Abstract
Background:
Ketamine is a commonly used sedative agent for procedural sedation and rapid sequence intubation. Ketamine can also be administered via continuous infusion as an adjunct agent for sedation and analgesia in mechanically ventilated patients requiring high levels of sedation. In individuals who are not catecholamine-depleted, ketamine induces the release of norepinephrine, epinephrine, and dopamine, leading to transient increases in cardiovascular function. This mechanism suggests that ketamine, when used as a sedative agent, could potentially reduce vasopressor requirements in patients undergoing vasopressor therapy to maintain stable hemodynamics while needing high levels of sedation. The existing literature on continuous infusion ketamine’s effect on vasopressor requirements is conflicting, with some studies reporting potential benefit and others showing no clinical difference.
Methods:
This retrospective cohort study was conducted at Saint Joseph Hospital, a 433 bed community hospital located in Lexington, KY. The study included mechanically ventilated patients admitted between August 7, 2018, and December 31, 2024 who received a vasopressor agent for at least 3 hours for hemodynamic support prior to the initiation of continuous infusion ketamine, and received both ketamine and vasopressors concomitantly for at least 3 hours. To be included, patients must have also been receiving vasopressors at a dose ≥0.1 mcg/kg/min norepinephrine equivalents (NEE). The primary outcome was change in vasopressor requirements in critically ill patients at 6, 12, 24 hours after initiation of a continuous ketamine infusion as an adjunct sedative.
Purpose:
The purpose of this research is to investigate whether the use of ketamine as an adjunct sedative and analgesic agent can decrease vasopressor requirements in critically ill patients.
Results:
Forty-one patients met inclusion criteria. There was a difference in mean vasopressor requirements from baseline to 6, 12, 24, hours after continuous ketamine infusion initiation (P = .035). Mean vasopressor requirements decreased from 0.218 NEE at baseline to 0.186 NEE at 6 hours (P = .002), 0.145 NEE at 24 hours (P = .022).
Conclusion:
Based on the results of this study, continuous infusion ketamine can reduce mean vasopressor requirements when used as an adjunct sedative in mechanically ventilated patients.
Introduction
Ketamine is a commonly used sedative agent for procedural sedation and rapid sequence intubation. Ketamine can also be administered via continuous infusion as an adjunct agent for sedation and analgesia in mechanically ventilated patients requiring high levels of sedation. In these settings, ketamine is often combined with other sedatives such as fentanyl, propofol, and/or midazolam.
Ketamine is classified as a nonbarbiturate dissociative anesthetic agent. It produces a sedative and analgesic effect by acting as an antagonist on the N-methyl-D-aspartate (NMDA) and glutamate receptors that block the HCN1 receptor. Ketamine is also thought to have partial mu-opioid agonist activity. 1 A well-known property of ketamine is its ability to transiently increase blood pressure and heart rate through its sympathomimetic effects. In individuals who are not catecholamine-depleted, ketamine induces the release of norepinephrine, epinephrine, and dopamine, leading to transient increases in cardiovascular function. 2 This mechanism suggests that ketamine, when used as a sedative agent, could potentially reduce vasopressor requirements in patients undergoing vasopressor therapy to maintain stable hemodynamics while needing high levels of sedation.
Another potential benefit of ketamine is its minimal impact on respiratory drive, as it possesses a high threshold for respiratory depression. 3 In critically ill patients, a medication that provides necessary sedation and analgesia while potentially lowering vasopressor requirements could be highly beneficial. This is particularly important given that high-dose vasopressors are associated with increased mortality in the intensive care unit (ICU). 4
The existing literature on the effects of ketamine on vasopressor requirements is conflicting. Some studies suggest a benefit by showing lower vasopressor requirements with the addition of ketamine, while others do not show any significant impact. The purpose of this research is to contribute to this body of literature by investigating whether the use of ketamine as an adjunct sedative and analgesic agent can decrease vasopressor requirements in critically ill patients. Determining if ketamine is a better alternative as an adjunct sedative compared to other medications could have significant implications for patient outcomes in the ICU.
Current literature is inconclusive when examining ketamine’s effects on vasopressor requirements. A 2022 retrospective cohort study of 200 patients assessed the effect of ketamine on vasopressor requirements when combined with one other sedatives such as fentanyl, propofol, dexmedetomidine, or a benzodiazepine versus the combined regimen of propofol and fentanyl. This study found a statistically significant decrease in vasopressor requirements when comparing ketamine as an adjunct sedative to the propofol and fentanyl group (P-value < .0001). 5
In contrast, another 2022 retrospective cohort study consisting of 68 patients found that at 1, 3, and 30 hours after initiation of ketamine there was no statistically significant effect on norepinephrine requirements when compared to control groups of midazolam or propofol resulting in no favorable hemodynamic effect. 6
When comparing continuous ketamine infusions in mechanical ventilated COVID-19 patients, a 2024 study of 84 patients found no difference in vasopressor requirements with ketamine versus propofol at 24, 48, and 72 hours after ketamine initiation. This study allowed the concomitant use of other sedatives such as fentanyl and dexmedetomidine to be administered with either ketamine or propofol. Notably, the study did find a statistically significant difference in patient’s mean arterial pressure at 24, 48, and 96 hours. The study also found a decrease in overall opioid requirements in the ketamine group at 24, 48, 72, and 96 hours when compared to propofol. 7
Lastly, a scoping review was published in 2023 and it assessed the current literature on continuous ketamine infusions for sedated patients in the intensive care unit utilized 27 different studies and case reports. This review found that results on improved hemodynamics (including assessment of MAP, SBP, and vasopressor requirements) was inconsistent, with some studies suggesting there is no benefit with ketamine use. 8
In many studies investigating ketamine’s impact on vasopressor requirements, the ketamine and control groups lack equal treatment conditions. While both ketamine and control groups may be conducted in similar populations (eg, critically ill), patients are unequally receiving other sedative agents such as propofol and/or dexmedetomidine that can cause hypotension and therefore cause patients to have increased vasopressor doses. This introduces confounding variables into the study that could skew the results. However, a study evaluating the effect on vasopressor requirement where the patient serves as their own control group (ie, vasopressor requirement before and after ketamine initiation) could be advantageous because the impact of hypotension causing sedative medications could be eliminated.
In summary, the idea of using ketamine as an adjunct for sedation is not novel, but current literature is conflicting on ketamine’s hemodynamic effect and ability to potentially lower vasopressor requirements in critically ill patients. Conducting a study in which the patient serves as their own control group could be a valuable study to add to the current literature. The purpose of this retrospective study is to determine if the initiation of ketamine in critically ill patients lowers vasopressor requirements. This study also plans to analyze if ketamine use in critically ill patients lowers opioid requirements. This analysis could help add to the body of literature that ketamine has a more recognized and valuable role in sedation and analgesia as an adjunctive agent, particularly in hemodynamically unstable patients.
Ethics Approval
Due to the retrospective nature of the study exempt status was granted by the Catholic Health Initiatives Institutional Review Board.
Methods
Study Design and Setting
This single-center, retrospective cohort study was conducted at Saint Joseph Hospital, a 433-bed acute care community hospital located in Lexington, KY. The study institution has 50 adult ICU beds and a 35 bed emergency department. The study included patients admitted to the institution between August 7, 2018, and December 31, 2024. Each patient served as their own control within the study to limit confounding variables such as unequal distribution of medications known to cause hypotension. The patients’ vasopressor requirements were compared from the initiation of ketamine to 6, 12, 24, and 48 hours after the initiation of continuous infusion ketamine to determine any changes in vasopressor requirements. This study adhered to ethical guidelines for retrospective studies.
Selection of Participants
We included patients aged ≥18 years of age, who received at least one vasopressor agent (norepinephrine, phenylephrine, epinephrine, or vasopressin) and continuous sedation and/or analgesia (propofol, fentanyl, dexmedetomidine, midazolam) for at least 3 hours prior to ketamine continuous infusion. These agents needed to be given concomitantly with a ketamine continuous infusion for at least 3 hours. Included patients also needed to receive ≥ 0.1 mcg/kg/min or greater of norepinephrine equivalents (NEE) at the time of ketamine initiation. Due to hospital policy, these patients had to be located in an intensive care unit or the emergency department at the time of ketamine initiation. Patients were excluded from the study if they did not meet the previously mentioned criteria, or if they were pregnant or lactating. All patients that met inclusion criteria were included in the study.
Data Screening and Recording
Patients were identified using the SlicerDicer feature within the EPIC electronic health record based on a retrospective search of both a ketamine continuous infusion order and at least one vasopressor agent order on a patient’s profile during an admission. All patients identified in the retrospective search were screened for inclusion to prevent selection bias. The list of eligible patients had their charts reviewed retrospectively, and the data was recorded on an external spreadsheet. Patient confidentiality was maintained by anonymizing data and securely storing records. Vasopressor and fentanyl requirements were collected at 6, 12, 24, and 48 hours after ketamine initiation. The number of vasopressors and particular sedation agents used at each time point was also collected. Additionally ICU length of stay, duration of mechanical ventilation, midodrine initiation and average daily dose, Charlson Comorbidity Index, diagnosis of opioid use disorder and treatment if applicable, and the use of concomitant steroids were also collected. Any data that was deemed missing during analysis would be retrospectively collected in the patients’ electronic health record, if applicable.
Outcome Measures
The primary outcome of this retrospective cohort study was the change in vasopressor requirement, measured in mcg/kg/min NEE from the time of ketamine initiation (baseline) to 6, 12, 24 hours after initiation of continuous ketamine infusion. Norepinephrine equivalents were calculated based on Table 1 which is extrapolated based on current literature. 9 Secondary outcomes of the study included vasopressor requirement from baseline to 48 hours after initiation of ketamine, the number of vasopressor agents the patients were on at baseline compared to 6, 12, 24, and 48 hours after ketamine initiation, the number of sedative agents the patients were on at baseline compared to 6, 12, 24, and 48 hours after ketamine initiation, and continuous infusion opioid requirements at 6, 12, 24, and 48 hours after ketamine initiation. A planned secondary analysis was performed. In this analysis, the original inclusion criteria was altered to remove the vasopressor requirement of ≥0.1 mcg/kg/min NEE with all other criteria remaining unchanged.
Norepinephrine Equivalent Chart.
Statistical Analysis
The study population was to create both a control and intervention group. Each patient served as their own control for this study. The control group consisted of the study population at the time point of ketamine initiation, while the intervention group was the study population after the initiation of ketamine. All statistical analysis for the control group was based on patient values at the time of ketamine initiation. This time point will be considered the baseline. The intervention group was defined as any time after the initiation of continuous infusion ketamine.
The collected data was evaluated for normality of distribution using a Shapiro-Wilk test. For the primary outcome, a paired samples t-test or repeated measures ANOVA was used to determine statistical significance. For secondary outcomes with continuous variables, a paired samples t-test or Wilcoxon signed-rank test as appropriate. For categorical data a Chi Square or Fisher’s Exact test will be used as appropriate for sample size. All analysis will utilize a 95% confidence interval and an alpha level of .05 for significance.
Results
Between August 1, 2018 and December 31, 2024, 41 patients met inclusion criteria for this study (Figure 1). Baseline characteristics for the study population included a mean age of 52.5 years, and 68.3% of the included patients were male. The most common comorbidities seen in the included patients was diabetes mellitus (34.1%), COPD (26.8%), and hepatic disease (24.4%). The mean Charlson Comorbidity Index for the study population was 2.9. The mean ICU length of stay was 20.195 days, and the mean number of days on mechanical ventilation was 14.805. Baseline characteristics can be found in Table 2.
Study participants.
Baseline Characteristics.
The primary outcome data showed that mean vasopressor dose decreased at 6, 12, and 24 hours compared to baseline (P = .035, Figure 2). Mean baseline vasopressor dose was 0.218 mcg/kg/min NEE compared to 0.186 mcg/kg/min NEE at 6 hours, 0.184 mcg/kg/min NEE at 12 hours, and 0.145 mcg/kg/min NEE at 24 hours (Table 3). When broken down at each time point, seen in Table 4, there is a statistically significant difference in baseline to 6 hours (P = .002), and baseline to 24 hours (P = .022), but there was not noted to be a statistically significant difference between baseline and 12 hours (P = .129). Relevant to these findings, the target mean arterial pressure for included patients was generally 65 mmHg, though alterations to this target, if any, was left to the clinical judgment of the individual provider. There was no pre-defined criteria for patient inclusion in the study.
Vasopressor requirements versus time.
Study Outcomes.
Study Outcomes Comparing Baseline to Each Time Point.
Analysis of secondary outcomes show that the mean number of vasopressor agents, seen in Figure 3, decreased from 1.439 at baseline compared to 1.390 at 6 hours, 1.341 at 12 hours, and 1.00 at 24 hours after ketamine initiation (P = .010). When the number of vasopressor agents was analyzed comparing baseline to individual time points, this outcome was only statistically significant from baseline to 24 hours (P = .008).
Number of vasopressors and sedatives versus time.
Mean fentanyl dose decreased as well from 250.000 mcg/hr at baseline compared to 237.195 mcg/hr at 6 hours, 213.415 mcg/hr at 12 hours, and 160.366 mcg/hr at 24 hours (Figure 4). When comparing baseline fentanyl dose to 6, 12, and 24 hours there was a statistically significant difference found (P = .002). When comparing baseline fentanyl dose to each individual time point, this outcome was also only statistically significant when comparing fentanyl dose at baseline to 24 hours (P = .002).
Fentanyl dose versus time.
Mean number of sedative agents (excluding ketamine), as shown in Figure 3, was found to be 1.780 at baseline compared to 1.585 at 6 hours, 1.561 at 12 hours, and 1.366 at 24 hours after ketamine initiation (P < .001). When comparing baseline to each individual time point, this secondary outcome was statistically significant at all time points respectively (P = .031, .027, <.001). Similar to the mean arterial pressure goals previously mentioned, sedation goals were also left the discretion of the prescribing provider. The use of specific sedatives associated with hypotension was also analyzed from baseline to 24 hours. The number of patients receiving propofol decreased from 24 patients at baseline to 21 patients after 24 hours. Midazolam use remained constant at five patients at both time points, while dexmedetomidine use decreased slightly from eight patients at baseline to seven patients after 24 hours.
The difference in vasopressor requirements from baseline to 48 hours was not included in the primary outcome data due to three patients being excluded due to death between 24 and 48 hours after ketamine initiation. This analysis showed a statistically significant difference (P = .005) when comparing a mean vasopressor dose at baseline of 0.213 mcg/kg/min NEE to a mean vasopressor dose at 48 hours of 0.118 mcg/kg/min NEE. This data can be found in Table 5.
Study Outcomes at 48 Hours.
The mean ketamine infusion rate was also recorded and was found to be statistically significant between baseline and 24 hours (P < .001). The mean initial ketamine infusion rate was found to be 0.339 mg/hr. The mean ketamine infusion rate at 24 hours after starting the infusion was found to be 1.10 mg/hr. The escalation of ketamine dosing, along with adjustments to other sedatives, was based on provider discretion. There was no study protocol for sedative adjustments (Figures 3 and 4).
In the pre-planned secondary analysis, inclusion criteria were altered to remove the vasopressor requirement of greater than or equal to 0.1 mcg/kg/min NEE. Twenty-eight patients were originally excluded for having a vasopressor requirement <0.1 mcg/kg/min NEE at ketamine initiation despite meeting all other inclusion criteria. Sixty-nine patients were included in this secondary analysis. The results were found to be similar to the results of the original primary analysis. There was a decrease in mean vasopressor requirements across all time points compared to baseline. However, there was only a statistically significant difference when comparing baseline mean vasopressor requirements to 6 hours (P = .003), and baseline to 24 hours (P = .049) after ketamine initiation. The mean vasopressor requirements comparing baseline to 12 hours after ketamine initiation was not found to be statistically significant (P = .077). This analysis can be found in Table 6.
Secondary Analysis Without Vasopressor Dose Requirement.
Discussion
The findings of this study suggest that using ketamine as an adjunct sedative in mechanically ventilated, critically ill patients can lower total vasopressor requirements both shortly after initiation as well as continually up to 24 to 48 hours after the initiation of a continuous ketamine infusion. There was a significant difference in vasopressor requirements at baseline compared to a composite endpoint of 6, 12, and 24 hours after initiation of ketamine. When this data was further broken down comparing each time point there was a significant difference between baseline and 6 hours (P = .002), and baseline and 24 hours (P = .022). In the baseline to 48 hour secondary analysis there was also a significant difference (P = .005). This suggests that ketamine can have a direct effect on vasopressor requirements likely through catecholamine release following ketamine administration. The statistical significance shown at both the 24 and 48 hour time points demonstrates that the catecholamine release associated with ketamine may have a longer duration of action instead of the previously noted transient effect on hemodynamics that ketamine can cause. The difference between vasopressor requirements from baseline and 24 hours was 0.073 mcg/kg/min NEE, which is larger than the initial infusion rate in NEE of phenylephrine, epinephrine, or norepinephrine at the study institution. It is also approximately the same vasopressor requirement as vasopressin at 0.3 units/minute. The difference from baseline to 48 hours was 0.095 mcg/kg/min NEE. This indicates that on average, ketamine may give the ability to stop an entire vasopressor agent 24 hours after the initiation of the continuous infusion ketamine. Based on the data shown in this study, ketamine may be a very valuable adjunct sedative agent for hemodynamically unstable patients with high vasopressor requirements.
The secondary analysis conducted strengthens the findings of the primary outcome. This data shows a similar trend that ketamine can lower vasopressor requirements in patients requiring any hemodynamic support, not just those at a dose ≥0.1 mcg/kg/min NEE. The main difference between the primary and secondary analysis is the difference of mean vasopressor requirement reduction. Due to the mean baseline vasopressor requirement being much lower in the secondary analysis versus the primary analysis (0.156 vs 0.218 mcg/kg/min NEE), the overall change in vasopressor requirements from baseline is reduced. There was still a statistically significant difference however at the baseline to 6 hour, and baseline to 24 hour time points, similar to that of the primary analysis. This consistent finding once again highlights that ketamine may have both a short-term and extended effect on vasopressor requirements after initiation.
Additionally, the number of vasopressor agents decreased at each time point. However, this decrease in mean number of vasopressor agents was only statistically significant when comparing baseline to 24 hours, and baseline to 48 hours. This data is showing that patients are being able to be weaned down on vasopressor dose in the short interval after ketamine initiation and potentially weaned completely off vasopressor agent(s) at the 24 and 48 hour time points, which is consistent with the findings of the primary outcome.
Furthermore, after ketamine initiation there was a decrease in the number of sedatives at all time points. Once again, similar to the number of vasopressor agents, this data was only statistically significant when comparing baseline to 24 and 48 hours. Meanwhile, the mean fentanyl dose decreased at all time points compared to baseline, but was statistically significant at all time points. This data coincides with the statistically significant difference in ketamine infusion rates between baseline and 24 hours. Clinically, the addition of ketamine and further uptitration of the infusion would allow the weaning of other sedatives depending on the goal level of sedation. The addition of a ketamine infusion in patients requiring high levels of continuous opioid infusions could be particularly helpful in patients who would benefit from minimal respiratory drive inhibition but also provide adequate sedation.
This study has several limitations including its retrospective nature, small sample size, single-center location, inclusion of COVID-19 patients, and variations of goal sedation level. The inclusion of COVID-19 patients may have impacted the external validity of the study due to differences in immune response, preexisting comorbidities, ICU length of stay, and mean ventilator days compared to other critically ill populations. Additionally, numerous patients included in the study with and without a COVID-19 diagnosis developed Acute Respiratory Distress Syndrome resulting in the use of paralytic agents. For these patients, a deeper level of sedation would be desired to prevent awareness of paralysis. This could result in increased use and increased doses of numerous sedative agents, some of which could impact hemodynamics.
Conclusion
Based on the results of this study, continuous infusion ketamine can reduce mean vasopressor requirements when used as an adjunct sedative in mechanically ventilated patients. This study showed that this vasopressor reduction can be seen as soon as 6 hours after initiation and up to 48 hours after initiation of a continuous infusion. This research supports the use of ketamine in patients who have high vasopressor requirements for hemodynamic support who also need sedation for mechanical ventilation. More studies should be conducted in this area with a larger sample size and fixed ketamine dosing to attempt to further understand the effect ketamine has on hemodynamics.
Footnotes
Ethical Considerations
Due to the retrospective nature of this study exempt status was granted by the Catholic Health Initiatives Institutional Review Board (IRB).
Author Contributions
Noah Gann, PharmD: Primary author. Patrick Ratliff, PharmD, BCPS, BCCCP: Conceptualization, methodology, formal analysis, supervision. Avery Guerrero, PharmD: Conceptualization, methodology, formal analysis, supervision. Robert Hughes, PharmD, BCPS: Conceptualization, methodology, formal analysis, supervision. Lauren Cottingham, PharmD, BCGP: Conceptualization, supervision.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article. The work was supported using the resources available at Saint Joseph Hospital located in Lexington, Kentucky.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.