Effect of Ketamine/Propofol Admixture on Peri-Induction Hemodynamics: A Systematic Review and Meta-Analysis

To evaluate the effectiveness of an admixture of ketamine and propofol on peri-induction hemodynamics during airway manipulation, we searched electronic databases of randomized controlled trials from January 1, 2000, to October 17, 2018. Trial screening, selection, and data extraction were done independently by two reviewers with outcomes pooled across included trials using the random-effects model. We included 10 randomized trials (722 patients, mean age of 53.99 years, 39.96% female). American Society of Anesthesiologists physical status was reported in 9 trials with classes I and II representing the majority. Ketamine/propofol admixture was associated with a nonsignificant increase in heart rate (weighted mean difference, 3.36 beats per minute (95% CI, −0.88, 7.60), I2 = 88.6%), a statistically significant increase in systolic blood pressure (weighted mean difference, 9.67 mmHg (95% CI, 1.48, 17.86), I2 = 87.2%), a nonsignificant increase in diastolic blood pressure (weighted mean difference, 2.18 mmHg (95% CI, −2.82, 7.19), I2 = 73.1%), and a nonsignificant increase in mean arterial pressure (weighted mean difference, 3.28 mmHg (95% CI, −0.94, 7.49), I2 = 69.9%) compared to other agents. The risk of bias was high and the certainty of evidence was low. In conclusion, among patients undergoing airway manipulation and needing sedation, the use of a ketamine/propofol admixture may be associated with better hemodynamics compared to nonketamine/propofol sedation. This trial is registered with CRD42019125725.


Introduction
Peri-intubation hypotension, defined by either systolic blood pressure or mean arterial pressure (MAP) below a certain threshold (i.e., <90 mmHg or <65 mmHg) or the introduction of vasopressors, has been recognized as a potential target area for research given its association with patientcentered outcomes. For example, peri-intubation hypotension has been associated with both increased length of stay and mortality [1][2][3]. Furthermore, this association has been identified not only in the critically ill but also in elective surgical patients [4,5]. Several studies indicate that the frequency with which peri-intubation hypotension occurs is fairly high, with one report indicating an incidence of greater than 80% [1,2,5,6]. However, the data presented on the frequency is dependent on definitions currently used in the literature for which no standard consensus exists. Perhaps, the best evidence for peri-intubation hypotension incidence in the critically ill comes from Green and colleagues [7]. ey performed a systematic review of emergent intubations performed outside the operating room and found that the incidence of peri-intubation hypotension ranged from 5 to 440 cases per 1,000 intubations (0.5-44%) with a pooled estimate of 110 cases per 1,000 intubations (11%) [7]. us, peri-intubation hypotension is not uncommon, and given the associations observed in the literature, prevention of peri-intubation hypotension is likely to improve patientcentered outcomes.
Several putative risk factors have been implicated in the pathway to peri-intubation hypotension. Age, illness severity, and preintubation hemodynamic derangement have consistently been implicated in the development of peri-intubation hypotension [2,3,8,9]. One modifiable risk factor potentially leading to peri-intubation hypotension is the choice of intravenous anesthetics with some anesthetics (i.e., propofol and barbiturates) more likely to lead to peri-intubation hypotension than others (i.e., etomidate and ketamine) [10]. Lately, a novel intravenous anesthetic admixture has gained popularity based on potential hemodynamic preservation postadministration [11][12][13].
e admixture involves the combination of propofol with its vasodilatory effects balanced by the vasoconstricting properties of ketamine [10]. e majority of studies on ketamine/propofol admixture have evaluated critically ill patients in the emergency department with the evidence demonstrating a potential sparing effect on hemodynamics along with improved pain relief and sedation quality.
ese studies have evaluated ketamine/propofol admixture from the standpoint of a continuous infusion for procedural sedation and analgesia [14][15][16]. ere have been a couple of systematic reviews on ketamine/propofol admixture sedation, demonstrating that ketamine/propofol admixture appears safe and efficacious for procedural sedation and analgesia and is possibly better than propofol only at reducing cardiorespiratory problems [17,18]. e wealth of the evidence above has mainly focused on ketamine/propofol admixture use in terms of infusions for procedural sedation and analgesia. ere are limited studies addressing the potential hemodynamic preservation effects of the admixture when administered as an induction agent for endotracheal intubation. Given the above associations between peri-intubation hypotension and increased patient morbidity and mortality, and the mounting evidence with ketamine/propofol admixture as an agent that allows potential maintenance of hemodynamics when administered for endotracheal intubation, our aim was to perform a systematic review and meta-analysis on the hemodynamic effects of ketamine/propofol admixture when administered as an induction agent for airway manipulation such as endotracheal intubation.

Methods
We used the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) 2015 statement to report the trial [19][20][21][22]. All reviews were conducted by two independent reviewers (MS and AM). Data collection was performed from January 1, 2000, to October 17, 2018. A formal protocol does not exist for this systematic review and meta-analysis. e review was registered with the International Prospective Register of Systematic Reviews (CRD42019125725).

Eligibility Criteria.
Randomized controlled trials published in peer-reviewed journals and written in English were eligible if they included the following: (1) adult patients who received ketamine/propofol admixture as an induction agent in procedural areas and underwent airway manipulation such as endotracheal intubation or supraglottic device placement; (2) a comparison to other induction agents including propofol, ketamine, etomidate, and sodium thiopental or any other combination; and (3) a report of hemodynamic effects (heart rate, systolic and diastolic blood pressure, and MAP) during the first 10 minutes after the induction of general anesthesia in American Society of Anesthesiologists I II, III, and IV patients. e intervention must have equal or similar doses (e.g., ketamine/propofol admixture 1 : 1 or 1 : 2 ratios). No restrictions were placed on trial location, clinical procedure, or patient severity. We excluded pediatric patients, patients who did not undergo induction of anesthesia, or patients who did not undergo airway manipulation. We also excluded observational studies, review articles, erratum, letters, and notes.

Trial Selection.
Reviewers, working independently, screened abstracts and titles for eligibility using the above inclusion and exclusion criteria. Full-text articles were then further screened using the same criteria. At the level of fulltext screening, any disagreements were resolved by consensus between the two reviewers (MS and AM) or by consulting a third reviewer (NJS).

Outcomes.
e primary outcomes were hemodynamics (heart rate, systolic and diastolic blood pressure, and MAP) at 5 and 10 minutes following intravenous anesthetic administration. e secondary outcome of interest was pain score as assessed by the visual analog scale (VAS) during the 24 hours postdrug administration.

Methodological Quality and Certainty of Evidence.
e risk of bias was assessed using the Cochrane Collaboration's tool for randomized clinical trials. We assessed random sequence generation, allocation concealment, blinding of providers, outcome assessors, and patients, incomplete outcome data, selective outcome reporting (based on the availability of protocol and inclusion of all prespecified outcomes), and other sources of bias (conflict of interest, source funding, risk of bias due to deviations from the intended interventions, etc.). Disagreements were resolved by consensus between two reviewers (MS and AM). e overall certainty across trials for each outcome was appraised by discussion between the two reviewers using the grading of recommendations assessment, development, and evaluation (GRADE) approach. Using the grading of recommendations assessment, development, and evaluation approach, randomized trials, as in this systematic review, would provide a starting level of certainty that is high. is level can be downgraded based on the risk of bias of the ((evidence adj based) or (meta adj analys * ) or (systematic * adj3 review * ) or (control * adj3 study) or (control * adj3 trial) or (randomized adj3 study) or (randomized adj3 trial) or (randomised adj3 study) or (randomised adj3 trial) or "pragmatic clinical trial" or (random * adj1 allocat * ) or (doubl * adj blind * ) or (doubl * adj mask * ) or (singl * adj blind * ) or (singl * adj mask * ) or (tripl * adj blind * ) or (tripl * adj mask * ) or (trebl * adj blind * ) or (trebl * adj mask * ) or "Latin square" or placebo * or nocebo * or cohort * or retrospectiv * or prospectiv * or (random * and (trial or study))).mp,pt.
OR placebo * OR nocebo * OR cohort * OR retrospectiv * OR prospectiv * OR (random * and (trial or study))) 3 PUBYEAR individual trials, inconsistency in the results, indirectness, imprecision, and other considerations to provide a global assessment of the certainty warranted by the body of evidence [23].
2.6. Data Extraction. Two reviewers (MS and AM) independently extracted the following information from each trial: author, publication year, patient characteristics, intervention, comparison, and outcomes.

Data Synthesis and Statistical Analysis.
We extracted or calculated the weighted mean difference (WMD) for continuous outcomes and relative risk for binary outcomes with associated 95% confidence intervals (CIs) from the included studies. Outcomes without measures of variations (e.g., standard deviation, standard error, and CI) were not included in the meta-analysis. e DerSimonian and Laird random-effects models were used to generate combined effects [24]. A two-sided P value of less than or equal to 0.05 was used to determine significance for the secondary outcome. To evaluate heterogeneity, we calculated the I 2 statistic [25,26] and heterogeneity P values, where I 2 more than 50% or a P value of less than 0.05 suggests high heterogeneity. All statistical analyses were performed using Stata, version 15.0 (StataCorp 2017, College Station, Texas, United States).

Trial Inclusion.
e searches identified 820 trials. After excluding the irrelevant trials, 33 full-text articles were assessed for eligibility, of which 10 randomized controlled trials (30%) met all criteria and were included in data analysis (n � 722) ( Figure 1). All the trials were reported in full-length journal articles.

Trial Characteristics.
ere were a total of 10 trials with 722 patients included. e number of patients in each trial ranged from 40 to 100 (mean, 72.22) with a mean age of 53.99 years. e proportion of participants who were female ranged from 0 to 100% (mean, 39.96%).
e American Society of Anesthesiologists physiologic status among participants ranged from I to III. Publication dates ranged from 2000 to 2018 (median, 2011 (all trials published after 2000)).
e included trials were conducted in several different countries (four trials in Turkey, three trials in Iran, and one trial from India, Japan, Egypt, and United States). e characteristics of the included trials with dosing are summarized in Table 2.
3.6. Diastolic Blood Pressure. Diastolic blood pressure at 5 minutes was reported in four trials (n � 305) [11,[28][29][30] and one trial (n � 84) [11] reported diastolic blood pressure at 10 minutes. All trials assessed ketamine/propofol admixture, two compared to propofol and two compared to etomidate at 5 minutes and one trial compared to propofol at 10 minutes. Patients in the ketamine/propofol admixture group had a nonsignificant increase in diastolic blood pressure at 5 minutes (WMD, 2.18 mmHg (95% CI, −2.82 to 7.19), I 2 � 73.1%) compared to patients in the nonketamine/propofol admixture group. At 10 minutes, there was a statistically significant increase in diastolic blood pressure in the ketamine/propofol admixture group compared to the nonketamine/propofol admixture group (WMD, 4.80 mmHg (95% CI, 0.24 to 9.36), I 2 � not applicable) ( Table 4 and Figures 4(a) and 4(b)).

Discussion
e present systematic review and meta-analysis demonstrate that ketamine/propofol admixture results in a potentially better hemodynamic profile as compared to other agents used for induction of anesthesia. In particular, systolic blood pressure was significantly higher with ketamine/ propofol admixture sedation as compared to nonketamine/ propofol admixture sedation when induction of anesthesia is required for airway manipulation. Although not statistically significant, heart rate, diastolic blood pressure, and MAP were higher when ketamine/propofol admixture-based sedation was employed versus nonketamine/propofol admixture based-sedation for induction of anesthesia. Furthermore, one study indicated a possible beneficial effect on pain scores in the immediate 24 hours postdrug administration [27].
Evidence has demonstrated that maintaining hemodynamics around the time of airway instrumentation such as endotracheal intubation is vitally important, in both the critically ill and noncritically ill. For example, in the critically ill, peri-intubation hypotension was associated with increased odds of dying (39% peri-intubation hypotension versus 30% no peri-intubation hypotension, P � 0.045) and increased odds of experiencing intensive care unit length of stay greater than 14 days, duration of mechanical ventilation longer than 7 days, and requiring renal replacement therapy (odds ratio 2.0, 95% CI: 1.30-3.07, P � 0.0017) [1]. ese   observations have been demonstrated in other populations of critically ill patients [36]. In noncritically ill patients undergoing noncardiac surgery, a significant fraction of all hypotensive events occurred before skin incision and, therefore, due to anesthetic management as demonstrated in one study. In addition, this study revealed that the odds of developing acute kidney injury increased with MAP less than 65 mmHg (odds ratio 1.02, 95% CI: 1.01-1.04, P � 0.004) [37]. Moreover, patients undergoing noncardiac surgery who experienced intraoperative hypotension defined as MAP less than 65 mmHg or a MAP decrease of 20% from baseline had increased odds of developing both myocardial and kidney injury [38]. Likewise, patients who developed greater than a 30% decrease in MAP from baseline intraoperatively had increased odds of developing postoperative stroke (odds ratio 1.013/min hypotension, 99.9% CI: 1.000-1.025, P ≤ 0.001) [39]. A recent systematic review suggested poor outcomes with increased end-organ injury in patients undergoing noncardiac surgery who experience MAP decreases <80 mm Hg for ≥10 minutes [40]. By the same token, elevated MAP thresholds have been postulated to be beneficial in the critically ill population. For example, one study demonstrated improved microcirculation in septic shock patients with previous hypertension when MAP was increased above a threshold of 65 mmHg [41]. Furthermore, others have shown that maintaining MAP well above 65 mmHg (i.e., 85 mmHg) in septic, critically ill patients may result in less myocardial and kidney injury as well as decreasing overall mortality [42].
Given that periprocedure (i.e., endotracheal intubation) hemodynamics are influenced by sedation and the preponderance of evidence demonstrating that preventing hypotensive episodes, especially with MAPs less than 65 mmHg or SBPs less than 90 mmHg, and maintaining perfusion pressure near baseline improves outcomes, the selection of an intravenous anesthetic agent during a procedure (i.e., endotracheal intubation) ought to have this goal  Low due to risk of bias and imprecision CI: confidence interval; HR: heart rate; WMD: weighted mean difference; RCT: randomized controlled trial; SBP: systolic blood pressure; DBP: diastolic blood pressure; MAP: mean arterial pressure; N/A: not applicable.    tailored to each patient as high pressure/heart rate may be harmful to some. For example, ketamine may exacerbate myocardial ischemia in noncompensated coronary artery disease or hypertension through its effects on central nervous system stimulation and inhibition of norepinephrine reuptake (i.e., cardiac population) [10].
e current systematic review and meta-analysis are consistent with prior reviews of ketamine/propofol admixture administration for procedural sedation. Prior reviews have demonstrated better hemodynamics with ketamine/propofol admixture-based procedural sedation than with nonketamine/propofol admixture-based  assessed adverse respiratory events and recovery times with ketamine/propofol admixture-based procedural sedation as compared to propofol-based procedural sedation. e authors noted that adverse respiratory events were significantly reduced with ketamine/propofol admixture as compared to propofol (risk ratio 0.82, 95% CI: 0.68-0.99, P � 0.04); however, recovery times were similar in both groups. Hemodynamic thresholds were not assessed in this review [44]. Although the above reviews point towards the safety of this combination, the admixture can, by its very nature, lead to some undesired effects. Apart from the use of ketamine in the cardiac population (discussed above), ketamine causes initial release of glutamate through nicotinamide adenine dinucleotide phosphate hydrogen oxidase 2 leading to emergence delirium with eventual blockade of glutamate's effects through N-methyl-D-aspartate antagonism. Emergence delirium could be devastating to patients with underlying mental illnesses such as schizophrenia, bipolar disorder, or posttraumatic stress disorder to name a few. is emergence delirium is dose dependent and also affected by cointerventions such as benzodiazepine administration. Interestingly, propofol has been shown to blunt glutamate's effects through gamma-aminobutyric acid A agonist activity [10]. is may explain the findings in the literature demonstrating a reduced rate of emergence delirium with the admixture (∼30% with ketamine only vs. <5%) [10-12, 14, 31].
Our study has several benefits. First, we focused on a theoretical advantage of ketamine/propofol admixture over other agents used for induction of anesthesia during procedures such as endotracheal intubation, namely, hemodynamic preservation. Furthermore, we extracted hemodynamic data in the immediate period after drug administration, thereby limiting the effects of other cointerventions. Second, we extracted data on all drug comparisons to ketamine/propofol admixture.
ird, we included trials that had either similar or near equivalent dosing of ketamine/propofol admixture.

4.1.
Limitations. Similar to other systematic reviews and meta-analyses, our results are limited by clinical trial quality. Although our search strategy was comprehensive, we included only English-language articles and published articles after 2000 and thus may have missed trials not published in English or published before the year 2000. Hemodynamic data were not recorded similarly in all trials nor was the frequency of hemodynamic data collected. Furthermore, 5 and 10 minutes may be an oversimplification of hemodynamic representation as intravenous anesthetics used for induction of anesthesia have a short initial elimination halflife and, therefore, collecting hemodynamic data every minute would have been ideal [10]. Moreover, we did not calculate MAP from trials reporting systolic and diastolic blood pressure because individual patient data were not reported. ree studies met our inclusion criteria and were included in the systematic review. However, we did not include these studies in the meta-analysis as we were unable to extract measures of variation [32,33,35]. Data on vasoactive medications which could alter hemodynamics and affect our estimates were not reported in the majority of included trials. We do not report subgroup meta-analyses for the different patient populations or interventions studied due to the limited number of studies in each meta-analysis (only 1 meta-analysis included 3 studies and the remainder had 2 studies). us, any subgroup and sensitivity analyses would reduce the number further with only 1 study, which is already presented well in the forest plots. Lastly, our systematic review illustrated moderate to high heterogeneity in the analyses and high risk of bias of the included studies. With such small numbers of studies in the meta-analysis, advanced bias evaluations were not appropriate. us, our results should be interpreted with caution.

Conclusions
We demonstrated a potentially better hemodynamic profile with ketamine/propofol admixture-based induction of anesthesia versus nonketamine/propofol admixture-based induction of anesthesia for procedures requiring airway manipulation such as endotracheal intubation. However, given moderate to high heterogeneity of trials included in the current meta-analysis, our results should be interpreted with caution.

Conflicts of Interest
e authors declare that they have no conflicts of interest.