Podcast: Play in new window

Subscribe: RSS

Date: January 17, 2026

Reference:  Casey et al. RSI Investigators and the Pragmatic Critical Care Research Group. Ketamine or Etomidate for Tracheal Intubation of Critically Ill Adults. NEJM. 2025 Dec

Guest Skeptic: Dr. Scott Weingart is an ED Intensivist from New York. He did fellowships in Trauma, Surgical Critical Care, and ECMO. Scott is best known for talking to himself about Resuscitation and Critical Care on the podcast EMCrit, which has been downloaded more than 50 million times.

Scott and I will both be presenting at Incrementum 2026 in Spain. 

Case:  You’re working an evening shift in a busy tertiary-care emergency department (ED). Emergency Medical Services (EMS) rolls in a 62-year-old woman with a history of hypertension and type 2 diabetes. She’s febrile at 39.2°C, tachycardic at 125 beats/min, hypotensive at 86/52 mm Hg despite 2 L of crystalloid and breathing 32/min on a non-rebreather with oxygen saturation of 88%. Chest X-ray shows a right‑lower‑lobe infiltrate; lactate is 5.6 mmol/L. She’s now on a norepinephrine infusion at 0.15 µg/kg/min and still looks exhausted and altered.

You decide she needs emergent rapid sequence intubation (RSI) for worsening work of breathing and impending respiratory failure. The respiratory therapist is at the bedside, the pharmacist has arrived with the RSI box, and your resident says: “For induction, should we go with ketamine because she’s septic and hypotensive, or etomidate because we’re worried about pushing her over the edge?”

Background: Intubating critically ill patients can be one of those high-stakes, high-adrenaline things we do often in emergency medicine (EM), but the physiology is stacked against us. These patients are often hypoxic, hypotensive, acidotic and catecholamine-depleted before we even reach for the laryngoscope. Emergency airway registries and multicentre cohorts consistently report serious peri-intubation complications (profound hypotension, hypoxemia, cardiac arrest, failed or difficult intubation) in roughly 10% to 20% of critically ill adult intubations in the ED and ICU [1].  Even a single episode of severe hypotension or hypoxemia during intubation is associated with increased mortality and organ failure in the ICU population [2].  So, the choice of induction agent seems to matter.

Etomidate became the darling of emergency RSI because it has a rapid onset, short duration, and relatively preserved hemodynamics compared with agents like thiopental or high‑dose propofol [3].  The flip side is adrenal suppression: a single dose transiently inhibits 11‑β hydroxylase and measurably blunts cortisol production for 24–72 hours. Observational studies and post‑hoc analyses in septic shock raised alarms that etomidate might increase mortality by worsening relative adrenal insufficiency, leading some guidelines and regulators to discourage or even remove etomidate in sepsis [4]. But those were mostly non-randomized data, and prior RCTs comparing etomidate with other agents were small and gave conflicting signals about mortality.

Ketamine, by contrast, is a dissociative NMDA antagonist with a very different vibe. It provides profound amnesia and analgesia, maintains airway reflexes to some degree, and has sympathomimetic properties that can increase heart rate and blood pressure by catecholamine release [5]. Those properties have made ketamine attractive in shocked patients where we’re worried that propofol or midazolam will “tank the pressure”. However, in catecholamine-depleted septic shock, ketamine’s direct myocardial depressant effects may become more apparent, and registry data suggest its hemodynamic advantage over etomidate is not as clear as many of us were taught on shift.  Old concerns that ketamine raises intracranial pressure have largely been debunked in modern neurocritical care literature, further widening its appeal. 

Other agents are still in the mix. Propofol remains widely used in operating theatre practice and some EDs because of its familiarity and ease of titration, but it predictably causes vasodilation and negative inotropy, making it a frequent offender in peri‑intubation hypotension among critically ill patients [6]. Benzodiazepines (midazolam) have a slower onset, a more variable effect and a longer half‑life, and when used as primary induction agents in shock, they’ve been associated with more hypotension and delirium compared with etomidate or ketamine. Fentanyl and other opioids are often layered on for analgesia or “blunting” the sympathetic surge, but they can also precipitate abrupt hypotension and apnea in the already fragile patient.

So for years we’ve been stuck between the theoretical adrenal toxicity of etomidate and the hoped‑for hemodynamic benefits of ketamine, without a large, definitive randomized trial in ED/ICU patients powered for patient‑important outcomes like mortality.

Clinical Question: In critically ill adults undergoing emergency tracheal intubation in the ED or ICU, which induction agent is better, ketamine or etomidate?

Reference:  Casey et al. RSI Investigators and the Pragmatic Critical Care Research Group. Ketamine or Etomidate for Tracheal Intubation of Critically Ill Adults. NEJM. 2025 Dec

  • Population: Critically ill adults (≥18 years) undergoing or planned to undergo tracheal intubation in an ED or ICU, where the treating clinician intended to use either ketamine or etomidate for induction, conducted at six EDs and eight ICUs in the US.
    • Excluded: Known or suspected pregnancy, prisoners. acute trauma as the primary indication for intubation, immediate “crash” intubations where there was no time for randomization, known or suspected allergy or contraindication to either ketamine or etomidate, situations where the treating clinician believed that either ketamine or etomidate was specifically required or contraindicated (strong clinician preference), or patients previously enrolled in the trial.
  • Intervention: The Ketamine group received 1.0–2.0 mg/kg given immediately before intubation.  Treating clinicians chose neuromuscular blocking agents, pre‑ and post‑intubation vasopressors, fluids, and other medications according to local practice.
  • Comparison: The Etomidate group received 0.2–0.3 mg/kg. Again, all other aspects of airway management and resuscitation were at the clinician’s discretion.
  • Outcomes:
    • Primary Outcome: In-hospital death by day 28 after randomization. 
    • Secondary Outcomes: A composite of “cardiovascular collapse” during the interval between induction of anesthesia and 2 minutes after tracheal intubation, defined as any of: Systolic BP <65 mm Hg, or A new vasopressor started or an increase in vasopressor dose, or Cardiac arrest. 
    • Exploratory Outcomes
  • Type of Study: A therapeutic, parallel‑group, multicenter randomized controlled trial

Authors’ Conclusions: “Among critically ill adults undergoing tracheal intubation, the use of ketamine to induce anesthesia did not result in a significantly lower incidence of in-hospital death by day 28 than etomidate.”

Quality Checklist for Randomized Clinical Trials:

  1. Did the study population include or focus on ED patients? Yes
  2. Were patients adequately randomized? Yes 
  3. Was the randomization process concealed? Yes
  4. Were patients analyzed in the groups to which they were randomized (intention‑to‑treat)? Yes
  5. Were patients recruited consecutively (no selection bias)? Unsure
  6. Were both groups similar with respect to prognostic factors? Yes
  7. Were all participants (patients, clinicians, outcome assessors) blinded? No
  8. Were all groups treated equally except for the intervention? Unsure
  9. Was follow‑up complete (≥80% in both groups)? Yes
  10. Were all patient‑important outcomes considered? Yes
  11. Was the treatment effect large enough and precise enough to be clinically significant? No
  12. Who funded the trial? The trial was funded by the Patient-Centred Outcomes Research Institute (PCORI), the US National Heart, Lung, and Blood Institute, and the US Department of Defence, among others; the funders had no role in study design, conduct, analysis, or manuscript preparation. 
  13. Did the authors declare any conflicts of interest? Several authors reported research funding and consulting fees from various organizations (NIH, DoD, Octapharma, monitoring companies), although none were manufacturers of ketamine or etomidate.

Results: They randomized 2,365 patients with a median age of 60 years, ~42% female, and ~56% intubated in the ED.  Nearly half had sepsis or septic shock, the median APACHE II score was 18, and about 22% were receiving vasopressors in the hour before intubation.  

Key Result: For critically ill adults undergoing emergency tracheal intubation in the ED/ICU, ketamine and etomidate produced similar 28-day in-hospital mortality, but ketamine was associated with more cardiovascular collapse and hypotension around the time of intubation.

  • Primary Outcome: In-hospital death by day 28:
    • Ketamine 28.1% vs Etomidate 29.1%
    • Risk difference adjusted for trial site: −0.8% (95% CI: −4.5% to 2.9%).
  • Secondary Outcomes:

  1. Open Label

    Open‑label: This was an open-label trial, with neither clinicians nor research staff being masked to assignment. The lack of blinding can bias both cointerventions (performance bias) and outcome measurement (detection bias), especially for outcomes that rely on clinician judgment. The primary outcome (death by day 28) may not be impacted by this, but the key secondary outcome, like cardiovascular collapse, is partly defined by starting or escalating vasopressors, which is inherently a clinician’s decision. If clinicians believed ketamine was safer or more pressure-friendly, they might delay vasopressors or, conversely, might treat etomidate patients more aggressively with prophylactic pressors, biasing the composite in either direction. Because group assignment was known and thresholds for vasopressor adjustment weren’t standardized or blinded, we can’t be entirely sure how much of the difference in collapse reflects actual hemodynamic physiology versus clinician behaviour influenced by expectations.

  2. Composite Secondary Endpoint: The key secondary outcome, cardiovascular collapse, comprises SBP <65 mm Hg, new or increased vasopressor use, and cardiac arrest. Composite outcomes can be misleading when their components differ substantially in clinical importance and frequency. This is true when the treatment effect is driven by the more frequent but less patient-oriented outcome (POO). In this trial, most of the difference in cardiovascular collapse appears driven by increased vasopressor use and moderate hypotension, while cardiac arrest (arguably the biggest POO) was rare and similar between groups. Patients probably care far more about avoiding arrest or death than about an extra push of norepinephrine. Although the composite difference is statistically significant, we should be skeptical of interpreting it as indicating that ketamine is clinically worse.
  3. Selection Bias & Exclusion Criteria: The trial excluded several important groups, such as trauma patients, pregnant patients, crash intubations with no time for randomization, and any case where the clinician believed ketamine or etomidate was specifically indicated or contraindicated. It was also conducted in 6 US academic centres, with no community, rural (critical access hospitals), or prehospital sites. We need to reflect on whether the study patients differ sufficiently from ours that the results may not apply. Here, the results may not generalize to trauma, obstetrics, prehospital RSI, or low-resource/community settings, where patients, logistics, and operator experience differ. In addition, allowing clinicians to exclude patients whenever they had a strong preference for one agent means that potentially the sickest or most fragile patients, the very ones we worry about most, may have been preferentially excluded from the trial. This type of discretion in enrolment is a form of selection bias and a threat to the trial’s internal and external validity. Consecutive patient enrollment, without subjective exclusion of patients, can mitigate this type of bias.
  4. Non-Protocolized Cointerventions and Performance Bias:This circles back to the trial being open-label. Apart from specifying the induction drug and its dose range, the protocol left all other aspects of care (choice and dose of neuromuscular blocker, timing and amount of fluid bolus, use and titration of vasopressors, steroid administration, and ventilator management) entirely to the clinician’s discretion. In an open-label trial, this introduces the potential for performance bias. Clinicians might manage ketamine and etomidate patients differently because of pre-existing beliefs about those drugs. For example, they may be more likely to give pre-emptive vasopressors before etomidate because of fear of adrenal suppression, or they might push higher ketamine doses in sicker-looking patients. Unequal cointerventions threaten the assumption that only the randomized treatment (ketamine or etomidate) differs between groups. This suggests that some observed hemodynamic differences (greater vasopressor escalation with ketamine) may partly reflect management style rather than pharmacology alone.
  5. Power: The trial was powered to detect a relatively large absolute difference in mortality (~5%) between groups, with the chosen sample size of 2,365 patients. The observed risk difference in 28-day mortality adjusted for trial site was −0.8% (95% CI, −5% to 2.9%). This 95% confidence interval is reasonably tight but still allows up to a 3% absolute difference in mortality, which some clinicians might consider clinically relevant in critically ill patients. Also, the study wasn’t primarily powered for subgroup analyses (patients in septic shock on vasopressors), yet multiple subgroup comparisons are presented. These subgroup analyses are underpowered and at risk of spurious findings unless pre-specified and strongly biologically plausible. They should be considered hypothesis-generating and not practice-changing. So, while the overall result strongly suggests no large mortality difference, modest benefits or harms in particular subgroups can’t be definitively excluded.

SGEM Bottom Line: For critically ill adults being intubated in the ED or ICU, there is no high-quality evidence from this trial to choose ketamine over etomidate to improve survival, and etomidate does not appear to carry an increased mortality.

Case Resolution: Back to your 62-year-old woman with septic shock who needs an emergent airway. Based on this trial, you can say that choosing ketamine instead of etomidate is very unlikely to change her chance of being alive in the hospital at day 28, but ketamine may give her a higher risk of BP dropping and needing more vasopressor during and immediately after the tube goes in. Etomidate does transiently suppress cortisol, but in this large RCT, it did not translate into higher mortality, even in the many patients with sepsis.  For this patient, already on norepinephrine with a MAP that is too low for comfort, a reasonable choice would be etomidate at 0.3 mg/kg, combined with pre-emptive hemodynamic optimization (fluids as appropriate, vasopressor up-titration, and, if you use them, a small push-dose vasopressor during induction). If you or your team are much more experienced with ketamine RSI, it’s also reasonable to use ketamine with an honest expectation that you may need aggressive vasopressor support and close BP monitoring. Either way, you would optimize preload and vasopressor support before induction, use a paralytic you’re comfortable with, plus meticulous pre-oxygenation and anticipate hypotension and have vasopressor boluses/infusion ready.

Clinical Application:

  • Mortality: This trial makes it very unlikely that one of these agents confers a substantial mortality advantage over the other in a general ED/ICU critically ill population. Your choice of induction agent should not be driven by fear that etomidate kills people via adrenal suppression or that ketamine is magically protective. 
  • Hemodynamics: In this trial, ketamine was associated with more peri-intubation hypotension and greater vasopressor escalation than etomidate. Don’t assume ketamine is always the hemodynamically safe option in catecholamine-depleted shock. You still need to aggressively resuscitate and support BP.
  • Patient & Context Factors: Etomidate may be a preferable default for patients in profound shock, where hypotension is the primary concern, and adrenal suppression is less concerning in light of this RCT. Ketamine might still be preferred in patients where you want bronchodilation (severe asthma), where etomidate is contraindicated, or when avoiding myoclonus is important (some neuro cases), remembering that ketamine’s hemodynamic effects are not uniformly benign in severe sepsis.
  • Local Practice: The evidence should be integrated with clinical expertise and local resources. If your team is more comfortable and faster with one agent, that may reasonably influence your choice, as long as you understand the trade-offs.
  • Not “One Size Fits All”: This trial covers ED/ICU adults, not trauma, pregnancy, prehospital RSI, or pediatrics. For those populations, we must still extrapolate cautiously from other data.

What Do You Tell the Family?  We need to give a medication to put her to sleep quickly and safely so we can place a breathing tube. There are two main medicines we use for this. We’re going to choose the drug that we think best fits her situation, and at the same time support her blood pressure with fluids and medications. We’ll be right at the bedside watching her closely the whole time.

Keener Question: Last week’s winner was Dr. Steven Steltz from NZ. He knew the sphygmomanometer (blood pressure cuff) had been generally credited to Scipione Riva-Rocci, an Italian physician, who invented it in 1896.

Other FOAMed:

Remember to be skeptical of anything you learn, even if you heard it on the Skeptics’ Guide to Emergency Medicine.

 References:

  1. April MD, Schauer SG, Nikolla DA, Casey JD, Semler MW, Ginde AA, Carlson JN, Long BJ, Brown CA 3rd. Association between multiple intubation attempts and complications during emergency department airway management: A national emergency airway registry study. Am J Emerg Med. 2024 Nov;85:202-207. doi: 10.1016/j.ajem.2024.09.014. Epub 2024 Sep 11. PMID: 39288499.
  2. Lin LC, April MD, Douin DJ, Winkle JM, Jenson WR, Rizzo JA, Schauer SG. Airway management in trauma patients: A seven-year review of emergency department intubations. Am J Emerg Med. 2026 Jan;99:306-312. doi: 10.1016/j.ajem.2025.10.023. Epub 2025 Oct 15. PMID: 41135359.
  3. Emergency Medicine Residents’ Association (EMRA) 2025. A Review of Rapid Sequence Intubation Pharmacologyhttps://www.emra.org/students/newsletter-articles/rapid-sequence-intubation-2025 Accessed Dec 18, 2025
  4. Sprung CL, Annane D, Keh D, Moreno R, Singer M, Freivogel K, Weiss YG, Benbenishty J, Kalenka A, Forst H, Laterre PF, Reinhart K, Cuthbertson BH, Payen D, Briegel J; CORTICUS Study Group. Hydrocortisone therapy for patients with septic shock. N Engl J Med. 2008 Jan 10;358(2):111-24. doi: 10.1056/NEJMoa071366. PMID: 18184957.
  5. George B, Joachim N. Evolving Techniques in RSI: Can the Choice of Induction Agent Matter in Securing a Definitive Airway in Emergency Settings? Indian J Crit Care Med. 2022 Jan;26(1):15-17. doi: 10.5005/jp-journals-10071-24100. PMID: 35110838; PMCID: PMC8783234.
  6. EMRA. Rapid Sequence Intubations https://www.emra.org/students/newsletter-articles/rapid-sequence-intubation-2025