Abstract
Objective
Specific sedative hypnotic agents, administered to facilitate endotracheal intubation (ETI) in critically ill adults, may lead to adverse outcomes such as peri-intubation cardiovascular collapse. However, little is known from systematic investigations of the impact these individual agents have on cardiovascular function or other clinical outcomes.
Data sources
MEDLINE, Embase, CENTRAL, ClinicalTrials.gov, Scopus and Web of science databases.
Study selection
We conducted a systematic search for randomized and non-randomized studies that evaluated adult (≥18 years) critically ill patients who were sedated to facilitate ETI with ketamine, propofol, ketamine/propofol, etomidate, or a benzodiazepine and who had data on peri-intubation hemodynamics and at least one other outcome involving acute kidney injury, delirium, opioid use, intubation difficulty, sequential organ failure assessment, length of stay, or mortality. Eighty-five studies were identified for eligibility assessment with 23 included in the analysis.
Data extraction
Two reviewers independently screened articles, extracted data from selected articles, and assessed risk of bias using ROBINS-I for observational studies and revised Cochrane Risk of Bias tool for randomized controlled trials.
Data synthesis
Acute cardiovascular dysfunction (peri-intubation hemodynamic instability and/or cardiac arrest) was similar between etomidate and ketamine with more events seen when propofol versus non-propofol sedation was administered. However, exploratory meta-analysis demonstrated no difference between etomidate and ketamine (OR 1.05 [95%CI 0.60-1.84]) or between etomidate and propofol (OR 0.91 [95%CI 0.33-2.46]). Compared to ketamine, etomidate demonstrated lower survival to hospital discharge in the included studies in exploratory meta-analysis OR 0.76 (95%CI 0.62-0.92). Limited data existed for other outcomes with no discernible differences between sedative agents.
Conclusions
Acute cardiovascular dysfunction was more common when propofol, as compared to non-propofol sedation, was administered, although not statistically significant in exploratory meta-analysis. In addition, etomidate conferred lower survival to hospital discharge versus non-etomidate sedation, which was confirmed in exploratory meta-analysis of etomidate versus ketamine.
Introduction
Endotracheal intubation (ETI) is commonly required in critically ill adults, either due to acute respiratory failure or to support essential clinical interventions such as endoscopy where airway protection is needed.1–3 Several studies have demonstrated that airway management performed in critically ill adults is associated with higher rates of physiologic complications compared to airway management conducted in a more controlled setting, namely the operating room.4–8 Furthermore, studies have demonstrated that physiologic complications during airway management are associated with adverse clinical outcomes, such increased length of stay and mortality.9,10 Research into potentially modifiable risk factors for two of the most common physiologic complications, peri-intubation hypotension and hypoxemia, has garnered increasing interest over the last decade.
Russotto and colleagues in their multicenter prospective observational study across 29 countries evaluating ETI practices demonstrated that older age, history of heart failure, hematologic malignancy, lower systolic blood pressure (SBP), administration of fluid bolus before ETI, higher heart rate, cardiovascular instability as the reason for ETI, high risk of pulmonary aspiration, use of video laryngoscopy, and lower SpO2/FiO2 ratio were associated with major adverse peri-intubation events. 5 Another multicenter prospective observational study conducted by Smischney and colleagues derived and validated a scoring system to quantify the risk of peri-intubation hypotension and found many of the same variables independently associated with peri-intubation hypotension as Russotto and colleagues. However, these authors identified etomidate as protective in the pathway to peri-intubation hypotension. 4
The aforementioned studies did evaluate whether sedative hypnotic agents were associated with adverse events during ETI, namely the development of peri-intubation hypotension, one of the most common complications of ETI. However, they and many others have not specifically focused on the impact that ETI sedative hypnotic agents have on both short- and long-term outcomes in the critically ill. Although several recent studies have explored ETI complications and their impact on long-term outcomes, such as length of stay and mortality, limited data exist on the impact sedative hypnotic agents, when given for ETI, have on both short- and long-term outcomes in the critically ill. Numerous studies have explored etomidate's association with mortality when given for ETI in critically ill adults with equivocal results.11–14 Others that have examined the impact of sedative hypnotic selection for ETI on such outcomes may have been under-powered to detect significant differences in all relevant outcomes besides just mortality.13,15
This systematic review aimed to assess the existing clinical evidence of the effects that sedative hypnotic agents used during ETI have on both immediate (eg, within 24 h of administration) and delayed (length of stay and mortality) outcomes in critically ill adults.
Methods
This systematic review was conducted at Mayo Clinic, Rochester using the resources of the Center for the Science of Health Care Delivery and the Plummer Library. The current study adheres to the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) 2020 guidelines. 16 The protocol was registered a priori in the PROSPERO database (CRD42022304366). Due to the established limited sample sizes of eligible studies, our team chose to present this study as a systematic review with exploratory meta-analysis results. This approach was taken as meta-analyses with very small sample sizes (eg, 3-5 studies) may lead to conclusions drawn from over-represented studies (ie, studies with very large effect sizes) and are more susceptible to greater variance on effect size estimations, likely to be impacted by publication bias, and demonstrate unreliable estimations of heterogeneity. 17 Due to the high heterogeneity in interventions and study designs among our included studies, we deemed the sample sizes insufficient to perform individual between-intervention comparisons or network meta-analysis.
Eligibility Criteria
Eligible studies included randomized controlled trials, prospective and retrospective longitudinal studies including cohort studies, and case series that reported using either ketamine, propofol, etomidate, ketamine/propofol admixture (“ketofol”) or benzodiazepines for sedation during ETI in critically ill adults. Studies were excluded if they did not report on granular peri-intubation hemodynamic data and at least one other outcome (acute kidney injury, delirium, opioid use, intubation difficulty, sequential organ failure assessment [SOFA], length of stay, or mortality). The rationale for this was to explore the effects sedative hypnotic agents have on hemodynamic and non-hemodynamic clinical outcomes, akin to prior studies.11,12,15 Studies were also excluded if they reported on pediatric patients (age <18 years), animals, case reports, reviews, editorials, letters, viewpoints, and book chapters. Studies reporting on non-critically ill patients, such as from the operating room, were excluded. Studies published prior to 2001 were excluded due to temporal shifts in airway management. No restrictions were placed on language of the published articles.
Data Sources
An experienced medical librarian (EAKV) searched electronic health and science literature databases and clinical trial registers on 9/29/2021, with an updated search performed from 9/20/2021 to 11/30/2022 and again from 11/20/2022 to 02/29/2024. The names and dates of the databases and register are provided in
Article Selection and Data Extraction
Article title and abstracts were screened by two independent reviewers (NJS, RGP) based on the previously described inclusion and exclusion criteria. Any disagreements were adjudicated by a third independent reviewer (GW). Candidate articles included in the full text review were randomly assigned to the two independent reviewers (NJS, RGP) with disagreements adjudicated by the third independent reviewer (GW). If consensus was not reached by the third independent reviewer during each stage of the review, then consultation with a senior investigator (AKK) was conducted to arrive at a conclusion. Article screening was performed using Covidence software (Melbourne, Australia).
Data extraction from the included and agreed upon final articles was performed by two independent abstractors (ARP, BB) using a structured tool with disagreements adjudicated by a third independent investigator (AKK) in consultation with a senior investigator (GW) if necessary. Data abstracted included details regarding publication information, study design, patient characteristics, sedative hypnotic agent used with dosing, peri-intubation hemodynamics within 60 min and outcomes of peri-intubation cardiovascular instability (defined as mean arterial pressure < map> < 65mm Hg, SBP <90mm Hg, need for vasoactive medications, or increase in vasoactive infusion), peri-intubation cardiovascular collapse or cardiac arrest within 60 min, intubation difficulty score, oral morphine equivalents within 24 h of ETI, acute kidney injury within 24 h of ETI, intensive care unit (ICU) and hospital mortality, change in sequential organ failure assessment scores within 24 h of ETI, ICU free days or length of ICU stay, and delirium within 24 h of ETI. We choose the hemodynamic outcomes as primary and the other clinical outcomes as secondary.
Risk of Bias Assessment
For non-randomized studies, the risk of bias was assessed using the ROBINS-I tool. 18 For the randomized trials, the risk of bias was assessed using the revised risk of bias tool for randomized trials. 19 For both types of study designs, the tools were applied by two independent assessors (ARP, BB) and disagreements were adjudicated by a senior investigator (GW). Summary statistics for the risk of bias information was performed using R version 4.4.1 (R Core Team, R Foundation for Statistical Computing, Vienna, Austria, 2024) and the robvis version 0.3.0 package. 20
Data Analysis
Data from both randomized trials and non-randomized observational studies are summarized and reported. Exact p-values and corresponding effect size measures (eg, odds ratios, mean differences, etc, where applicable) of included studies are organized, summarized in descriptive tables, and reported narratively in the results sections. As previously outlined, meta-analyses are reported as exploratory owing to constraints in the published literature (significant heterogeneity and limited numbers). We performed exploratory meta-analyses on the cardiovascular outcome and presented forest plots and funnel plots for visualization. For these exploratory analyses, we categorized the sedative interventions as: (1) etomidate, (2) propofol, (3) ketamine, (4) midazolam, and (5) other interventions. Studies with treatment arms using ketamine/propofol admixture (“ketofol”) were considered under the propofol category for analysis. We also combined events of (1) cardiovascular collapse including cardiac arrest and (2) hypotension and/or hemodynamic instability into one composite, binary cardiovascular outcome (ie, acute cardiovascular dysfunction)
Results
Included Studies
The initial search identified 3083 studies with an additional 734 studies retrieved in the updated searches. Following removal of 1441 duplicates and 2291 irrelevant studies, 85 full-text studies were assessed for eligibility with 23 (27.1%) studies included in the analysis.4,13,15,21–40 Please refer to Figure 1 for further details.

Flow diagram.
Study and Patient Characteristics
Of the 23 included studies, 6 were randomized trials with 17 non-randomized studies. In total, there were 8590 participants across the 23 studies with 13 performed in the United States, 1 performed in United States, Spain, and Greece, 28 3 in France, 2 in Egypt, and the remaining 3 in Saudi Arabia, Turkey, and United Kingdom with one study being conducted in several countries 35 with dates spanning 2009 to 2023. In total, there are 8590 participants across the 23 included studies. The majority of participants were male 5428 (63%) with a pooled averaged mean age of 56.4 years and an average weight of 81.6 kg. The most common indications for ETI in most studies were respiratory failure, shock, airway protection, and neurologic deterioration. The most common sedatives administered for ETI were etomidate and ketamine. Characteristics of study participants are detailed in Table 1.
Study Characteristics.
Note on Alday and VanBerkel - the above demographics are for the propensity-matched subset of participants.
Abbreviations: ACE: angiotensin-converting enzyme; APACHE: acute physiology and chronic health evaluation; ARDS: acute respiratory distress syndrome; ARF: acute respiratory failure; BMI: body mass index; CAD: coronary artery disease; CHF: congestive heart failure; CKD: chronic kidney disease; COPD: chronic obstructive pulmonary disease; CVC: cardiovasculare collapse; DM: diabetes mellitus; GI: gastrointestinal; HIV: human immunodeficiency virus; HTN: hypertension; kg: kilogram; m: meter; mg: milligram; mL: milliliter; NR: not reported; OHS: obesity hypoventilation syndrome; q SOFA: quick sequential organ failure assessment; SAPS: simplified acute physiology score; SD: standard deviation; SOFA: sequential organ failure assessment; VTE: venous thromboembolism
Data reported as mean [±standard deviation].
Data reported as median (interquartile range).
Patient on cardiac medications before intubation (24 h) [Alpha blocker, beta blocker, Angiotensin-converting enzyme (ACE) inhibitor, calcium channel blocker, nitrates, midodrine, anti-arrhythmic]
Patient on vasopressors before intubation (24 h)
Outcomes (Studies Reporting Outcomes Between Sedatives)
Acute Cardiovascular Dysfunction
Four studies did not report on granular hemodynamic data within 60 min of ETI.23,28,31,36 Of the studies that did report on granular hemodynamic data and available for analysis, six were randomized trials13,15,22,24,25,27 and 13 were non-randomized studies.4,21,26,29,30,32–35,37–40 Of the six randomized trials, only one demonstrated a significant difference in the rate of acute cardiovascular dysfunction between etomidate and ketamine (etomidate 69 [17.4%] vs ketamine 99 [25.1%], difference −7.6 [95% CI −13, −2]). 22 Five non-randomized studies demonstrated a significant difference in the rate of acute cardiovascular dysfunction between sedative hypnotic agents21,33,37,38,40 with two demonstrating a significant difference between etomidate and ketamine (Van Berkel et al 33 : etomidate 84 [73%] vs ketamine 59 [51.3%], p = 0.001; Kunkel et al 37 : etomidate 25 [9.6%] vs ketamine 2 [4.2%], p = 0.028), two demonstrating a significant difference between propofol and non-propofol sedation (Elsharkawy et al 21 : p = 0.007 for propofol and acute cardiovascular dysfunction; Gamage et al 40 : propofol 36 [71%] vs non-propofol 3 [3%], p = 0.000), and one study 38 demonstrating a significantly higher rate of acute cardiovascular dysfunction in patients receiving propofol or ketamine compared to etomidate (propofol 35 [16%] vs ketamine 26 [18%] vs etomidate 51 [6%], p=<0.001).
Delirium (24hr)
Three studies reported delirium or mental status changes post-intubation, two randomized trials13,25 and one case series. 26 Both randomized trials demonstrated no significant difference between etomidate and ketamine/propofol admixture (“ketofol”) (7 [13%] vs 4 [6%], p = 0.233; adjusted for benzodiazepine use, p = 0.177) 13 and between etomidate and ketamine (etomidate 2 [16%] vs ketamine 5 [50%], p = 0.172). 25 The case series evaluated ketamine/propofol admixture (“ketofol”) in six patients for which 3 (50%) developed new-onset delirium 24 h post-intubation. 26
SOFA (24hr)
Four studies reported on SOFA scores post-intubation, three randomized trials15,22,25 and one non-randomized study. 33 The randomized trials did not show any significant difference in SOFA scores at 24 h between etomidate and ketamine (Jaber et al ketamine 9.6 ± 3.9 vs etomidate 10.3 ± 3.7, p = 0.056; Matchett et al ketamine 10.9 ± 4.5 vs etomidate 10.7 ± 4.6, difference −0.2 [95% CI −0.9, 0.4]; Cinar et al ketamine 8 vs etomidate 7).15,22,25 Likewise, the non-randomized study did not show a difference between etomidate and ketamine (ketamine 6.2 ± 2.9 vs etomidate 6.4 ± 3.1, p = 0.58). 33
ICU Length of Stay
Six studies reported ICU length of stay, two randomized trials22,25 and four non-randomized studies.29,32–34 The randomized trials did not show a difference between etomidate and ketamine (Cinar et al etomidate 12.3 ± 22 vs ketamine 4.3 ± 27, p = 0.368; Matchett et al etomidate 8 [4-16] vs ketamine 9 [5-14], p = 0.302). Similarly, the non-randomized studies did not demonstrate a difference in ICU length of stay with three studies evaluating etomidate to all comers (Alday et al etomidate 7 [3.9-13] vs non-etomidate 9 [5-15], p = 0.19; Jung et al etomidate 12 [6-22] vs non-etomidate 9 [4-13], p = 0.06; Perier et al etomidate 6 [3-8] vs non-etomidate 6 [4-14], p = 0.24) and one evaluating etomidate to ketamine (Van Berkel et al etomidate 9 [5-14] vs ketamine 8 [4-17], p = 0.74).
New Vasopressor Requirement Within 24 h of ETI
Out of the seven studies reporting on post-intubation vasopressor administration,4,22,26,28,34,35,40 only three studies22,28,34 reported on post-intubation vasopressor administration per sedative agent received of which one was a clinical trial. 22 The clinical trial demonstrated a significant reduction in post-intubation vasopressors between etomidate and ketamine (Matchett et al etomidate 65 [18.8%] vs ketamine 92 [26%], difference −7.1 [95% CI −13, −1]) while the two non-randomized studies did not show any significant difference in post-intubation vasopressor need between etomidate and non-etomidate (Perier at al. etomidate 14 [25.9%] vs non-etomidate 7 [16.3%], p = 0.32; Leou et al etomidate 94 [96%] vs non-etomidate 73 [100%], p = 0.78).
Survival to Hospital Discharge
In total, 12 studies reported on survival to discharge per sedative hypnotic agent received13,15,22,23,25,27–30,33,34,38 with one study reporting overall survival to discharge. 36 Of the 12 studies reporting in-hospital survival, five studies13,15,22,25,27 were clinical trials with the remaining being non-randomized studies.23,28–30,33,34,38 None of the randomized trials showed a significant difference in survival to discharge between etomidate and other sedatives. Two non-randomized trials28,38 demonstrated a significant difference in survival to hospital discharge between etomidate and other sedatives (Kuza et al etomidate 66.1 [74.5%], ketamine 122 [82.9%], propofol 222 [98.2%], p=<0.001; Leou et al etomidate 21 [21.4%], non-etomidate 32 [43.8%], p = 0.003) with etomidate demonstrating the lowest in-hospital survival while the other non-randomized trials did not demonstrate a significant difference in survival to hospital discharge.
Findings in Exploratory meta-Analyses
Our exploratory meta-analysis showed that there was no significant difference in the occurrence of acute cardiovascular dysfunction events in either the etomidate versus propofol or the etomidate versus ketamine comparisons. Similarly, we observed a trend toward decreased in-hospital length of stay but failed to obtain sufficient studies for detection of the difference in ICU length of stays. We did however detect a significant difference in survival to hospital discharge between etomidate and ketamine with ketamine demonstrating higher survival to hospital discharge. We deliberately did not report any summary statistics in this section to avoid creating over-simplified presentations and overconfidence of our analytical estimates given the exploratory nature of the meta-analysis. Several randomized controlled trials and non-randomized observational studies were excluded from the forest plots because data on the specified outcomes of interest were not available for demonstration. The absence of these data highlights a limitation in the reporting of certain trials, which could affect the robustness and generalizability of the findings in the analysis. Therefore, we reiterate that the results of this meta-analysis are purely exploratory, and the study is designed primarily as a systematic review.” Please refer to Figures 2–4 for all analytical statistics and visualizations.

Acute cardiovascular dysfunction exploratory meta-analyses (etomidate vs. ketamine). Several randomized controlled trials and non-randomized observational studies were excluded from the forest plots because data on the specified outcomes of interest were not available for demonstration. The absence of these data highlights a limitation in the reporting of certain trials, which could affect the robustness and generalizability of the findings in the analysis. Therefore, we reiterate that the results of this meta-analysis are purely exploratory, and the study is designed primarily as a systematic review.

Acute cardiovascular dysfunction exploratory meta-analyses (etomidate vs. propofol with P/K-propofol/ketamine admixture included). Several randomized controlled trials and non-randomized observational studies were excluded from the forest plots because data on the specified outcomes of interest were not available for demonstration. The absence of these data highlights a limitation in the reporting of certain trials, which could affect the robustness and generalizability of the findings in the analysis. Therefore, we reiterate that the results of this meta-analysis are purely exploratory, and the study is designed primarily as a systematic review.

Survival to hospital discharge exploratory meta-analyses (etomidate vs. ketamine). Several randomized controlled trials and non-randomized observational studies were excluded from the forest plots because data on the specified outcomes of interest were not available for demonstration. The absence of these data highlights a limitation in the reporting of certain trials, which could affect the robustness and generalizability of the findings in the analysis. Therefore, we reiterate that the results of this meta-analysis are purely exploratory, and the study is designed primarily as a systematic review.
Risk Bias Assessment
Six out of the 23 studies were randomized controlled trials13,15,22,24,25,27 in which the revised Cochrane Risk of Bias tool was used to assess risk and for the remaining non-randomized observational studies, ROBINS-I was used to assess risk of bias. Table 2 demonstrates the detailed assessment for each domain. Overall, the risk of bias was low for the randomized trials (one trial had high risk of bias), 22 and moderate for the non-randomized observational studies due to bias in the measurement of outcomes, confounding, and in the classifications of interventions.
Risk of Bias. (a) Revised Cochrane Risk of Bias Tool for Randomized Controlled Trials. (b) Risk of Bias in Non-Randomized Studies of Interventions.
Discussion
Our systematic review of observational and interventional studies comparing sedative hypnotic agents and reporting discrete hemodynamic outcomes within 60 min after ETI demonstrates that certain sedative hypnotic agents used to facilitate ETI in critically ill adults may be associated with worse outcomes as compared to others. For example, we found that propofol is associated with more acute cardiovascular dysfunction among the included studies with the exploratory meta-analysis demonstrating no significant difference in acute cardiovascular dysfunction between etomidate and ketamine or etomidate and propofol, albeit with ketamine/propofol admixture (“ketofol”) included in the propofol group. Although only three studies reported on mental status changes at 24 h post drug administration, there was no significant difference in the rate between sedatives. Moreover, we found no association with organ injury as evidenced by SOFA score at 24 h and choice of sedative for ETI. We did not find an association between ICU length of stay and any sedative administered for ETI, although the sample size was likely underpowered to detect these differences along with significant heterogeneity among the included studies. We did observe that etomidate, as compared to ketamine, was associated with lower survival to hospital discharge in non-randomized observational studies which was not observed in the randomized controlled trials.
Hypotension in the critically ill has gained much attention as of late. Although past clinical trials have not found a significant association with higher blood pressure targets and outcomes in the critically ill,41,42 recent cohort studies have demonstrated a significant association between hypotension (MAP ≤ 65mm Hg) and negative outcomes.43,44 For example, Khanna and colleagues demonstrated that exposure to increasing amounts of hypotension in the critically ill was associated with myocardial injury, renal injury, and mortality. 43 Likewise, Smischney and colleagues found that the risk of 30 day major adverse cardiac or cerebrovascular events in postoperative critically ill patients increased as one progressed down the hypotensive ladder (MAP ≤ 65mm Hg: hazard ratio 1.52 [98.4% CI 1.17-1.96]; MAP ≤ 55mm Hg: hazard ratio 2.02 [98.4% CI 1.50-2.72]). 44 Not only does hypotension portends a negative outcome for the critically ill, it is the most common complication experienced when performing ETI, which may be partially explained by the choice of sedative agents, thus, making this a modifiable risk factor.4,5 We demonstrate in our review that exposure to hypotension, even if limited in duration, is more likely when propofol, as compared to non-propofol sedation, is administered to critically ill adults for ETI compared to other sedatives. However, we did not find more acute cardiovascular dysfunction with any sedative hypnotic agent used for ETI on exploratory meta-analysis (etomidate vs ketamine odds ratio 1.05 [0.60, 1.84]; etomidate vs propofol odds ratio 0.91 [0.33, 2.46]). Although we did not observe worse hypotension events with ketamine as compared to etomidate, a recent systematic review and meta-analysis of randomized trials comparing the administration of ketamine to etomidate concluded that ketamine probably results in more hemodynamic instability during the peri-intubation period. 45 However, these results should be interpreted cautiously as the definition of hemodynamic instability and the examined timeframes are heterogeneous throughout the literature, as highlighted in our study.
Several studies have identified an association between intraoperative hypotension and acute kidney injury.46–48 Similar associations have been found between hypotension and acute kidney injury in the critically ill.49,50 Lehman and colleagues found that the risk of acute kidney injury was related to the severity of hypotension with an odds ratio of 1.03, 95% CI 1.02–1.04 (p = <0.0001) and that for each additional hour MAP was less than 70, 60, and 50mm Hg, the risk of acute kidney injury increased by 2% (odds ratio 1.02, 95% CI 1.00-1.03, p = 0.0034), 5% (odds ratio 1.05, 95% CI 1.02-1.08, p = 0.0028), and 22% (odds ratio 1.22, 95% CI 1.04-1.43, p = 0.0122). 49 Moreover, a recent systematic review and meta-analysis found that on subgroup analysis, a higher MAP may reduce the risk of undergoing renal replacement therapy in those with chronic hypertension as compared to those without chronic hypertension (relative risk 0.83, 95% CI 0.71-0.98, I2 0%, p = 0.02). 51 Our review was limited by a lack of data on 24 h acute kidney injury and thus we cannot discern a difference between rates of acute kidney injury and choice of sedative used for ETI. In addition, studies have demonstrated an association with ICU hypotension and delirium.52,53 We did not find any association with delirium and the use of certain ETI sedatives. Although we did not detect any such association, ketamine is known to cause hallucinations and a recent observational study evaluating ketamine's effect on incident delirium among the critically ill showed that ketamine use was among the risk factors for delirium (adjusted odds ratio 5.60 [95% CI 1.09-29.15]). Apart from the use of ketamine, age (adjusted odds ratio 1.03 [95% CI 1.01-1.06]), coma (adjusted odds ratio 2.10 [95% CI 1.15-3.78]), opioid use (adjusted odds ratio 171.17 [95% CI 66.45-553.68]), and benzodiazepine use (adjusted odds ratio 34.07 [95% CI 8.12-235.34]) were each significantly associated with delirium. 54 Based on literature, an admixture of ketamine and propofol (“ketofol”) appears to have reduced rates of incident delirium, perhaps relating to propofol acting at the gamma-aminobutyric acid α receptor, inherent analgesic properties of ketamine, and reduced doses of ketamine when combined with propofol all while avoiding benzodiazepines and opioids.13,55
Examinations concerning the impact of sedatives used for ETI on illness severity, as measured by the SOFA score and others, are limited in the literature. Jabre and colleagues demonstrated, in a randomized controlled trial between ketamine and etomidate, that the mean maximum SOFA score between the two agents did not differ significantly (ketamine: 9.6 ± 3.9, etomidate: 10.3 ± 3.7; mean difference 0.7 [95% CI 0.0-1.4], p = 0.056). 15 Several studies report on the anti-inflammatory effects of propofol, which may be related to the lipid emulsion of the drug.56–58 However, a study did show increased levels of proinflammatory cytokines (interleukin-1, interleukin-6, tumor necrosis factor alpha) following long-term propofol infusions while long-term benzodiazepine infusions were associated with reduced levels of proinflammatory cytokines. 59 The implications of the immunomodulatory effects of sedative agents used for ETI on illness injury remain to be determined, however given the evidence thus far, it stands to reason that limiting long-term exposure, such as targeting light sedation and daily awakening trials, is prudent. We did not find any association between sedative agents used for ETI and worse SOFA scores in critically ill adults.
Several studies have evaluated the association of sedative agents used for ETI with long-term outcomes such as length of stay and mortality. Most of this literature has focused on the association of etomidate with mortality through its suppression of cortisol in the critically ill. Although this literature is heterogeneous in its conclusion between etomidate and increased mortality, a recent study of over 12 000 patients demonstrated that compared to a propofol cohort, the etomidate cohort showed 57% lower odds of 48 h survival (odds ratio 0.43 [98.75% CI 0.27-0.73]). Interestingly, the investigators found that the odds of mortality increased by 1.36 times per 0.1milligrams/kilogram of etomidate administered (odds ratio 1.36 [95% CI 1.23-1.49]). 60 In addition, a meta-analysis of over 8000 critically ill patients demonstrated an increased overall relative mortality rate with etomidate that correlated with the severity of critical illness scores. 61 Moreover, a recent bayesian meta-analysis of nearly 3000 patients demonstrated a moderate probability (83.2%) that ketamine administered during ETI is associated with a reduced risk of mortality as compared to etomidate (376/1475 [25%] vs 411/1503 [27%]; relative risk 0.93 [95% Credible interval 0.79-1.08). 62 This is in-line with a recent large retrospective analysis demonstrating more favorable survival with ketamine compared to etomidate administered on the day of mechanical ventilation initiation. 63 We observed a reduction in survival to hospital discharge in those who were given etomidate versus an alternative sedative for ETI in the non-randomized observational studies that was not observed in the randomized trials. Our exploratory meta-analysis on survival to hospital discharge demonstrated reduced survival with etomidate compared to ketamine sedation, which was primarily weighted by the non-randomized observational studies (odds ratio 0.76 [95% CI 0.62, 0.92]).
Our study has several strengths. First, we attempted to collect data on several outcomes related to sedative hypnotic agents utilized for ETI in critically ill adults. Previous studies have focused on a limited number of outcomes related to ETI sedative hypnotic agents in critically ill adults and thus the outcome metrics evaluated in this review have not been explored previously. Second, we evaluated several outcomes with a timeframe that could potentially be associated with the sedative hypnotic agent administered during the intubation of the critically ill adult, thereby limiting further confounding between these outcomes and sedative hypnotic agents. Third, we focused on studies that provided granular hemodynamic data given the significant impact hypotension in critically ill adults has on several of the reported outcomes. Lastly, we observed associations between some outcomes and sedative hypnotic agents used for ETI that warrant further research in future studies. Our study adds to the growing body of literature on etomidate and survival to hospital discharge as well as on evaluations of acute cardiovascular dysfunction surrounding ETI.
Limitations
As with other systematic reviews and exploratory meta-analyses, our results are limited by study quality which may be more relevant for our study given we included both randomized controlled trials and non-randomized observational studies. Although we feel our search strategy was comprehensive, we only included studies conducted after 2000 and therefore may have missed earlier studies that could have altered the observed results. However, we wanted to limit temporal shifts in airway management in the critically ill such as changes in devices or drugs that were utilized during ETI from earlier time periods. Second, we did not evaluate intubations performed in non-critically ill patients, and therefore our results may not apply to the non-critically ill adult. However, we did study all critically ill adults, whether in the ICU or elsewhere. Therefore, our results are generalizable to the adult critically ill population as a whole and not just to those in the ICU. Third, we may have missed studies on the critically ill by including only those studies with granular peri-intubation hemodynamic data reported within 60 min after ETI. A recent systematic review and meta-analysis comparing ketamine and etomidate for ETI in critically ill adults included data from two studies that did not meet our inclusion criteria, and it is possible that inclusion of these or other studies into our analyses could yield different results.62,64,65 However, given the prevalence of peri-intubation hypotension and its impact on patient related outcomes, we felt it was necessary to focus on studies that reported granular peri-intubation hemodynamic data within a clear timeframe. Moreover, the timeframe of 60 min for hemodynamic data to evaluate acute cardiovascular dysfunction may have confounded our findings as the effects of ETI sedative agents typically last minutes. With that said, this timeframe is consistent with that of numerous other studies evaluating peri-intubation hemodynamics.66,67,68 In addition, we attempted to extract data on vasoactive agents from the included studies as these medications could alter hemodynamics and affect our estimates. However, we were unable to collect this data in all included studies. Of the 23 included studies, only a small number of studies were included in the exploratory meta-analysis due to the inability to extract measures of variation in the studies not included in the meta-analysis. Subgroup exploratory meta-analyses are not reported for all outcomes due to the limited number of studies in each analysis, which has limited similar efforts. 45 Fourth, our review is likely influenced by not only the above factors, but also by the choice of drug utilized at the time of ETI as well as the dose administered. Fifth, our review demonstrated high heterogeneity in the analyses and moderate risk of bias in the non-randomized observational studies with low risk of bias in the randomized controlled trials. Advanced bias evaluations were not conducted due to the limited number of studies in our exploratory meta-analysis. Therefore, caution is advised when interpreting our results. Furthermore, additional assessments of methodological quality such as the GRADE framework were not conducted and thus, this further adds to caution when interpreting our results. Finally, we were unable to evaluate the outcomes of acute kidney injury and oral morphine equivalents at 24 h post drug administration or intubation difficulty score due to the lack of reporting.
Conclusions
In this systematic review and exploratory meta-analysis, we observed more acute cardiovascular dysfunction when propofol, as compared to non-propofol sedation, was used for ETI; however, exploratory meta-analyses did not demonstrate an association between any sedative agent and acute cardiovascular dysfunction. We did observe worse survival to hospital discharge in critically ill patients who were given etomidate as compared to those critically ill adult patients given non-etomidate sedatives. Our exploratory meta-analysis revealed lower survival to hospital discharge with etomidate as compared to ketamine. We did not find any association between sedative hypnotic agent and the other outcomes of delirium, new vasopressor requirement, and illness severity at 24 h post drug administration nor did we observe any association with sedative hypnotic agent choice and ICU length of stay. Although we did not observe any associations between sedative hypnotic agent choice administered during ETI and specific outcome metrics, this may reflect sample size and high heterogeneity of included studies.
Supplemental Material
sj-docx-1-jic-10.1177_08850666251337702 - Supplemental material for Outcomes of Sedative Hypnotic Agents Used for Endotracheal Intubation in Critically Ill Adults: A Systematic Review with Exploratory Meta-Analysis
Supplemental material, sj-docx-1-jic-10.1177_08850666251337702 for Outcomes of Sedative Hypnotic Agents Used for Endotracheal Intubation in Critically Ill Adults: A Systematic Review with Exploratory Meta-Analysis by Nathan J. Smischney, George Williams, Craig S. Jabaley, Ashish K. Khanna, Bethany Bouldin, Andrew R. Petrilli, Hao Deng, Elissa A. Kinzelman-Vesely and Ronald G. Pearl in Journal of Intensive Care Medicine
Supplemental Material
sj-docx-2-jic-10.1177_08850666251337702 - Supplemental material for Outcomes of Sedative Hypnotic Agents Used for Endotracheal Intubation in Critically Ill Adults: A Systematic Review with Exploratory Meta-Analysis
Supplemental material, sj-docx-2-jic-10.1177_08850666251337702 for Outcomes of Sedative Hypnotic Agents Used for Endotracheal Intubation in Critically Ill Adults: A Systematic Review with Exploratory Meta-Analysis by Nathan J. Smischney, George Williams, Craig S. Jabaley, Ashish K. Khanna, Bethany Bouldin, Andrew R. Petrilli, Hao Deng, Elissa A. Kinzelman-Vesely and Ronald G. Pearl in Journal of Intensive Care Medicine
Supplemental Material
sj-docx-3-jic-10.1177_08850666251337702 - Supplemental material for Outcomes of Sedative Hypnotic Agents Used for Endotracheal Intubation in Critically Ill Adults: A Systematic Review with Exploratory Meta-Analysis
Supplemental material, sj-docx-3-jic-10.1177_08850666251337702 for Outcomes of Sedative Hypnotic Agents Used for Endotracheal Intubation in Critically Ill Adults: A Systematic Review with Exploratory Meta-Analysis by Nathan J. Smischney, George Williams, Craig S. Jabaley, Ashish K. Khanna, Bethany Bouldin, Andrew R. Petrilli, Hao Deng, Elissa A. Kinzelman-Vesely and Ronald G. Pearl in Journal of Intensive Care Medicine
Footnotes
Abbreviations
Acknowledgements
None.
Author Contributions
NJS, GW, CSJ, BB, ARP, HD, EAKV, RGP, and AKK made substantial contributions to the conception and design of the study, the acquisition and interpretation of data, have drafted the work or substantively revised it, have approved the submitted version, and have agreed both to be personally accountable for the author's own contributions and to ensure that questions related to the accuracy or integrity of any part of the work, even ones in which the author was not personally involved, are appropriately investigated, resolved, and the resolution documented in the literature.
Ethical Considerations
Ethical approval was not required. This systematic review and meta-analysis was registered aprior: Prospero registration (CRD42022304366).
Consent to Participate and Publish
Not applicable.
Data Availability
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
Declaration of Conflicting Interests
NJS receives consulting fees from Edwards Life Sciences. AKK consults for Medtronic, Edwards Lifesciences, Philips Research North America, GE Healthcare, Potrero Medical, Sentinel Medical, Retia Medical, Nihon-Kohden USA, Caretaker Medical, TNO, Bayer Corporation, Fifth Eye Inc., Pharmazz Inc., Viatris, Innoviva Specialty Therapeutics, and Zynex Medical.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Supplemental Material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
