Table 1.

Characteristics of the included studies.

References	Substance of abuse and sample characteristics	Ketamine/control intervention	Main outcome measures	Results	Study limitations
COMPLETED STUDIES
(19)	Cocaine 8 non-treatment and non-abstinence seeking participants aged 21–52, meeting DSM-IV criteria for cocaine dependence.	Three arm, cross-over design RCT in which participants were hospitalized for 9 days, undergoing infusions of 0.41 mg/kg ketamine, 0.71 mg/kg ketamine, and 2 mg lorazepam at intervals of 48 h, with assessments performed 24 h post-infusion.	Motivation levels to stop cocaine use (assessed using the URICA) and levels of cue-induced craving (using the VAS).	Motivation to quit trended toward significance (p = 0.05) with a median improvement in URICA scores from 0.71 to 4.87. There was a significant reduction in cocaine use compared to baseline as quantified by median amount ($149.30/using day at baseline vs. $10.50 post-treatment, p < 0.001) and frequency (22/28 using day pre-treatment, compared to 5/28 post-treatment, p = 0.01).	Small sample size with narrow demographic profile (55–87.5% male, 75–87.5% African American) limits extensive external comparisons. Cross-over design limits ability to analyze duration of effects.
(20)	Cocaine 20 non-depressed, non-treatment seeking participants aged 21–55, meeting DSM-IV criteria for cocaine dependence as well as a minimum criteria for use (at least 2 days of cocaine use per week, spending more than $40 per use).	Two arm, cross-over design RCT in which participants were hospitalized on 3 occasions at 2 week intervals and underwent choice sessions during which they were offered either 25 mg cocaine or $11. Baseline rates were established using a saline infusion during hospitalization #1. During hospitalizations #2 and #3, participants underwent either ketamine infusion (0.71 mg/kg) or midazolam (0.025 mg/kg) 28 h prior to the choice session.	Number of cocaine self-administration choices at 28 h post-infusion relative to baseline. Secondary outcome measures included cocaine use rates following discharge #2 and stress reactivity scores (measured using the FFMQ).	Self-administration rates were reduced significantly in the ketamine group (1.61 cocaine choices vs. 4.33 in the midazolam group, p < 0.0001). Non-reactivity rates were significantly improved in the ketamine group at 3.46 (vs. 2.92 in the midazolam group, p < 0.05). Ketamine use was significantly reduced in the acute time period (1–6 days) post discharge ($22.45 vs. $3.20, p < 0.05), but lost significance over longer follow-up. Additionally, 2/10 participants were voluntarily abstinent in the 2 weeks post-discharge in the ketamine group, compared to 0/10 in the midazolam group. Craving rates were significantly reduced at time of discharge by 59.6% in the ketamine group compared to 15.3% in the midazolam group, although this lost significance at subsequent follow-up time points.	Small sample size with narrow demographic profile limits extensive external comparisons. Cross-over design limits ability to analyze duration of effects.
(21)	Opioids 70 participants aged 18–30 meeting DSM-IV criteria for heroin dependence who had recently undergone a 2-week inpatient detoxification.	Two arm RCT in which participants were randomized to a single session of high dose (2 mg/kg) or low dose (0.2 mg/kg) IM ketamine in conjunction with existentially-oriented psychotherapy.	Self-reported abstinence rates following discharge assessed on a monthly basis for 2 years.	One and two year complete heroin abstinence rates of 24% and 17% in the high dose group vs. 6 and 2% in the low dose group, p < 0.05).	Lack of clarity regarding the event of possible discordance in abstinence outcomes suggests possible reporting bias. No placebo group (only low dose comparison group). No follow-up on opioid substitution pharmacotherapy.
(22)	Opioids 59 participants aged 18–35 meeting DSM-IV criteria for heroin dependence who had recently undergone a 2-week inpatient detoxification	Two arm RCT in which all participants underwent existentially-oriented psychotherapy in conjunction with 2 mg/kg IM ketamine. Participants were then randomized to either 2 general addiction counseling sessions on a monthly interval or 2 additional ketamine-guided psychotherapy sessions at a dose of 2 mg/kg IM.	Self-reported abstinence rates following discharge assessed on a monthly basis for 2 years.	At 1 year follow-up, 13/26 (50%) of participants in the multiple ketamine infusion group remained fully abstinent from heroin vs. 6/27 (22.2%) subjects in the single session group.	No placebo group (only low dose comparison group). No follow-up on opioid substitution pharmacotherapy.
(23)	Opioids 58 participants aged 18–35 meeting DSM-IV criteria for opioid dependence.	Two arm RCT in which all participants underwent general anesthesia for rapid induction of opioid withdrawal. Ketamine (0.5 mg/kg/h) vs. saline placebo were compared as part of maintenance of anesthesia prior to opiate antagonist (naltrexone) induction while under general anesthesia.	Cardiovascular response (blood pressure, pulse), respirations, renal output and gastrointestinal output as well as cortisol levels were measured. Subjective and Objective Opiate Withdrawal Scales (SOWS and OOWS) were assessed during the early post-anesthetic period. Abstinence rates at 4 months were also secondarily assessed.	The ketamine group showed significant lower increases in blood pressure, pulse and cortisol levels during rapid opiate antagonist induction under general anesthesia. The ketamine group had lower OOWS at hours 1 and 2 of during the anesthesia phase. In the early post-anesthesia phase, patients in the ketamine group required less adjuvant carbamazepine and clonazepam to maintain equivalent opioid withdrawal symptoms compared to the placebo group (p < 0.001). There were no significant differences in terms of abstinence rates at 4-month follow-up.	Subsequent opioid antagonist use in the follow-up phase was not assessed. Abstinence rates were not tracked in interim follow-up, limiting the discernment of duration of effects.
(24)	Alcohol 111 treatment-seeking alcohol dependent participants and 100 treatment seeking controls, aged 23–56.	Two arm, case-control study in which participants initially completed a 3 month inpatient detoxification program. The ketamine group then underwent a single session of 2.5 mg/kg IM ketamine in conjunction with existentially oriented psychotherapy. The control group received routine follow-up.	Self-reported abstinence rates following discharge assessed on a monthly basis for 2 years.	Total abstinence rates at 1 year follow-up of 73/111 (65.8%) in the ketamine treatment group as compared to 24/100 (24%) in the treatment as usual group. Two year follow-up rates of total abstinence were 40.7% in the ketamine group, although comparisons could not be made to the control group, which was only followed for a 1 year interval.	Study methodology suggests potential risk of selection bias. The addition of a behavioral intervention (existential psychotherapy) for the treatment but not comparison group limits isolation of ketamine's effects.
(25)	Alcohol 23 patients aged 47–54 admitted to a medical ICU for management of alcohol withdrawal syndrome.	Retrospective review of 23 patients who were admitted for management of severe alcohol withdrawal symptoms and administered ketamine as an adjunct to conventional treatment with benzodiazepines.	Median change in benzodiazepine requirements to maintain Withdrawal Assessment Scale values < 10, frequency of delirium tremens and other alcohol withdrawal symptoms.	While 75% of patients developed symptoms of delirium tremens early in their hospitalization, no AWS complications (delirium tremens, hallucinations, or seizures) occurred after ketamine was initiated. There was a trend toward decreased benzodiazepine requirements at 12 and 24 h after initiation of ketamine infusion with median changes of −40 mg and −13.3 of diazepam equivalents (p = 0.11 and p = 0.33, respectively).	Retrospective review in which ketamine infusions were not standardized in dose or frequency. Adjunctive anesthesia agents were not standardized between patients, and patients were neither randomized nor matched on withdrawal severity.
ONGOING STUDIES
Ketamine for Reduction of Alcoholic Relapse/Celia Morgan, PhD	Alcohol	Estimated enrollment of 96 recently detoxified participants, ages 18-60, meeting DSM-V criteria for alcohol use disorder and minimum of mild depression (BDI-II score >14).	RCT with 4 groups, comparing ketamine with saline placebo and manualized relapse prevention psychotherapy with simple education about alcohol effects. Both groups will also receive 12 weeks of motivation enhancement therapy and 4 weeks of mindfulness based relapse prevention.	Arm 1: 0.8 mg/kg IV ketamine + manualized relapse prevention psychotherapy (RPT). Arm 2: a 0.8 mg/kg IV ketamine session + simple alcohol education (AE). Arm 3: saline placebo + RPT. Arm 4: saline placebo + AE.	Primary outcome measures are abstinence rates at 3 and 6 months, as well as relapse rates to alcohol at 6 months. Secondary outcome measures include depressive symptoms, craving, and quality of life.
Glutamatergic Modulation of Disordered Alcohol Use /Elias Dakwar, MD	Alcohol	Estimated enrollment of 40 treatment seeking participants, ages 21-69, meeting DSM-IV criteria for active alcohol dependence.	RCT with 2 groups, both groups also receive 5 weeks of motivational enhancement therapy.	Arm 1: 0.71 mg/kg IV ketamine in wk2 Arm 2:0.025 mg/kg ketamine IV in wk2.	Primary outcome measure will be change from baseline to wk 5 in # drinking days (TLFB).
Ketamine for the Rapid Treatment of Major Depressive Disorder and Alcohol Use Disorder/Gihyun Yoon, MD	Alcohol	Estimated enrollment of 65 participants, ages 21–65 meeting DSM-V criteria for MDD and AUD.	RCT with 3 groups. All groups will receive weekly infusions of ketamine or saline placebo. Participants will also receive 2 intramuscular (IM) injections of either naltrexone or saline placebo.	Arm 1: 0.5 mg/kg IV ketamine weekly × 4 wks + monthly IM naltrexone (380 mg). Arm 2: 0.5 mg/kg IV ketamine weekly × 4 wks + monthly IM saline placebo. Arm 3: psychoactive placebo 0.045 mg/kg midazolam) + Placebo weekly × 4 weeks.	Primary outcome measures will be: depression severity from baseline to day 21 (MADRS) and rate of complete abstinence from alcohol from baseline to day 21 (TLFB).
Glutamatergic Modulation to Facilitate the Behavioral Treatment of Cocaine Use Disorders/Elias Dakwar, MD	Cocaine	Estimated enrollment of 150 treatment-seeking participants, ages 21–60, meeting DSM-V criteria for cocaine use disorder.	RCT with 2 groups, both groups also receive 12 weeks of MET and 4 weeks of MBRP.	Arm 1: 0.71 mg/kg IV ketamine in wk1+ wk5. Arm 2: 0.025 mg/kg ketamine IV in wk1+ wk5.	Primary outcome measure is abstinence from cocaine use from baseline to wk12 (TLFB).
Facilitating Rapid Naltrexone Initiation/Elias Dakwar, MD	Opioids	Estimated enrollment of 100 treatment-seeking participants, ages 18–60, undergoing acute inpatient detoxification over the course of 5 days.	RCT with 2 groups, both groups also receive 12 weeks of mindfulness based relapse prevention and motivational interviewing. Participants will undergo an initial ketamine infusion when they begin experiencing moderate withdrawal and be up-titrated on oral naltrexone as tolerated during hospitalization.	Arm 1: Two infusions of 0.11 mg/kg IV ketamine bolus + 1.3 mg/kg IV ketamine over 90 min, over the course of 2 consecutive days. Arm 2: Two infusions of saline bolus + 0.0125 mg/kg IV ketamine over 90 min over the course of 2 consecutive days.	Primary outcome will be rates of initiation on extended-release naltrexone.
Facilitating the Behavioral Treatment of Cannabis Use Disorder/Elias Dakwar, MD	Cannabis	Estimated enrollment of 15 treatment seeking participants, ages 21–60, meeting DSM-IV criteria for cannabis dependence.	Single-blind trial, with participants given the understanding that they may receive any of several medications in conjunction with a 2 week course of MET and then a 4 week course of MBRP.	Arm 1: Initial infusion of 0.71 mg/kg IV ketamine in week 2 + possible second infusion in week 3 or 4.	Primary outcome will be rates of cannabis use from baseline to week 6 (UDS).

RCT, Randomized controlled trial; URICA, University of Rhode Island Change Assessment; VAS, Visual Analog Scale; FFMQ, Five Facet Mindfulness Questionnaire; BDI-2, Beck's Depression Inventory-II; TLFB, Timeline Followback; MADRS, Montgomery–Asberg Depression Rating Scale; RPT, Relapse prevention psychotherapy; AE, Alcohol education; UDS, Urine drug screen; MET, Motivational enhancement therapy; MBRP, Mindfulness based relapse prevention.