the predetermined dosing of an induction agent without titration can easily lead to hemodynamic instability (13), the stated benefit of RSI is to prevent pulmonary aspiration of gastric contents (16). Recent data bring this assumption into question. The PreVent (Preventing Hypoxemia with Manual Ventilation during Endotracheal Intubation) trial randomized 401 critically ill patients to receive mask ventilation or no ventilation during the period between induction and intubation (17). Patients receiving mask ventilation experienced a lower incidence of severe hypoxemia, whereas the rates of pulmonary aspiration did not differ significantly between the no-ventilation (2.5%) and ventilation (4%) arms. Although the trial was not designed to precisely detect between-group differences in the secondary outcome of pulmonary aspiration, these data offer support for avoiding RSI when intubating a critically ill patient.

Has etomidate reached its “best before” date when it comes to its routine use in endotracheal intubation? Wunsch and colleagues provide convincing findings that require attention and consideration. As we outline above, safer alternatives to etomidate exist, and airway management of critically ill patients should be considered to constitute a suite of interventions, with the induction agent (and dose) being just one component (13, 14). While we await the results of other RCTs (NCT06179485, NCT05277896), we recommend careful consideration of the choice of induction drug when proceeding with endotracheal intubation in critically ill patients.