The Reemergence of Ketamine for Treatment in Critically Ill Adults

Kimberly P. Hurth, PharmD, BCCCP; Anthony Jaworski, PharmD, BCCCP; Kristen B. Thomas, PharmD, BCPS; William B. Kirsch, PharmD, BCPS; Michael A. Rudoni, PharmD, BCPS, BCCCP; Kevin M. Wohlfarth, PharmD, BCPS, BCCCP, BCCP


Crit Care Med. 2020;48(6):899-911. 

In This Article

Potential Uses for Ketamine

Evidence for each off-label use is displayed within Table 2.

Acute Pain Management

Rationale. Many studies investigating ketamine for pain management include patients presenting to the emergency department (ED). There are limited pain management recommendations for ketamine in nonoperative settings in the ICU, outside of those resulting from a randomized, double-blind study, which demonstrated a reduction in morphine consumption (p < 0.05) without a significant difference in pain score at 24 or 48 hours compared with morphine.[11–13] Additionally, there was lower mean morphine equivalent use (p = 0.015) in a retrospective, case-control cohort involving 30 ICU patients.[14] Due to ketamine's effects on delta-, kappa-, and mu-opioid receptors in addition to NMDA inhibition, there is increasing interest in its utilization for acute pain.

Evidence. The majority of studies use a variety of ketamine doses, including weight-based and fixed-dosed, for pain,[15–23] showing it may reduce pain scores by at least 3 on a numeric pain scale out of 10.[19,21–24] Ketamine has demonstrated pain score reduction similar to morphine, although duration of effect may be shorter.[22–24]

Risks.Although multiple studies discussed report more adverse effects with ketamine compared with control groups, the low severity of these effects did not require an intervention, and many patients received concomitant opioids which could confound findings. Ahern et al[18] found that 53% of patients reported weak to modest side effects, and 34% reported bothersome effects. Most common side effects were dizziness, fatigue, nausea, and feelings of unreality. Bowers et al[20] reported more patients experienced adverse effects with ketamine compared with placebo (27 vs 12 patients), with the majority experiencing lightheadedness or dizziness (30%). Mitchell and Fallon[24] found that 33% of patients receiving ketamine compared with 6% in placebo reported feeling emotional at 24 hours, with no other differences in adverse effects.

Clinical Application. Despite the majority of evidence resulting from an ED population, low-dose ketamine for acute pain has demonstrated reduced pain scores. However, most included trials had a small sample size, prior receipt of opioids, were single-center, and used convenience sampling for enrollment. Given the quality of evidence from these studies, ketamine can reasonably be used as an adjunctive agent for acute pain in the ED (level 2 evidence); however, further evidence is needed to extrapolate results to the ICU setting. There are several ongoing trials continuing to investigate ketamine for acute pain management.[25–28]

Adjunctive Analgosedation in Mechanical Ventilation

Rationale. There are limited data on the use of ketamine as an adjunctive agent for sedation outside of procedural or general anesthesia, with no formal recommendation from guidelines.[1,4,5] Ketamine may be an ideal adjunctive agent in mechanically ventilated patients since it can reduce opioid requirements while not depressing hemodynamic parameters like alternative sedative medications.

Evidence. Buchheit et al[29] observed the use of subanesthetic doses of ketamine as an adjunctive option for intubated patients and found significant reductions in opioid and propofol requirements 24 hours post-ketamine introduction (p ≤ 0.001 and p = 0.02, respectively). Additionally, vasopressor requirements were significantly reduced 6 hours after initiating ketamine (p = 0.019), however, not at 24 hours (p = 0.236).[29] Another retrospective review of mechanically ventilated patients investigated ketamine as an alternative agent and found, the utilization of ketamine significantly reduced opioid and propofol requirements (p < 0.05).[30] A third retrospective cohort reduced or discontinued alternative sedatives in 63% of patients within 24 hours of adding ketamine infusion.[31]

Risks. Rates and severity of adverse effects were not reported in all included studies. No significant difference in blood pressure or respiratory rates (RRs) before and 24 hours after ketamine was noted; however, there was a significant increase in heart rate (p = 0.01).[29] Another study found no significant difference in heart rate after ketamine initiation.[31] However, reported adverse effects included dysphoria, nystagmus, tachycardia, and increased agitation. Only one study reported increased amounts of dexmedetomidine and ziprasidone following ketamine initiation.[30]

Clinical Application. Despite limited data, retrospective studies suggest efficacy and opioid-sparing properties of ketamine as an adjunctive sedative in mechanically ventilated patients (level 3 evidence). Considering its limited evidence, the role of ketamine as a continuous infusion should be reserved as an adjunctive agent in refractory situations. Given the retrospective study designs, small sample size, and varying dosing regimens, larger, prospective, randomized studies with traditional sedation are warranted to further assess clinical outcomes.

Status Asthmaticus

Rationale. Ketamine causes bronchodilation by increasing circulating catecholamines, inhibiting vagal outflow, and relaxing airway smooth muscle.[1,32] Given this effect, ketamine has been considered a potential treatment for severe asthma exacerbations refractory to standard therapy.

Evidence. Experimental use of ketamine in adult patients presenting with status asthmaticus is limited to one randomized controlled trial. Howton et al[32] evaluated the safety and efficacy of ketamine in nonintubated adults with an acute, severe asthma exacerbation. Fifty-three patients with a peak expiratory flow less than 40% of the predicted value following initial albuterol therapy were randomized to receive a ketamine 0.2 mg/kg IV bolus over 5 minutes, followed by an IV continuous infusion at a rate of 0.5 mg/kg/hr or placebo; both groups received standard therapy. Due to dysphoria that occurred in three of the initial six ketamine patients, subsequent bolus doses were decreased to 0.1 mg/kg IV and the first nine patients were eliminated from the analysis. There was overall improvement seen in peak flow, Borg score, RR, and forced expiratory volume in 1 second among both patient groups; however, there were no statistically significant differences detected between the two groups when comparing changes in these parameters over a 3-hour time period. Noting the possibility of β-error, the authors concluded that ketamine at a dose low enough to avoid dysphoria demonstrated no clinical benefit in treating refractory asthma exacerbations in an adult population when compared with standard therapy.

Risks. There were no statistically significant differences in terms of safety, although overall adverse effect rates were numerically higher with ketamine compared with placebo. Rates of individual adverse effects were not reported; however, dysphoria and dizziness were noted to be the most common. One patient was intubated in the ketamine group compared with three in the placebo group, but this was stated to not be a statistically significant finding.

Clinical Application. The literature on ketamine use for status asthmatics in adults is sparse, and this study failed to show a significant difference in efficacy compared with standard therapy alone (level 3 evidence). Well-conducted, randomized controlled trials with a patient-oriented primary outcome are still needed to identify whether there is any clinical benefit of using ketamine in this population. Thus, ketamine should not be routinely used for status asthmaticus.

Alcohol Withdrawal Syndrome

Rationale. Chronic alcoholics experience a down-regulation in GABA receptors and an up-regulation of NMDA receptors.[1,9] Symptom-triggered benzodiazepine therapy is the mainstay in management of alcohol withdrawal syndrome (AWS), although there are no guidelines for the assessment and management of AWS in the critically ill population.[33] Oftentimes, patients develop resistance to benzodiazepines, typically defined as needing greater than 40 mg of diazepam equivalents over the course of an hour. Major sequelae of uncontrolled AWS include seizure and delirium tremens (DTs), and therapies are needed to prevent these effects.[33] Ketamine exhibits a mechanism of action similar to ethanol by blocking NMDA, a site of action that is uncommonly used in current AWS management.

Evidence. Two recent studies reported the use of ketamine infusions as an adjunct to symptom-triggered therapy in the ICU population.[34,36] Wong et al[34] were unable to find a difference in benzodiazepine requirements before and after initiation of ketamine in 23 ICU patients with AWS (p = 0.11). Conversely, Shah et al[36] found decreased lorazepam requirements 24 hours after ketamine initiation (p < 0.05), and all 30 patients achieved symptom control within 1 hour of ketamine. Despite this reporting, six of 14 nonintubated patients eventually required mechanical ventilation after initiation of ketamine. This study used higher infusion rates compared with Wong et al,[34] possibly accounting for the difference in outcomes. The study populations differed between both analyses, reporting differences in rates of intubation prior to ketamine, and the reasons for initiation of ketamine (i.e., delirium tremens, benzodiazepine resistance).

Another recent study evaluated a protocol implemented for the management of nonintubated patients with DTs, where a ketamine infusion was immediately added to symptom-triggered therapy at the diagnosis of DTs.[35] The authors found significant reductions in mean ICU length of stay by 2.83 days (95% CI, −5.58 to –0.089 d; p = 0.043), likelihood for intubation (odds ratio, 0.14; 95% CI, 0.04–0.49), and total benzodiazepine requirements (p = 0.02) when compared with symptom-triggered therapy with benzodiazepines alone.[35]

Risks. Few adverse events are noted with ketamine in AWS. The need for intubation may be of concern, as it was exemplified by the more aggressive dosing strategy reported by Shah et al.[36] Emergence reactions were not reported.

Clinical Application. Although ketamine appears to be a safe option to consider as an adjunct to AWS management in the ICU, current evidence is fairly discrepant concerning appropriate timing, dosing strategies, and monitoring of ketamine in AWS (level 3 evidence).[35,36] There is a need for further well-designed studies to confirm ketamine's place in therapy in managing AWS.

Status Epilepticus

Rationale. Strong evidence exists for the administration of benzodiazepines for status epilepticus (SE), but evidence for other medications is based on very limited data. In many cases, patients develop refractory SE (RSE), or the failure to control seizures with benzodiazepines and a second antiepileptic drug (AED).[38] The down-regulation of GABA receptors and subsequent pharmacoresistance to benzodiazepines may contribute to disease progression from SE to RSE.[37] Reduced response to common first-line drugs due to altered GABAA receptor and augmentation of p-glycoprotein-mediated export of phenytoin and phenobarbital can be seen 30 minutes after seizure onset.[40] Coupled with increased excitotoxicity due to up-regulation of NMDA receptors at the synaptic membrane, ketamine presents itself as a possible treatment option.

Evidence. In a multicenter study, Gaspard et al[39] observed the use of ketamine for the management of RSE in adults and children. The authors assessed the impact of ketamine on seizure control, defining a "possible response" as permanent control of SE within 24 hours of starting of ketamine. In the study, 19 of 60 patients (32%) achieved a "possible response" to ketamine.[39] Response rates were highest when ketamine was started early (at a median of 12 hr) after seizure onset and when fewer concurrent antiepileptics were present at the time of ketamine initiation (4 vs 8 AEDs; p < 0.01). The median infusion rate was 2.75 mg/kg/hr and administration of a bolus dose did not correlate with permanent seizure control. Variables associated with lower mortality included younger age and achieving a "possible response" to ketamine (p = 0.001 and p ≤ 0.001, respectively). Other retrospective reviews describe success in controlling super-RSE (RSE persisting for 24 hr) with ketamine, although the quality of evidence remains low.[40–42]

Risks. A case of cardiac arrest with possible association to ketamine was reported, although more data are needed to confirm this adverse event.[43] Ketamine's effect on ICP while managing RSE has also not been fully assessed.

Clinical Application. It may be reasonable to consider ketamine in the management of RSE, especially in the setting of cardiac depression from other anesthetic agents (level 3 evidence). It is unclear at what point of therapy ketamine should be added and some have advocated to start ketamine early in the course of RSE due to the pharmacoresistance that can develop early in RSE.[40] Nevertheless, further studies are needed to truly elucidate the place in therapy and optimal dosing strategy of ketamine in RSE.

Acute Agitation

Rationale. Patients may present in a prehospital or ED setting with acute agitation or excited delirium, defined as altered sensorium and aggressive behavior in which patients experience hyperadrenergic autonomic dysfunction and metabolic acidosis that may result in death.[44–46] Ketamine may be an appropriate treatment option due to its rapid onset and duration, particularly when given intramuscularly.[45]

Evidence. A prospective, observational study compared ketamine to a benzodiazepine or haloperidol (or combination) in 98 patients and found that those who received ketamine were less agitated at 5, 10, and 15 minutes (p = 0.001, p ≤ 0.001, p = 0.032, respectively) based on an agitation score of less than or equal to 2 on a 6-point scale, indicating patients to be mildly aroused/pacing, settled, or asleep (scores 2, 1, and 0, respectively).[45] Other studies have described ketamine in the prehospital setting.[44,46–49]

Keseg et al[44] conducted a retrospective review of 35 patients who received ketamine for agitation and reported subjective improvement in 91% of patients without using a standardized assessment methodology. Cole et al[46] investigated intramuscular ketamine as first-line therapy in 49 patients with profound agitation defined as Altered Mental Status Scale (AMSS) of +4 and found that patients achieved adequate sedation in a median of 4.2 minutes (95% CI, 2.5–5.0 min; range, 1–25 min). Parsch et al[47] performed a retrospective review before and after implementation of ketamine practice guidelines for a transport team and found that intubation incidence decreased (36% vs 7.14%; p < 0.01). Another recent prospective, open-label trial compared ketamine 5 mg/kg intramuscular to haloperidol 10 mg intramuscular in patients with AMSS +2 or +3 and found that patients who received ketamine achieved adequate sedation within a median of 5 minutes versus 17 minutes for haloperidol (p < 0.0001).[48]

Risks. High intubation rates after ketamine administration have been reported, however, this may be related to individual provider practice as described in two studies.[46,49] Parsch et al[47] reported higher complication rates in the ketamine group (49% vs 5%; p < 0.0001), including intubation (39% vs 4%; p < 0.0001). A retrospective cohort study compared single ketamine doses related to intubation incidence and found that both "high dose" (> 5 mg/kg) and "low dose" (≤ 5 mg/kg) groups had similar rates.[49] Other complications included hypersalivation, emergence reaction, laryngospasm, and vomiting.[44–48]

Clinical Application. Although data are limited, studies suggest ketamine's efficacy for acute agitation in ED and prehospital settings. However, given the higher reported incidence of adverse effects, ketamine could be considered as an adjunct agent with close monitoring until further prospective data are available (level 3 evidence).