Abstract
Background Ketamine has been shown to decrease sedative requirements in intensive care unit (ICU). Randomized trials are lacking on patient-centered outcomes. We aimed to compare the clinical characteristics and outcomes of mechanically ventilated adult ICU patients receiving ketamine as an adjunct analgosedative with those receiving standard of care (SOC) alone. We also described the feasibility during COVID-19 pandemic.
Methods In this randomized, open-label trial either ketamine or SOC, in a 1:1 ratio, was administered to patients who were intubated within 24 hours (medical, surgical, or transplant/oncology ICUs), expected to require mechanical ventilation (MV) for the next calendar day, and had the institutional pain and sedation protocol initiated. Ketamine infusion was 2 μg/kg/min on day 1 and 1 μg/kg/min on day 2. The primary outcome was the 28-day MV duration and ventilator-free days as co-primary outcome. Cox-proportional regression analysis was used to assess factors associated with probability for weaning off MV.
Results A total of 83 patients (43 in SOC and 40 in ketamine) were included. Demographics were balanced between the groups. The median duration of MV was not significantly different between the groups [median (interquartile range): 7 (3-9.25) for ketamine and 5 (2-8) for SOC, p= 0.15]. The median ventilation-free days was 19 days (IQR 0-24.75) in the ketamine and 19 days (IQR 0-24) in the SOC (p=0.70). Surgical and transplant/oncology ICU patients had a higher probability of weaning off MV than those in medical ICU [hazard ratio (95% confidence interval): 2.09 (1.06–4.14) for surgical ICU, 2.11 (1.02–4.35) for transplant/oncology ICU]. More patient was at goal RASS in ketamine compared to SOC. The sedatives and vasopressors cumulative doses were similar between the two arms at 48 hours. We found no difference in 28-day mortality rate, ICU and hospital length of stay, and hemodynamic changes. The consent rate was adequate and the protocol adherence rate was 97.5%.
Conclusions Ketamine as an adjunct agent for sedation did not decrease the duration of MV and appeared to be safe, feasible, and effective in subgroups of ICU patients. No effect was noted in sedative and pressors requirements, or on hemodynamics.
Trial registration ClinicalTrials.gov: NCT04075006 and current-controlled trials: ISRCTN14730035
Introduction
Analgo-sedation or analgesia-first-sedation has gained popularity in recent years [1]. This approach has been developed to avoid or decrease the use of sedatives, and facilitate mechanical ventilation (MV) weaning [2]. Ketamine has gained much interest in recent years in critical care settings because of its unique characteristics [3,4]. It acts by inhibiting N-methyl-D-aspartate (NMDA) receptors and the activation of the opioid μ- and κ-receptors. It is an attractive agent in intensive care unit (ICU) because favorable hemodynamic profile, preserves respiratory drive, and does not affect gastric motility [5]. Anesthetists have been using ketamine for years for acute and chronic pain, procedural sedation, and rapid sequence intubation. It has been used in postoperative pain control in surgical and trauma patients [as part of multimodal opioid-sparing analgesia in enhanced recovery after surgery (ERAS)], as an analgo-sedative agent in mechanically ventilated patients, status asthmatics, status epilepticus, alcohol withdrawal, and agitation [6,7]. Its use has been extended during the coronavirus disease (COVID-19) pandemic due to a shortage of other sedatives. It has been shown to decrease the use of fentanyl, propofol, and dexmedetomidine, and reduce opioid-induced hyperalgesia and opioid tolerance [8,9,10,11]. However, the evidence provided in the Pain, Agitation-Sedation, Delirium, Immobility, and Sleep Disruption (PADIS) guideline in support of adjunct ketamine as an analgo-sedative in mechanically ventilated patients was insufficient due to the limited number of randomized controlled trials (RCTs)[12]. Data on whether ketamine has an effect on patient-centered outcomes and its safety in RCTs, as compared with standard of care (SOC), are unclear. We sought to compare the clinical characteristics and outcomes of Analgo-sedative adjuncT keTAmine Infusion iN Mechanically vENTilated ICU patients (ATTAINMENT trial) compared to SOC alone, using a randomized trial design. We also described the feasibility of conducting this trial as a non-pandemic critical care research during the COVID-19 pandemic.
Methods
The ATTAINMENT trial was registered with ClinicalTrials.gov: NCT04075006, current controlled trials: ISRCTN14730035, and Saudi Food and Drug Authority: SCTR #19063002. The full study protocol has been published previously [13]. The study was conducted according to the Consolidated Standards of Reporting Trials (CONSORT) guidelines.
Design and setting
The trial was an investigator-initiated, single-center, prospective, parallel-group, randomized open-label clinical trial. Participants were recruited from the King Faisal Specialist Hospital and Research Center (KFSH&RC). It is located in Riyadh and is a major referral center that provides tertiary and quaternary care, receiving cases from all around the country of Saudi Arabia. The ICU department is composed of several intensive care units (medical, surgical, transplant/oncology). During the COVID-19 pandemic, new units were opened to accommodate the surge of patients. Recruitment began in September 2019 and was completed in November 2020.
Participants
Patients were eligible for inclusion if they were admitted to any of our three adult ICUs, if MV had been initiated within the previous 24 hours and was expected to continue on the next calendar day, initiated on institutional pain and sedation protocol, and no objection from ICU attending or primary treating team. Patients were excluded if they had a history of dementia or psychiatric disorders, or were comatose on admission due to hepatic encephalopathy. The full inclusion and exclusion criteria are detailed in Supplementary Appendix (Table S1). Other exclusion criteria were considered: pulmonary hypertension, tracheostomy at baseline, ketamine administration at baseline prior to randomization, intellectual disability which precluded delirium assessment, transfer from an external hospital, history of substance abuse, and situations where high blood pressure could trigger dangerous complications such as aortic dissection. We also excluded those with repeated ICU admissions within the same hospital visit and those who participated in another interventional trial. We recorded all patients who were eligible but not randomized for any of the following reasons: (1) patient or next of kin declined consent; (2) patient was unable to consent and the next of kin was not available (missing consent); and (3) ICU physician or primary team declined the patient’s enrolment.
Ethics approval and consent
The study was conducted in accordance with the principles of the Declaration of Helsinki. Our institution’s research ethics committee (REC) approved the study through the Research Advisory Council (RAC) number 2191 187. Given the need to enroll patients in an expedited manner within the 24 hours window, verbal consent from a guardian/next of kin was allowed and documented in the chart. Written consent was obtained as soon as the next of kin became available. During the COVID-19 pandemic, where patients’ family visits were prohibited, we were unable to obtain written consent. Therefore, we granted approval from our research ethics committee for the following consenting process: verbal consent from the family over the phone or a conference call to explain the nature of the study, as detailed in the written informed consent. A witness (not part of this study) was present during the verbal consenting process and documentation was completed in the patient chart.
Randomization
Patients were randomized in a 1:1 allocation using a computer-generated, pre-determined randomization list created by an independent biostatistician; no stratification was performed. The overall study population included all patients who were enrolled, randomly assigned, and received at least one dose of the study medication, constituting the modified intention-to-treat (mITT) population. Confirmation of the inclusion and exclusion criteria by local principal investigators in an electronic screening form in Research Electronic Data Capture (REDCap). Group allocation was concealed until after randomization. The investigators and all others involved in the trial were masked to the outcome data during the trial. Since this is an open-label study and because of lack of funding, blinding the investigators and the treating team was not possible at this phase. However, patients, and families were unaware of the group assignments.
Trial interventions
After randomization, both groups were started on the KFSH&RC ICU sedation and analgesia protocol. This is the SOC for all units in our practice. Since it was a nurse-driven protocol, the treating team placed an order regarding the target Richmond Agitation-Sedation Scale (RASS), and the nurse adjusted the infusions of other sedatives according to the protocol. The control group was called SOC and was continued on the sedation protocol only. The intervention group, “The Ketamine group,” was started on ketamine continuous infusion for 48 hours. During the first 24 hours, ketamine was administered at a rate of 2 μg/kg/min (0.12 mg/kg/h). During the following 24 hours, ketamine was administered at a rate of 1 μg/kg/min (0.06 mg/kg/h). The infusion was stopped at 48 hours or earlier if sedation was weaned off in preparation for patient extubation. The dose of ketamine was reported in μg per kilogram of actual body weight per min as per institutional practice. When the patient was deemed to be excessively sedated, the other sedatives were held first (or decreased) until the subject reached the team’s desired RASS goal. In situations where excessive sedation persisted, ketamine was discontinued. When the patient was deemed to be agitated, the other sedative rate was increased, and boluses or dexmedetomidine infusion were added, as required. Patients who were mechanically ventilated beyond the 48 hours study drug period were then sedated according to the standard practice of the KFSH&RC ICU.
Standard treatment common to both arms
The other aspects of care, including MV, fluid management, vasopressor use, blood products, enteral nutrition, renal replacement therapy, and early mobilization at the discretion of the treating team, were similar in both the groups. Other ICU supportive measures were applied as appropriate, including prone positioning of patients who met the criteria based on established guidelines for patients with ARDS and venous thromboembolism (VTE) prophylaxis. Septic patients with the expected source of infection were managed according to the latest survival sepsis campaign guidelines. Patient-ventilator asynchrony was systematically assessed and managed through inter-professional collaboration, prioritizing analgesia, and management of MV and respiratory-drive related factors to avoid the unnecessary use of neuromuscular blocking blockers [NMB]. The spontaneous awakening trial (SAT) was assessed every morning with a SAT safety screen, unless the patients were receiving sedative infusion for status epilepticus or undergoing sedative dose escalation due to ongoing agitation, or were started on NMB post randomization. Patients who passed the SAT were immediately managed using the spontaneous breathing trial (SBT) protocol. Both groups received a basic analgesic regimen that included paracetamol and epidural analgesia for hyperthermic intraperitoneal chemotherapy (HIPEC) patients. If delirium treatment was needed, non-pharmacological measures (reassurance or mobilization, and family support) was applied first. If not sufficient, the protocol allowed the use of antipsychotic agents and the decision was left to the ICU attending physician.
Outcomes and Variable Definitions
Clinical
The primary outcome was the median duration of MV, with ventilator-free days (VFDs) up to day 28 as the co-primary outcome. This outcome was chosen as patient-centered and highly influenced by mortality [14]. Duration of MV was recorded as either the number of calendar days from intubation to extubation or until ICU discharge or death, whichever occurred first. Successful extubation was defined as the absence for the need of reintubation within 48 hours. The secondary outcomes included the following up to 28 days: the ICU and hospital length of stay (LOS), mortality rate, and the percentage of participants with adverse events (AEs). We collected the daily cumulative dose of sedatives and analgesics [fentanyl, propofol, midazolam, and dexmedetomidine] and of antipsychotic medications over 48 hours post-randomization. Data only on sedatives administered in the ICU were collected while those given outside the ICU setting during anesthesia or intraoperative management were not collected. The presence of delirium was assessed using the confusion assessment method for the ICU (CAM-ICU), which was measured at baseline and 48 hours post randomization. If the CAM-ICU scores were not available, an electronic progress note was reviewed to detect any evidence of delirium. Hemodynamics changes were assessed until 48 hours after randomization included tachycardia, hypertension, and hypotension. Tachycardia was defined as a heart rate > 130 beats per minute, and hyper or hypotension were classified as systolic blood pressure ≥ 180 mmHg and ≤ 90 mmHg, respectively. Data were stored online in REDCap web application and data quality assessments were executed routinely to ensure completion and accuracy. Details about the variables extracted from the electronic medical record were reported previously [13].
Feasibility and process of care
Feasibility was assessed by evaluating the consent rate, recruitment success, and protocol adherence. Consent rate was deemed to be adequate if > 70% of surrogate decision makers (SDMs) or patients chose to participate upon being approached. Successful recruitment was defined as > 3 patients enrolled per month. Protocol adherence was defined as patients receiving ketamine according to the prescribed protocol if they remained ventilated for 48 hours for >75%. This threshold was chosen after examining other pilot studies on complex interventions. We conducted educational sessions for clinicians, nurses, and hospital pharmacies to facilitate the implementation of the protocol [9]. Protocol deviation was defined as not starting ketamine immediately after randomization (ideally within 4 hours) due to pharmacy delay or non-placement of the order.
Sample size calculation and Statistical analysis
The study was designed to recruit 40 ketamine-treated and 40 untreated subjects for analysis (i.e., total sample size 80). A statistician blinded to the study group allocation conducted all analyses based on the mITT principle. Based on previous evidence, the median time for extubation for patients with ketamine is 1.44 days (IQR 0.58– 2.66) and 2.44 (IQR 1.60 -4.00) for patients without ketamine[15]. A sample size of 80 was determined to have 80% power to detect a clinically meaningful and achievable reduction of one day less in the Ketamine group compared to the SOC, with two sided α = 0.05. Extra patients were recruited towards the end of the study to make up for the missing patients. Statistical analyses were performed using R statistical software Version 3.5.0 (The R Foundation for Statistical Computing, Vienna, Austria) and SAS/JMP version 15.0 (SAS Institute, Cary, NC). Counts and percentages were used to summarize the distribution of categorical variables. Continuous variables were summarized using either mean ± standard deviation of the median/interquartile range, based on the results of normality testing (performed using the Shapiro-Wilk test). Histograms were also checked for normality. The chi-square test of independence was used to assess whether the distribution of categorical variables was significantly different between the groups. The unpaired t-test or Mann-Whitney test was used to compare the distribution of normal and non-normal continuous variables. Variables with more than one possible response were dummy coded, and the percentage of each response was calculated from the total sample size. The analysis was performed separately for each response after dummy coding (changing it to yes/no variables). Statistical analysis was performed on VFDs after assigning VFD=0 for patients who died during 28 days and were calculated by subtracting the number of ventilation days from 28.
Kaplan-Meier curve with a corresponding log-rank test was used to estimate the probability of weaning-off MV. Patients who were not weaned-off MV during their ICU stay were censored at the last follow-up date. Kaplan-Maier estimates with a corresponding log-rank test were also used to compare the survival probability, hospital and ICU LOS between patients who received ketamine and patients who did not. Patients who alive or stayed in the ICU were censored at the last follow-up date. Cox-proportional regression analysis was used to assess factors associated with probability for weaning off MV and was reported as hazard ratio (HR) and 95% confidence interval (CI). The adjustment of observed effects was undertaken using a list of a priori defined covariates (age, Acute Physiology And Chronic Health Evaluation II (APACHE-II) score, Sequential Organ-Failure Assessment (SOFA) score, comorbidities, and ICU type]. Multivariate logistic regression was used to assess factors associated with all-cause mortality at 28-days with estimated odds ratios (ORs) and 95% CIs. The model was adjusted for age, the APACHE-II score, SOFA score, admission group (medical, surgical, or transplant), infection, metabolic acidosis, urea concentration, and the duration of MV. Backward stepwise logistic regression was used to eliminate variables that were not statistically significant. The likelihood ratio test was also used to assess whether the reduced model was significantly different from the full model (one with all the covariates). Sensitivity analysis for sedatives and vasopressors requirements excluding patients started on NBM post randomization was conducted. No imputation was done as there were no missing variables.
Study management and data monitoring
We ensured immediate data entry and identified missing data quickly, and issues were resolved in a timely manner. We used random source data verification to monitor data integrity and ensure data fidelity. An independent Research Advisory Council at our institution served as a Data Safety Monitoring Committee (DSMC) and reviewed all adverse events (including all deaths) [13].
Results
Participants
From September 2019 through November 2020, a total of 437 MV patients were screened; 83 patients met the inclusion criteria and 354 were excluded. Among the screened patients, 88 did not meet the eligibility criteria mainly because they were expected to require MV for < 24 hours. Among the included patients, 43 were in the SOC group and 40 were included in the ketamine group in the mITT analysis. The flow of the participants through the trial is shown in Figure 1. Other exclusion criteria assessed were pulmonary hypertension (3 patients), tracheostomy at baseline post-face flab or due to subglottic stenosis (11 patients), ketamine at baseline prior to randomization (1 patient), intellectual disability that precluded the delirium assessment (two patients), transferred from an external hospital (two patients), and history of substance abuse (three patients).
¶ deemed to be extubatable post randomization Abbreviations: ESLD, end stage liver diseases; HR, heart rate; SBP, systolic blood pressure; ECMO, Extracorporeal membrane oxygenation; DNR, do-not-resuscitate
Overall Characteristics
Overall, the median age was 60 years, with a higher proportion of males. About half of the patients were from the medical ICU, 26.5% from the surgical ICU, and 25.3% from the transplant/oncology ICU; median SOFA 8, and APACHE II, 20. The demographic characteristics were balanced between the groups, and the patients were similar in terms of sex, race, and most comorbidities, except for the prevalence of chronic obstructive pulmonary disease, which was higher in patients who receive SOC. We included a wide variety of ICU admission diagnoses and among those randomized to ketamine, 60% had ARDS and about 34% were recipients of solid organ transplant and solid malignancy. Other primary reasons for ICU admission included HIPEC (three patients: two in the ketamine group and one in the SOC group), COVID-19 pneumonia (two patients: one in each arm), Sickle cell disease (one patient in the ketamine arm). Ketamine-treated patients were noted to have higher lactate levels (median 2.2, IQR 1.58-3.4 mmol/L, p= 0.004) (Table 1). The median number of hours of ICU admission before study enrollment was 13 in SOC (IQR 6, 21.15) and 15 (IQR 12, 21) in ketamine treated patients (p=0.17). Post randomization, NMB was initiated in % 12.5 ketamine-treated patients compared to 4.65 % in SOC group.
Demographic and baseline characteristics
Clinical Outcomes
Primary outcome
The median duration of MV on day 28 was 7 days in the ketamine group (IQR 3-9.25) compared to 5 days in the SOC group (IQR 2-8) (Table 2). Among the 83 patients assessed, a similar proportion of patients had come off MV at 28 days: 25 of 40 (62.5%) in the ketamine group and 27 of 43 (62.8%) in the SOC. The distribution of VFDs at day 28 in the ketamine arm was not significantly different compared to the SOC (median 19, IQR: 0–24.75 days vs. median 19, IQR: 0–24 days, respectively; P = 0.70). Kaplan-Maier estimates for weaning off MV showed no statistically significant difference in the probability of the weaning off MV between the groups (P = 0.9) (Figure 2,A). The Cox-proportional hazard regression model showed that a lower probability of weaning off MV in patients with metabolic acidosis (HR [95% CI]): 0.48, [0.24-0.94], P = 0.032), whereas ICU admission for surgical reasons and transplant or oncology reasons was associated with a higher probability of weaning off MV [HR (95% CI): 2.09 (1.06–4.14) for surgical ICU, 2.11 (1.02–4.35) for transplant/oncology ICU] compared to medical admission (Supplementary table S2).
Efficacy outcomes
Panel A for duration of mechanical ventilation; Panel B represents the overall ICU length of stay (days); Panel C represents the overall hospital length of stay (days); Panel D represents 28-day mortality. Red line represents the standard of care group and blue line represents the ketamine group. The 95% confidence interval was illustrated as a band around the time-to-event curves. SOC donates to standard of care
Secondary outcomes
The median RASS in our trial was -2 at baseline, which gradually increased to -1 during 48 hours post randomization indicating light sedation and ability of patients to make eye contact with verbal stimulation. More patient was at goal RASS in 24 and 48 hours in ketamine arm (67.5 % in 24 hours and 73.5 % in 48 hours) compared to SOC (52.4 % in 24 hours and 66.7 % in 48 hours). However, this was not statistically different between groups. In addition, no significant differences were noted in the median pain score throughout the 48 hours interval post-randomization (Table 2). There was no difference in the baseline values of other vasopressors and sedative requirements prior to randomization. Ketamine-treated patients were noted to have a higher amount of vasopressin prior to randomization (median 39.6, IQR 30.5-64.2 units, P = 0.053). The cumulative doses of fentanyl and other sedatives were similar between the two arms at 48 hours post-randomization. Similar trends were observed with the cumulative dose of vasopressors in mg at 48 hours post-randomization (table 3). Box plots were used to visualize the distribution of cumulative doses for various sedatives and vasopressors and available in Supplementary figure S1, S2. Sensitivity analysis was conducted on sedative requirements, excluding those started on NMB post-randomization, and the findings were consistent with the primary analysis; no difference was observed between the two arms (Supplementary table S3). The median duration of hospital LOS and ICU LOS were comparable between the groups. Kaplan-Maier estimates were used to compare the hospital and ICU LOS between patients and the results showed that the probabilities of ICU discharge and hospital discharge were not significantly different between the groups (Figure 2,B-C).
Sedatives and vasopressors requirements
Table 4 illustrates the safety outcome. The proportion of patients who did not completed 48 hours of the trials was significantly different between groups (P = 0.01) and was higher in ketamine (37.5%) compared to SOC (11.63%). The main difference was weaning off sedation in preparation of extubation. The prevalence of delirium and hallucinations was higher in patients who received ketamine than those who did not (10% vs. 0%, P = 0.05). Thirty-six patients underwent CAM-ICU assessment (43.37 %), of which 2 were positive within 48 hours after randomization. However, this does not translate to either an increased use of antipsychotic agents, which was started in 3 ketamine-treated patients compared to 4 patients in the SOC arm, or increased dexmedetomidine initiation within 48 hours after randomization in the ketamine arm. The rates of re-intubation and tracheostomy 28 days post-randomization were not different between the two arms (P = 0.072). The median Clinical Pulmonary Infection Score (CPIS used to differentiate secretions caused by patients’ underlying lung pathology vs ketamine-associated hypersalivation) within the first 48 hours of randomization was higher in SOC (4, IQR 3 – 6) indicating a higher proportion of patients with ventilator-associated pneumonia, thus, higher frequency of hypersalivation and frequent suctioning in the SOC arm. With regard to the hemodynamic changes in HR and MAP at baseline, 24 hours, and 48 hours, we found no statistical difference between the patients randomized to ketamine as compared to SOC (figure 3).
Safety outcomes
SOC donates to standard of care. The middle black line is the median. The lower and upper lines of the box are the 25th and 75th percentile, respectively. Check this image. The lower and upper lines represent that percentile multiplied by 1.5 the Interquartile range
Abbreviation: HR, heart rate; MAP, mean arterial pressure
The 28-day mortality rate was not statistically different between group (Figure 2, D). The DSMC reviewed all deaths, and all were determined to have been due to underlying disease, with participation in the trial not being a contributing factor. Multivariate logistic regression was used to assess the factors associated with all-cause mortality at 28-days, showing that higher age was associated with a higher risk of mortality (OR [95% CI]: 1.04, [1.00 – 1.07], P= 0.04). Mortality was lower in patients admitted to the surgical ICU than in those admitted to the medical ICU (OR [95% CI]: 0.11, [0.02 – 0.64], P=0.016). None of the remaining factors were associated with 28-days mortality (Supplementary table S4).
Feasibility outcomes
The average patients enrollment was 3-4 patients/month. The consent rate was adequate; more than 70% of SDMs or patients when approached for consent chose to participate. The recruitment rate significantly decreased during the COVID-19 pandemic and was withheld for 1 month. We resumed recruitment at a slower rate in March 2020, with an average of 1-2 patients/month. In total, 12% of the patients were enrolled outside traditional working hours, that is, on weekends. This process was facilitated through close collaboration with the on-call ICU physician, the research coordinator, and the research investigators. Protocol adherence was 97.5%. and the median hours from consent or enrollment until ketamine started was 4.25 hours [IQR 2.08 - 5.88]. The adherence rate was lower than expected (90%) during the COVID-19 pandemic, and we were able to improve compliance using strategies such as pre-study education sessions for the research and clinical staff and routine clinical reminders, including documentation in the patient’s charts.
Discussion
Our trial has attempted to quantify the potential vasopressor and opiate-sparing effects of ketamine, as well as assess patient-centered outcomes and feasibility for conducting larger RCTs. We demonstrated that ketamine, as an adjunct in a dose of 1-2 μg /kg/min in mechanically ventilated ICU patients, did not decrease the duration of MV and appeared to be safe, feasible, and effective in subgroups of ICU patients. No effect was noted in sedative and vasopressor requirements and there was no impact on hemodynamics. The prevalence of delirium and hallucinations was slightly higher in patients who received ketamine, however, this does not translate to the increased use of antipsychotic or dexmedetomidine initiation, but is more likely attributed to the limited assessment and documentation of CAM-ICU.
Data regarding the ideal sedative agent in mechanically ventilated hemodynamically unstable ICU patients are limited. The interest in ketamine’s favorable hemodynamic, analgesic, and adverse effect profile has driven providers to pursue its use as analgosedative agent. When compared with other sedatives in the ICU, ketamine has the fastest onset (within 30-40 sec) and a 15 min duration of action. There is an increasing body of literature supporting the use of continuous ketamine infusion. Groetzinger et al. evaluated 91 patients who received ketamine at a median dose of 6.8 μg/kg/min (0.41 mg/kg/hours) for a median duration of 2.8 days. Within 24 hours of initiating ketamine, other sedatives were reduced or discontinued in 63% of patients. Additionally, 61% of patients were within the target sedation agitation scale goal after ketamine initiation, compared with 55% of patients prior to ketamine initiation. No adverse events were noted, and no increase in the use of antipsychotic medications was reported [16]. Similar results were reported in a study of 104 patients who received ketamine at doses of 5-7 μg/kg/min. At 24 hours after ketamine initiation, there was a 20% relative reduction in the total doses of the analgesic-sedative infusions and the median percent time within the goal RASS improved from 7% to 25% [17]. Another study on 36 patients with blunt trauma requiring MV and sedation showed a decrease in the use of other sedatives such as opioids and propofol. However, it was also associated with an increase in the dose of dexmedetomidine used, from 0.7 μg/kg/hr to 0.9 μg/kg/hr (p = 0.002), and ziprasidone, from a median cumulative dose of 120 mg to 220 mg (p = 0.018), to achieve target sedation[18]. Ketamine does not appear to have the potential adverse effects of nonsteroidal anti-inflammatory drugs on the gastrointestinal tract (bleeding) and kidneys (acute kidney injury). Also, in contrast to opioids, ketamine does not have negative effects on the mu receptors of the gastrointestinal tract associated with the ileus. Therefore, it has been used as a multimodal opioid-free analgesic as part of the ERAS protocols [19,20,21]. Further studies to control acute pain in traumatic rib fractures of severely injured individuals at sub-anesthetic doses resulted in a reduction of the pain scale score and morphine equivalent dose [22,23]. A systematic review and meta-analysis addressed the efficacy and safety of non-opioid adjunctive analgesics for ICU patients. Ketamine resulted in decreased opioid consumption (mean difference, 36.81 mg less; 95% CI, 27.32–46.30 mg less; low certainty). The authors concluded that clinicians should consider using adjunct agents to limit opioid exposure and improve pain scores in critically ill patients [24].
Most of the previous studies had a limited focus on patient centered outcomes, such as the duration of MV, as the primary outcome favoring surrogate outcomes, such as sedation scores and changes in analgesics and sedatives, leaving a significant knowledge gap, which is reflected by the wide variation in the use of this agent as a sedative agent in ICUs. To the best of our knowledge, our trial was the first and largest RCT on ketamine that reported a patient-centered outcome as the primary outcome and included diverse ICU population. We chose the duration of MV because ketamine preserves pharyngeal and laryngeal protective reflexes, lowers airway resistance, increases lung compliance, and is less likely to cause respiratory depression in low and slow infusions[4]. In addition, we reported VFDs as a co-primary outcome, which assumes that interventions may shorten ventilator duration, and improve mortality, thereby increasing efficiency as an outcome measure. In our trial, ketamine did not decrease the duration of MV and the median VFDs in our cohort is consistent with the one reported in the MENDS2 sedation trial; adjusted median, 23.7 days in dexmedetomidine vs. 24 days in propofol; OR, 0.96; 95% CI, 0.74 to 1.26) [25]. This neutral effect on VFDs could be because the majority of our population was from the medical ICU (48.2% of the entire cohort) and had moderate ARDS with a median baseline PaO2/FiO2 ratio of 152. Patients with ARDS may be under-represented in analgesia/sedation studies, and currently recommended strategies may not be feasible with light sedation. Although we found a high proportion of surgical patients being weaned off MV compared to medical ICU patients, this is because of the patient population nature in the surgical ICU; the majority without respiratory disorders and were admitted for postoperative management. Moreover, by increasing pulmonary airway pressure, there is a theoretical concern that ketamine could aggravate pulmonary hypertension; therefore, it should be used cautiously in patients with this condition.
In our trial, ketamine was most commonly used in conjunction with fentanyl and propofol infusions, which aligns with the PADIS guideline recommendations. In our trial, the cumulative doses of fentanyl and other sedatives were similar between the two arms. This could be explained by the fact that the proportion of patients who did not complete 48 hours of the trials was significantly different between groups (P = 0.027) and was higher in patients who received ketamine (37.5%) than in those who did not. This could also be due to starting NMB in 5 patients randomized to ketamine compared to 2 patients in the SOC after randomization. In addition, the adherence rate to our protocol and sedative titration was lower than expected (90%) during the COVID-19 pandemic, due to staff shortage and re-assignment of ICU nurses to COVID-19 ICUs. Subsequently, newly hired non-ICU nurses were assigned to cover manpower shortages in non-COVID-19 ICUs and could be unaware of the study protocol. Hence, efforts to reduce concomitant sedatives with ketamine may be conservative. For midazolam, two patients who were randomized to ketamine had status epilepticus at baseline, which could explain the higher midazolam requirement in ketamine-treated patients. The findings of our trial are in contrast to the trial by Guillou et al. which showed a reduction in opioid consumption with low-dose ketamine infusion for 48 hours [26]. However, patients in this trial underwent postoperative abdominal surgery and were able to use patient-controlled analgesia. It is difficult to extrapolate these findings to mechanically ventilated patients who are unable to self-report pain, have a higher severity of illness, and require a deeper level of sedation, as in our trial.
Another question to address pertains to ketamine dosing for analgosedation. It is well known that the severity of critical illness influences drug pharmacodynamics and pharmacokinetics [27]. Sympathetic overstimulation, hemodynamic instability, and acute septic brain dysfunction negatively affect organ function and thus absorption, distribution, metabolism, and drug dose-response relationships. It is a common finding that severely ill patients need much lower doses of sedatives to maintain deep sedation. Ketamine is highly lipophilic and is metabolized in the liver, generating active compounds (norketamine and hydroxynorketamine), and is eliminated almost entirely in the urine with an elimination half-life of 1.5-3 hours [28,29]. Published data for ketamine doses for analgosedative effects showed that it can be safely titrated up to 10-20 µg/kg/min, as needed, to achieve the desired level of analgosedation. We chose ketamine dosing as 1-2 µg/kg/minute because the majority of the ICU population included in our trial was relatively elderly (median age 61 years), with renal and hepatic dysfunction, which potentially alters the metabolism and excretion of ketamine and its active metabolite, resulting in increased sensitivity to continuous infusion ketamine, prolonged duration, drug accumulation, and a longer recovery process. Moreover, the dose described in this trial is in agreement with the existing literature describing the light sedation strategy and 2018 PADIS guideline recommendations [1,13]. This regimen is more conservative to minimize dose-related reactions, such as psychotomimetic episodes, that could lead to complex differential diagnoses in ICU patients who are prone to delirium and other CNS disturbances. We also did not observe notable severe confusion, nightmares, or emergence phenomena associated with ketamine use, which is consistent with the findings reported by Perbert et al. [30]. A retrospective cohort study evaluating the effect of ketamine versus non-ketamine-based sedation on delirium and coma in the ICU showed similar rates of delirium- and coma-free days in both groups (p = 0.25 and p = 0.51, respectively)[31]. Ketamine at these lower doses (< 5 μg /kg/min) does not appear to induce psychomimetic adverse effects that have been observed when using higher doses (>10 μg /kg/min). Although the mechanism is not completely clear, it has been hypothesized that the neurochemical basis for psychomimetic effects may be related to ketamine losing receptor selectivity at higher doses, leading to the depression of auditory and visual relay nuclei.
Ketamine itself has a sympathomimetic effect and can cause hypertension and tachycardia by acting as a catecholamine re-uptake inhibitor and has a negative inotropic effect. However, in a subgroup of patients, especially in the catecholamine-depleted state, it can sometimes cause hemodynamic compromise and hypotension[32]. It is best avoided in patients with a history of ischemic cardiac disease, hypertensive crisis or heart failure due to decreased cardiac index, increased pulmonary capillary wedge pressure, increased systemic vascular resistance, and myocardial depressant effect [33]. In our trial, we excluded patients with cardiogenic shock due to potential harm. Moreover, the addition of ketamine did not lower the vasopressor requirements, and not associated with clinically significant hemodynamic changes and appeared to be safe. We hypothesized that ketamine did not lower vasopressor requirements due to the lack of difference in dosing requirements of hemodynamic-altering sedative agents. In addition, ketamine-treated patients were sicker at baseline evident by higher lactate levels and higher vasopressin dose at baseline.
We noticed that the 28-day mortality rate in our cohort was 30.1% (32.6% in the SOC compared with 27.5% in ketamine treated patients], which is slightly higher than the mortality rate reported in old sedation trials; in the MIDEX trial, 21.1% midazolam patients and 27.3% dexmedetomidine patients died between randomization and follow-up at day 45, and in the PRODEX trial, 19.4% propofol patients and 17.1% dexmedetomidine patients died [34]. This was expected because we are a tertiary care hospital. Therefore, it is likely that our patients were sicker. The APACHE II and SOFA scores also suggest that these data were derived from a cohort of critically ill patients. This is comparable to the mortality rate in patients admitted to the ICU with severe sepsis and shock [35,36] and to the 90 day all-cause mortality rate reported in more recent sedation trials such as the SPICE III trial, which showed 29.1% in both the dexmedetomidine and usual-care groups, and the MENDS 2 trial, with 38% in the dexmedetomidine group vs. 39% propofol (HR, 1.06; 95% CI, 0.74 to 1.52) [25,37].
Even during the COVID-19 pandemic, we successfully completed the trial enrollment assessing the feasibility of conducting a larger multicenter trial. Achieving our threshold of recruitment and consent rate demonstrated that the trial is acceptable to patients, families, and clinicians. The major barriers faced were difficulties in continuing under lockdown conditions, infected research staff, shifting the staff to cover COVID-19 areas, and reorientation in clinical trial research towards COVID-19. We demonstrated that ketamine appears to be safe, and effective in subgroups of ICU patients, and may be considered as an alternative analgosedative agent. This is particularly important during the COVID-19 era, when the limited availability of many commonly used agents and medication shortages have necessitated the development of alternative strategies to keep patients on MV comfortable and synchronous[38]. Moreover, ketamine has several advantages over fentanyl. It has not been associated with chest wall rigidity precipitating insufficient ventilation, which has occasionally been described with fentanyl [39]. Additionally, propofol and dexmedetomidine associated-hypotension may necessitate the vasopressor support which may exclude patients from qualifying for COVID-19 antiviral medication (remdesivir), making ketamine an attractive alternative in those populations [40].
This study had several strengths. Firstly, it analyzed robust patient-centered outcomes, high rates of completed follow-up, and comprehensive assessments of AEs associated with ketamine use and its impact on hemodynamic response. We believe that our results provide incremental value in understanding the effects of ketamine. Randomization, blinded study statistician, and adherence to the mITT principle limit the potential sources of bias. Moreover, our trial included diverse ICU populations from the medical, surgical, transplant/oncology ICU, which potentially increased the trial generalizability and external validity. We also made every effort to include eligible patients within a narrow randomization window (within 24 hours of intubation) to eliminate potential confounders with other co-interventions.
Despite these strengths, we acknowledge the limitations of our trial. Although we enrolled our target sample size, it was small. Since it was a single center-study, the trial may not statistically powered enough to show a difference in MV duration or mortality between the trial groups. Moreover, we excluded neuro patients, such as those with severe traumatic brain injury (TBI) and hydrocephalus, which may limit the generalizability to the neurocritical care ICU patients. However, more recent systematic reviews of heterogeneous acute brain populations (subarachnoid hemorrhage, tumors, TBI) have concluded that ketamine causes only temporary changes in intracranial pressure without modifying cerebral perfusion pressure, and has no detrimental effect on ICU stay, outcomes, or mortality [41]. Furthermore, we did not collect detailed data on the frequency and duration of prone positioning for the ARDS patients who were made prone, or the median change in the PaO2/FiO2 ratio post randomization, limiting the determination of the real benefit of ketamine in oxygenation post-randomization. Although we made efforts to validate the diagnosis of delirium and delirium assessment with the CAM-ICU, we had a large proportion of patients (56.6 %) with un-assessed CAM-ICU, leaving a knowledge gap to be addressed by future trials. Finally, while all patients in our cohort were followed-up for 28 days, additional follow-up will be valuable to determine the full course of hospitalization and to characterize the long-term sequelae for survivors. Furthermore, the ketamine duration was limited to 48 hours and a longer duration with different ketamine dosing needs to be investigated by future multicenter trials.
Future studies should capture MV settings after randomization, co-interventions including, the use of bicarbonate, corticosteroids, prone position, and diuretics. Future studies should also consider looking at the analgosedative effect of ketamine in COVID-19 and neurocritical care ICUs. It was hypothesized that the anti-inflammatory effect of ketamine may have a beneficial effects in COVID-19 patients with pro-inflammatory activity compared to other sedatives. Finally, ketamine is not an expensive drug. Currently, no studies have evaluated the cost-effectiveness of ketamine in the management of sedation and analgesia and this should be considered in future studies.
Conclusions
Ketamine has been shown to be an attractive option for analgosedation. In our trial, ketamine did not decrease the duration of MV and appeared to be safe, feasible, and effective in subgroups of ICU patients. No effect was noted in sedative and pressor requirements and there was no negative impact on hemodynamics. However, this trial is not sufficiently powered to show a difference in important clinical outcomes, precluding definitive conclusions. While these data are encouraging and can help reform clinical practice, results from larger multicenter trials are awaited to shed light on the remaining questions and further investigate the association with patient-centered outcomes. Additional data are needed to further elucidate the role of ketamine in COVID-19 patients and as an alternative when resources are low and there is a shortage of essential sedative agents.
Data Availability
The datasets used and analyzed during the current report are available from the corresponding author on reasonable request.
Consent for publication
Not Applicable
Availability of data and material
The datasets used and analyzed during the current report are available from the corresponding author on reasonable request.
Competing interest
The authors declare that they have no conflicts of interest
Funding
This trial was investigator-initiated and all study authors are employees at King Faisal Specialist Hospital and Research Center (KFSH&RC), which has not provided any research grant for this particular project. All authors volunteered their time and used local resources to conduct the study.
Authors’ contributions
MB and MA: Conception and design, analytical plan, drafting of the manuscript, critical revision of the manuscript for important intellectual content, trial registration at Saudi FDA and clinicaltrial.gov, and approval of the final version to be published. KM, MH, RA, MR, KA, MS: acquisition, analysis, and interpretation of data, critical revision of the manuscript for important intellectual content, and approval of the final version to be published. ED: sample size calculation, analysis, and interpretation of data, assisted in submitting the documents to the Saudi FDA, and approved the final version to be published. M Altewerki, F Alkhaldi, S Alenazi, M Bano: acquisition and data collection, research coordinators, review final version to be published. We confirmed that the authorship followed the uniform requirements for manuscripts submitted to biomedical journals.
Authors’ information
MA: PharmD; BCPS, BCCCP; Critical Care Pharmacy Consultant; Medical and Surgical ICU; Adjunct Assistant Professor-Alfaisal University; Pharmaceutical Care Division (MBC 11); King Faisal Specialist Hospital & Research Center PO Box 3354, Riyadh 11211, Tel: +966114647272 Ext 70836.
MB: MD, FRCSC, FACS, Consultant Trauma and Acute Care Surgery, Critical Care Medicine, Department of Critical Care Medicine (MBC 94), King Faisal Specialist Hospital & Research Centre, P. O. Box 3354, Riyadh 11211, Tel: +966-11-216-2919, Ext: 24731
KM: MD, FCCP, MRCP (UK), Chairman and Consultant Intensivist, Department of Critical Care Medicine (MBC 94), King Faisal Specialist Hospital & Research Centre, P. O. Box 3354, Riyadh 11211, Tel: +966-11-442-4731.
MH: MD, FACP, FCCP, Consultant Intensivist, Department of Critical Care Medicine (MBC 94), King Faisal Specialist Hospital & Research Centre, P. O. Box 3354, Riyadh 11211, Tel: +966-11-464-7272 Ext 37165.
RA: PharmD, BCPS, BCCCP, critical care pharmacy specialist, medical and surgical ICU, pharmaceutical care division (MBC 11), King Faisal Specialist Hospital & Research Center, PO Box 3354, Riyadh 11211.
MS: MD, Assistant consultant Department of Critical Care Medicine (MBC 94), King Faisal Specialist Hospital & Research Centre, P. O. Box 3354, Riyadh 11211, Tel: +966-11-464-7272 Ext 37165.
KA: MD, Assistant consultant, Department of Critical Care Medicine (MBC 94), King Faisal Specialist Hospital & Research Centre, P. O. Box 3354, Riyadh 11211, Tel: +966-11-464-7272 Ext 37165.
MR: MD, Assistant consultant, Department of Critical Care Medicine (MBC 94), King Faisal Specialist Hospital & Research Centre, P. O. Box 3354, Riyadh 11211, Tel: +966-11-464-7272 Ext 37165.
ED: Chairman of Biostatistics and Epidemiology, Department of Biostats and Epidemiology, King Faisal Specialist Hospital & Research Centre, P.O Box 3354, Riyadh 11211, Tel: +966-11-464-7272 Ext 32504.
M Altewerki: Neurology Resident, Department of Neurosciences, Residency Training Program, King Faisal Specialist Hospital and Research Center, Riyadh, Saudi Arabia
F Alkhaldi, S Alenazi, M Bano: staff nurse I, Department of Oncology Liver Nursing, King Faisal Specialist Hospital and Research Center, Riyadh, Saudi Arabia
Supplemental Digital Content
Supplementary Data
Full inclusion and exclusion criteria
Cox-proportional regression analysis for weaning off mechanical ventilation
SOC donates to standard of care
SOC donates to standard of care
Sensitivity analysis for Sedatives and vasopressors requirements excluding patients started on atracurium post randomization
Factors associated with mortality at 28 days
Acknowledgments
We thank the ICU physicians, ICU nurses, ICU respiratory therapists, ICU physiotherapists, ICU satellite pharmacists, and ICU clinical pharmacists of KFSH&RC for their efforts to save the lives of ICU patients and their bravery in fighting against the COVID-19 pandemic. We also thank the Saudi Critical Care Trials Group for providing feedback on the study proposal and the participants and their families. Without their collective generosity, this trial would not have been possible.
List of Abbreviations
- APACHE II
- Acute Physiology and Chronic Health Evaluation II
- CRC
- Clinical Research Committee
- CAM-ICU
- Confusion Assessment Method for the ICU
- DNR
- Do-not-resuscitate
- ECMO
- Extracorporeal membrane oxygenation
- HR
- Heart rate
- ICU
- Intensive care unit
- IQR
- Interquartile range
- IRB
- Institutional Review Board
- KFSH&RC
- King Faisal Specialist Hospital and Research Centre
- LOS
- Length of stay
- MV
- Mechanical ventilation
- ORA
- Office of Research Affair
- PADIS
- Pain,Agitation,Sedation,Delirium, Immobility, and Sleep
- RCT
- Randomized controlled trial
- PF ratio
- Ratio of the partial pressure of arterial oxygen to the fraction of inspired oxygen
- REC
- Research Ethics Committee
- REDCap
- Research Electronic Data Capture
- RASS
- Richmond Agitation-Sedation Scale
- SOFA
- Sequential Organ Failure Assessment
- SAT
- Spontaneous awakening trial
- SBT
- Spontaneous breathing trial
- SBP
- Systolic blood pressure
References
