key: cord-0965472-x5gx0miu
authors: Amer, M.; Bawazeer, M.; Maghrabi, K.; Alshaikh, K.; Shaban, M.; Rizwan, M.; Amin, R.; De Vol, E.; Baali, M.; Altewerki, M.; Alkhaldi, F.; Alenazi, S.; Bano, M.; Hijazi, M.
title: Clinical Characteristics and Outcomes of Critically Ill Mechanically Ventilated Patients Receiving Adjunctive Ketamine for Sedation: an investigator-driven, open-label, randomized, controlled trial
date: 2021-04-29
journal: nan
DOI: 10.1101/2021.04.26.21256072
sha: 19095437203adbe567926a4514aaa9016594e199
doc_id: 965472
cord_uid: x5gx0miu

Background: Ketamine showed to decrease sedative requirements in intensive care unit (ICU). Randomized trials are lacking on patient centered outcome. We aimed to compare the clinical characteristics and outcomes of mechanically ventilated adult ICU patients receiving ketamine as an adjunct analgosedative agent 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 h (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 g/kg/min on day 1 and 1 g/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 h. 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

In addition, no significant differences were noted in the median pain score throughout the 48 hours interval post-282 randomization (Table 2) . There was no difference in the baseline values of other vasopressors and sedative 283 requirements prior to randomization. Ketamine-treated patients were noted to have a higher amount of 284 vasopressin prior to randomization (median 39.6, IQR 30.5-64.2 units, P = 0.053). The cumulative doses of 285 fentanyl and other sedatives were similar between the two arms at 48 hours post-randomization. Similar trends 286 were observed with the cumulative dose of vasopressors in mg at 48 hours post-randomization (table 3) . Box 287 plots were used to visualize the distribution of cumulative doses for various sedatives and vasopressors and 288 available in Supplementary figure S1, S2. Sensitivity analysis was conducted on sedative requirements, excluding 289 those started on NMB post-randomization, and the findings were consistent with the primary analysis; no 290 difference was observed between the two arms (Supplementary table S3) . The median duration of hospital LOS 291 and ICU LOS were comparable between the groups. Kaplan-Maier estimates were used to compare the hospital 292 and ICU LOS between patients and the results showed that the probabilities of ICU discharge and hospital 293 discharge were not significantly different between the groups (Figure 2,B-C) . 294 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 29, 2021. ; https://doi.org/10.1101/2021.04.26.21256072 doi: medRxiv preprint 13 Table 4 illustrates the safety outcome. The proportion of patients who did not completed 48 hours of the trials 295 was significantly different between groups (P = 0.01) and was higher in ketamine (37.5%) compared to SOC 296 (11.63%). The main difference was weaning off sedation in preparation of extubation. The prevalence of delirium 297 and hallucinations was higher in patients who received ketamine than those who did not (10% vs. 0%, P = 0.05). 298

Thirty-six patients underwent CAM-ICU assessment (43.37 %), of which 2 were positive within 48 hours after 299 randomization. However, this does not translate to either an increased use of antipsychotic agents, which was 300 started in 3 ketamine-treated patients compared to 4 patients in the SOC arm, or increased dexmedetomidine 301 initiation within 48 hours after randomization in the ketamine arm. The rates of re-intubation and tracheostomy 302 28 days post-randomization were not different between the two arms (P = 0.072). The median Clinical Pulmonary 303

Infection Score (CPIS used to differentiate secretions caused by patients' underlying lung pathology vs ketamine-304 associated hypersalivation) within the first 48 hours of randomization was higher in SOC (4, IQR 3 -6) indicating 305 a higher proportion of patients with ventilator-associated pneumonia, thus, higher frequency of hypersalivation 306 and frequent suctioning in the SOC arm. With regard to the hemodynamic changes in HR and MAP at baseline, 307 24 hours, and 48 hours, we found no statistical difference between the patients randomized to ketamine as 308 compared to SOC ( figure 3) . 309

The 28-day mortality rate was not statistically different between group (Figure 2, D) . The DSMC reviewed all 310 deaths, and all were determined to have been due to underlying disease, with participation in the trial not being 311 a contributing factor. Multivariate logistic regression was used to assess the factors associated with all-cause 312 mortality at 28-days, showing that higher age was associated with a higher risk of mortality ( CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 29, 2021. ; https://doi.org/10.1101/2021.04.26.21256072 doi: medRxiv preprint Feasibility outcomes 320

The average patients enrollment was 3-4 patients/month. The consent rate was adequate; more than 70% of 321 SDMs or patients when approached for consent chose to participate. The recruitment rate significantly decreased 322 during the COVID-19 pandemic and was withheld for 1 month. We resumed recruitment at a slower rate in March 323 2020, with an average of 1-2 patients/month. In total, 12% of the patients were enrolled outside traditional 324 working hours, that is, on weekends. This process was facilitated through close collaboration with the on-call ICU 325 physician, the research coordinator, and the research investigators. Protocol adherence was 97.5%. and the 326 median hours from consent or enrollment until ketamine started was 4.25 hours [IQR 2.08 -5.88]. The adherence 327 rate was lower than expected (90%) during the COVID-19 pandemic, and we were able to improve compliance 328 using strategies such as pre-study education sessions for the research and clinical staff and routine clinical 329 reminders, including documentation in the patient's charts. Our trial has attempted to quantify the potential vasopressor and opiate-sparing effects of ketamine, as 333 well as assess patient-centered outcomes and feasibility for conducting larger RCTs. We demonstrated that 334 ketamine, as an adjunct in a dose of 1-2 μg /kg/min in mechanically ventilated ICU patients, did not decrease the 335 duration of MV and appeared to be safe, feasible, and effective in subgroups of ICU patients. No effect was noted 336 in sedative and vasopressor requirements and there was no impact on hemodynamics. The prevalence of 337 delirium and hallucinations was slightly higher in patients who received ketamine, however, this does not 338 translate to the increased use of antipsychotic or dexmedetomidine initiation, but is more likely attributed to the 339 limited assessment and documentation of CAM-ICU. 340

Data regarding the ideal sedative agent in mechanically ventilated hemodynamically unstable ICU 341 patients are limited. The interest in ketamine's favorable hemodynamic, analgesic, and adverse effect profile has 342 driven providers to pursue its use as analgosedative agent. When compared with other sedatives in the ICU, 343 ketamine has the fastest onset (within 30-40 sec) and a 15 min duration of action. There is an increasing body of 344 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) Additionally, 61% of patients were within the target sedation agitation scale goal after ketamine initiation, 348 compared with 55% of patients prior to ketamine initiation. No adverse events were noted, and no increase in 349 the use of antipsychotic medications was reported [16] . Similar results were reported in a study of 104 patients 350 who received ketamine at doses of 5-7 μg/kg/min. At 24 hours after ketamine initiation, there was a 20% relative 351 reduction in the total doses of the analgesic-sedative infusions and the median percent time within the goal RASS 352 improved from 7% to 25% [17] . Another study on 36 patients with blunt trauma requiring MV and sedation 353 showed a decrease in the use of other sedatives such as opioids and propofol. However, it was also associated 354 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 355 ziprasidone, from a median cumulative dose of 120 mg to 220 mg (p = 0.018), to achieve target sedation [18] . 356

Ketamine does not appear to have the potential adverse effects of nonsteroidal anti-inflammatory drugs on the 357 gastrointestinal tract (bleeding) and kidneys (acute kidney injury). Also, in contrast to opioids, ketamine does not 358 have negative effects on the mu receptors of the gastrointestinal tract associated with the ileus. Therefore, it has 359 been used as a multimodal opioid-free analgesic as part of the ERAS protocols [19, 20, 21] . Further studies to control 360 acute pain in traumatic rib fractures of severely injured individuals at sub-anesthetic doses resulted in a reduction 361 of the pain scale score and morphine equivalent dose [22, 23] . A systematic review and meta-analysis addressed the 362 efficacy and safety of non-opioid adjunctive analgesics for ICU patients. Ketamine resulted in decreased opioid 363 consumption (mean difference, 36.81 mg less; 95% CI, 27.32-46.30 mg less; low certainty). The authors 364 concluded that clinicians should consider using adjunct agents to limit opioid exposure and improve pain scores 365 in critically ill patients [24] . 366

Most of the previous studies had a limited focus on patient centered outcomes, such as the duration of 367 MV, as the primary outcome favoring surrogate outcomes, such as sedation scores and changes in analgesics and 368 sedatives, leaving a significant knowledge gap, which is reflected by the wide variation in the use of this agent as 369 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 29, 2021. ; https://doi.org/10.1101/2021.04.26.21256072 doi: medRxiv preprint 16 a sedative agent in ICUs. To the best of our knowledge, our trial was the first and largest RCT on ketamine that 370 reported a patient-centered outcome as the primary outcome and included diverse ICU population. We chose 371 the duration of MV because ketamine preserves pharyngeal and laryngeal protective reflexes, lowers airway 372 resistance, increases lung compliance, and is less likely to cause respiratory depression in low and slow 373 infusions [4] . In addition, we reported VFDs as a co-primary outcome, which assumes that interventions may 374 shorten ventilator duration, and improve mortality, thereby increasing efficiency as an outcome measure. In our 375 trial, ketamine did not decrease the duration of MV and the median VFDs in our cohort is consistent with the one 376 reported in the MENDS2 sedation trial; adjusted median, 23.7 days in dexmedetomidine vs. 24 days in propofol; 377 OR, 0.96; 95% CI, 0.74 to 1.26) [25] . This neutral effect on VFDs could be because the majority of our population 378 was from the medical ICU (48.2% of the entire cohort) and had moderate ARDS with a median baseline PaO2/FiO2 379 ratio of 152. Patients with ARDS may be under-represented in analgesia/sedation studies, and currently 380 recommended strategies may not be feasible with light sedation. Although we found a high proportion of surgical 381 patients being weaned off MV compared to medical ICU patients, this is because of the patient population nature 382 in the surgical ICU; the majority without respiratory disorders and were admitted for postoperative management. 383

Moreover, by increasing pulmonary airway pressure, there is a theoretical concern that ketamine could aggravate 384 pulmonary hypertension; therefore, it should be used cautiously in patients with this condition. 385

In our trial, ketamine was most commonly used in conjunction with fentanyl and propofol infusions, 386 which aligns with the PADIS guideline recommendations. In our trial, the cumulative doses of fentanyl and other 387 sedatives were similar between the two arms. This could be explained by the fact that the proportion of patients 388 who did not complete 48 hours of the trials was significantly different between groups (P = 0.027) and was higher 389 in patients who received ketamine (37.5%) than in those who did not. This could also be due to starting NMB in 390 5 patients randomized to ketamine compared to 2 patients in the SOC after randomization. In addition, the 391 adherence rate to our protocol and sedative titration was lower than expected (90%) during the COVID-19 392 pandemic, due to staff shortage and re-assignment of ICU nurses to COVID-19 ICUs. Subsequently, newly hired 393 non-ICU nurses were assigned to cover manpower shortages in non-COVID-19 ICUs and could be unaware of the 394 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 29, 2021. ; https://doi.org/10.1101/2021.04.26.21256072 doi: medRxiv preprint study protocol. Hence, efforts to reduce concomitant sedatives with ketamine may be conservative. For 395 midazolam, two patients who were randomized to ketamine had status epilepticus at baseline, which could 396 explain the higher midazolam requirement in ketamine-treated patients. The findings of our trial are in contrast 397 to the trial by Guillou et al. which showed a reduction in opioid consumption with low-dose ketamine infusion 398 for 48 hours [26] . However, patients in this trial underwent postoperative abdominal surgery and were able to use 399 patient-controlled analgesia. It is difficult to extrapolate these findings to mechanically ventilated patients who 400 are unable to self-report pain, have a higher severity of illness, and require a deeper level of sedation, as in our 401 trial. 402

Another question to address pertains to ketamine dosing for analgosedation. It is well known that the 403 severity of critical illness influences drug pharmacodynamics and pharmacokinetics [27] . Sympathetic 404 overstimulation, hemodynamic instability, and acute septic brain dysfunction negatively affect organ function 405 and thus absorption, distribution, metabolism, and drug dose-response relationships. It is a common finding that 406 severely ill patients need much lower doses of sedatives to maintain deep sedation. Ketamine is highly lipophilic 407 and is metabolized in the liver, generating active compounds (norketamine and hydroxynorketamine), and is 408 eliminated almost entirely in the urine with an elimination half-life of 1.5-3 hours [28, 29] . Published data for 409 ketamine doses for analgosedative effects showed that it can be safely titrated up to 10-20 µg/kg/min, as needed, 410 to achieve the desired level of analgosedation. We chose ketamine dosing as 1-2 µg/kg/minute because the 411 majority of the ICU population included in our trial was relatively elderly (median age 61 years), with renal and 412 hepatic dysfunction, which potentially alters the metabolism and excretion of ketamine and its active metabolite, 413 resulting in increased sensitivity to continuous infusion ketamine, prolonged duration, drug accumulation, and a 414 longer recovery process. Moreover, the dose described in this trial is in agreement with the existing literature 415 describing the light sedation strategy and 2018 PADIS guideline recommendations [1, 13] . This regimen is more 416 conservative to minimize dose-related reactions, such as psychotomimetic episodes, that could lead to complex 417 differential diagnoses in ICU patients who are prone to delirium and other CNS disturbances. We also did not 418 observe notable severe confusion, nightmares, or emergence phenomena associated with ketamine use, which 419 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 29, 2021. ; https://doi.org/10.1101/2021.04.26.21256072 doi: medRxiv preprint is consistent with the findings reported by Perbert et al. [30] . A retrospective cohort study evaluating the effect 420 of ketamine versus non-ketamine-based sedation on delirium and coma in the ICU showed similar rates of 421 delirium-and coma-free days in both groups (p = 0.25 and p = 0.51, respectively) [31] . Ketamine at these lower 422 doses (< 5 μg /kg/min) does not appear to induce psychomimetic adverse effects that have been observed when 423 using higher doses (>10 μg /kg/min). Although the mechanism is not completely clear, it has been hypothesized 424 that the neurochemical basis for psychomimetic effects may be related to ketamine losing receptor selectivity at 425 higher doses, leading to the depression of auditory and visual relay nuclei. 426

Ketamine itself has a sympathomimetic effect and can cause hypertension and tachycardia by acting as 427 a catecholamine re-uptake inhibitor and has a negative inotropic effect. However, in a subgroup of patients, 428 especially in the catecholamine-depleted state, it can sometimes cause hemodynamic compromise and 429 hypotension [32] . It is best avoided in patients with a history of ischemic cardiac disease, hypertensive crisis or 430 heart failure due to decreased cardiac index, increased pulmonary capillary wedge pressure, increased systemic 431 vascular resistance, and myocardial depressant effect [33] . In our trial, we excluded patients with cardiogenic 432 shock due to potential harm. Moreover, the addition of ketamine did not lower the vasopressor requirements, 433

and not associated with clinically significant hemodynamic changes and appeared to be safe. We hypothesized 434 that ketamine did not lower vasopressor requirements due to the lack of difference in dosing requirements of 435 hemodynamic-altering sedative agents. In addition, ketamine-treated patients were sicker at baseline evident by 436 higher lactate levels and higher vasopressin dose at baseline. 437

We noticed that the 28-day mortality rate in our cohort was 30.1% (32.6% in the SOC compared with 438 27.5% in ketamine treated patients], which is slightly higher than the mortality rate reported in old sedation 439 trials; in the MIDEX trial, 21.1% midazolam patients and 27.3% dexmedetomidine patients died between 440 randomization and follow-up at day 45, and in the PRODEX trial, 19.4% propofol patients and 17.1% 441 dexmedetomidine patients died [34] . This was expected because we are a tertiary care hospital. Therefore, it is 442 likely that our patients were sicker. The APACHE II and SOFA scores also suggest that these data were derived 443 from a cohort of critically ill patients. This is comparable to the mortality rate in patients admitted to the ICU with 444 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 29, 2021. ; https://doi.org/10.1101/2021.04.26.21256072 doi: medRxiv preprint severe sepsis and shock [35, 36] and to the 90 day all-cause mortality rate reported in more recent sedation trials 445 such as the SPICE III trial, which showed 29.1% in both the dexmedetomidine and usual-care groups, and the 446 MENDS 2 trial, with 38% in the dexmedetomidine group vs. 39% propofol (HR, 1.06; 95% CI, 0.74 to 1.52) [25, 37] . 447

Even during the COVID-19 pandemic, we successfully completed the trial enrollment assessing the 448 feasibility of conducting a larger multicenter trial. Achieving our threshold of recruitment and consent rate 449 demonstrated that the trial is acceptable to patients, families, and clinicians. The major barriers faced were 450 difficulties in continuing under lockdown conditions, infected research staff, shifting the staff to cover COVID-19 451 areas, and reorientation in clinical trial research towards COVID-19. We demonstrated that ketamine appears to 452 be safe, and effective in subgroups of ICU patients, and may be considered as an alternative analgosedative 453 agent. This is particularly important during the COVID-19 era, when the limited availability of many commonly 454 used agents and medication shortages have necessitated the development of alternative strategies to keep 455 patients on MV comfortable and synchronous [38] . Moreover, ketamine has several advantages over fentanyl. It 456 has not been associated with chest wall rigidity precipitating insufficient ventilation, which has occasionally been 457 described with fentanyl [39] . Additionally, propofol and dexmedetomidine associated-hypotension may 458 necessitate the vasopressor support which may exclude patients from qualifying for COVID-19 antiviral 459 medication (remdesivir), making ketamine an attractive alternative in those populations [40] . 460

This study had several strengths. Firstly, it analyzed robust patient-centered outcomes, high rates of 461 completed follow-up, and comprehensive assessments of AEs associated with ketamine use and its impact on 462 hemodynamic response. We believe that our results provide incremental value in understanding the effects of 463 ketamine. Randomization, blinded study statistician, and adherence to the mITT principle limit the potential 464 sources of bias. Moreover, our trial included diverse ICU populations from the medical, surgical, 465 transplant/oncology ICU, which potentially increased the trial generalizability and external validity. We also made 466 every effort to include eligible patients within a narrow randomization window (within 24 hours of intubation) 467

to eliminate potential confounders with other co-interventions. 468 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 29, 2021. ; https://doi.org/10.1101/2021.04.26.21256072 doi: medRxiv preprint 20 Despite these strengths, we acknowledge the limitations of our trial. Although we enrolled our target 469 sample size, it was small. Since it was a single center-study, the trial may not statistically powered enough to 470 show a difference in MV duration or mortality between the trial groups. Moreover, we excluded neuro patients, 471 such as those with severe traumatic brain injury (TBI) and hydrocephalus, which may limit the generalizability to 472 the neurocritical care ICU patients. However, more recent systematic reviews of heterogeneous acute brain 473 populations (subarachnoid hemorrhage, tumors, TBI) have concluded that ketamine causes only temporary 474 changes in intracranial pressure without modifying cerebral perfusion pressure, and has no detrimental effect on 475 ICU stay, outcomes, or mortality [41] . Furthermore, we did not collect detailed data on the frequency and duration Table 1 : Demographic and baseline characteristics 520 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. 

The datasets used and analyzed during the current report are available from the corresponding author on 606 reasonable request. 607

The authors declare that they have no conflicts of interest 609

Funding 610

This trial was investigator-initiated and all study authors are employees at King Faisal Specialist Hospital and 611

Research Center (KFSH&RC), which has not provided any research grant for this particular project. All authors 612 volunteered their time and used local resources to conduct the study. 613 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 29, 2021. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 29, 2021. ; https://doi.org/10.1101/2021.04.26.21256072 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 29, 2021. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 29, 2021. ; Log-Rank P =0.608 Log-Rank P =0.838

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 29, 2021. ; https://doi.org/10.1101/2021.04.26.21256072 doi: medRxiv preprint is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 29, 2021. ; https://doi.org/10.1101/2021.04.26.21256072 doi: medRxiv preprint Data presented as n (%), mean ± sd, or median (interquartile range). SOC donates to standard of care . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 29, 2021. ; https://doi.org/10.1101/2021.04.26.21256072 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 29, 2021. ; https://doi.org/10.1101/2021.04.26.21256072 doi: medRxiv preprint

Continuous infusion of Ketamine for adjunctive sedation in medical intensive care unit: a case series

Ketamine Use in the Intensive Care Unit

Sedation of Mechanically Ventilated COVID-19 patients: Challenges and Special Considerations

The reemergence of ketamine for treatment in critically ill adults

Role of ketamine in acute postoperative pain management: a narrative review

A review on nonanesthetic uses of ketamine

Analgesia and Sedation Strategies in Mechanically Ventilated Adults with COVID-19

Ketamine for analgosedation in critically ill patients

Reappraisal of ventilator-free days in critical care research

Impact of Low-Dose Ketamine on the Usage of Continuous Opioid Infusion for the Treatment of Pain in Adult Mechanically Ventilated Patients in Surgical Intensive Care Units

Ketamine infusion for adjunct sedation in mechanically ventilated patients

Continuous infusion ketamine for adjunctive analgosedation in mechanically ventilated, critically ill patients

Impact of ketamine use on adjunctive analgesic and sedative medications in critically ill trauma patients

Low-dose ketamine infusion for analgesia during postoperative ventilator treatment

Ketamine versus hydromorphone patient-controlled analgesia for acute pain in trauma patients

Non-opioid analgesics as adjuvants to opioid for pain management in adult patients in the ICU: A systematic review and meta-analysis

Ketamine infusion for pain control in elderly patients with multiple rib fractures: Results of a randomized controlled trial

Ketamine infusion for pain control in adult patients with multiple rib fractures: Results of a randomized control trial

SOC donates to standard of care ¶ The Sequential Organ Failure Assessment (SOFA) is used to track organ failure in the ICU; scores range from 0 to 24, with higher scores indicating greater severity of illness ** The Acute Physiology and Chronic Health Evaluation (APACHE II) assesses the risk of

DELIRIC is a delirium prediction 24 hours after ICU admission to predict the chance to develop a delirium episode during the ICU stay. Abbreviations: AC, assist-control mode; ABG, arterial blood gas; APACHE II, Acute Physiology And Chronic Health Evaluation II; COPD, chronic obstructive pulmonary disease; CKD, chronic kidney disease; CVVH, continuous veno-venous hemofiltration

FiO2, fraction of inspired oxygen

Human immunodeficiency virus infection and acquired immunodeficiency syndrome

HSCT, hematopoietic stem-cell transplantation

Synchronized intermittent mandatory ventilation; SOFA, Sequential Organ Failure Assessment

Admitted to one of the following ICUs (Medical, Surgical, transplant/oncology or COVID-19)

• Patients with a history of dementia or psychiatric disorders or those on any antipsychotic or antidepressant medications at home

• Patients on dexmedetomidine as the primary sedative prior to randomization

• Patients with cardiogenic shock, acute decompensated heart failure, or myocardial infarction • History of end-stage liver failure

• Proven or suspected primary neurological injury (traumatic brain injury, ischemic stroke