The inclusion criteria clarifies only 'True Abstinence' is in regards to contraception in this trial is acceptable and to makes clear that periodic abstinence is not true abstinence

The addition of Columbia Rating Severity Scale at screening visit as an additional assessment tool to exclude potentially suicidal patients

Clarifications to the incl./excl. criteria of the protocol, specifically to correct the breathalyser reading to 0.00 to ensure participants cannot have consumed alcohol prior to participation.

Explanation to the randomisation procedure, specifically pre-randomisation participants will be given a screening ID, whereas post-randomisation this will be updated and participants will be given a patient ID.

Clarification that two blood samples will be collected for BDNF and Ketamine on infusion days.

Explanation of the urine drug testing regarding which testing will be performed: to correct an error in the previous protocol as alcohol cannot be screened for by the urine test. The alcohol screen is from the breathalyser reading thus there are no additional tests made by this change. Not all urine tests will be carried out using dip tests, some will be carried out using a urine test cup, therefore this has been made clear in this amendment.

The purpose of this amendment is to provide further clarification on the eligibility criteria to identify that participants only need to meet criteria within one DSM version not both. Additionally we are specifying that these criteria need to be present within the past 12 months.

The purpose of this amendment is to clarify that a positive THC urine drug result is not grounds for exclusion. Occasional cannabis use is acceptable though cannabis dependence is grounds for exclusion. Cannabis dependence will be assessed using standardised dependence-related questions.

The purpose of this amendment is to clarify that the SCID will be used to assess for psychosis as the SPC for ketamine cautions against use in individuals with a history or current diagnosis of psychosis. All other psychiatric diagnoses will be recorded based on self-report information or identification by their medical professional.

We are increasing the upper age limit for the trial from 60 to 65 years to be in line with the working age. There is no extra implication of this clinically following discussion with study team anaesthetists and psychiatrists.

We are making it easier for volunteers to meet criteria for an alcohol use disorder in that participants will only be required to have a minimum of a moderate alcohol use disorder as opposed to the original criteria which asked for a severe disorder.

Benzodiazepines are used in community and inpatient alcohol detox and for sleep problems. They have a relatively long detection window in urine (up to 7 days for long-acting) so we are removing a positive urine benzodiazepine screen from the exclusion criteria at screening as we may erroneously exclude people who have just completed a detox. We will still exclude people with a diagnosis of benzodiazepine dependence.

Both the Beck Depression Inventory-II and the Hamilton Depression rating Scale are reliable assessments of depression therefore we are including the HAMD here for completeness.

We want to expand recruitment to include participants that have minimal depression on the BDI and HAM-D. We are already excluding participants who are using anti-depressants, to exclude people who are not depressed is reducing the pool of participants considerably and feel this change would aid recruitment without altering the safety of participants (as we are expanding to include those who are psychologically healthier). This would also not alter the scientific value of the trial as synaptogenesis has been observed in healthy mice following ketamine (Zunszain et al., 2013), thus we think ketamine would still work via its proposed treatment mechanism.

We are clarifying this exclusion criteria to specify any relevant mental or physical health issues identified by a medically qualified personnel can be grounds for exclusion, specifically including a history of psychosis in the participant, however there is only a risk from a first degree relative if a diagnosis of schizophrenia has been given. There is a large co-morbidity between anxiety and both depression and alcohol use disorder and no clinical implications of including anxiety disorders (phobias, GAD, PTSD), as ketamine is also being trialled as a treatment of these in other settings (https://clinicaltrials.gov/ct2/show/NCT02083926).

The original criteria is for participants to be willing and able to wear the SCRAM-X data collection bracelet for the duration of the trial, 6 months. However, we have found resistance to this criteria, the idea of wearing the device for 6 months as well as the stigma associated with such a bracelet. Other unforeseen events have also arisen, such as patients having to remove the bracelet to fly and issues with data download if neither a landline or mobile telephone reception is present in patients homes meaning researchers would have to attend patients homes to download the data. Criteria are now revised so that patients will wear the bracelet from screening to the end of active treatment (roughly 2 months). The TLFB, an index of self-reported alcohol consumption, as well as drink diaries, will be used to assess our primary outcome, the gold standard used in other clinical trials of alcoholism.

Anxiolytics, sedatives and hypnotics (e.g. Librium) are often prescribed to help with alcohol detoxification and also sleep problems that are associated with alcohol dependence. Use of these drugs does not pose a risk alongside our ketamine administration, only when concurrently administered, therefore this criterion from the SPC is inappropriate for the use of ketamine in the manner in which we are using it during this trial. Halogenated anaesthetics are not taken as daily prescribed medications so this is an inappropriate criterion as pointed out by the REC committee during the original meeting. As part of screening procedures we check with the trial medical professionals (anaesthetist and psychiatrist) about all concomitant medications. Therefore it is more thorough and representative to include this criterion

The original criteria indicated that any liver function test (listed in the protocol) result outside of 3 times the normal range would be grounds for exclusion from the trial. The reason for this being that liver damage is contraindicated in administration of ketamine. This criteria has been updated to specify three specific tests which should be used as indictors of liver damage and therefore grounds for exclusion (bilirubin, ALT and AST). Other listed LFTs provide markers of recent alcohol consumption but not liver impairment and so updating this criteria will not impact on patient safety and is more in line with criteria used in other trials (http://archpsyc.jamanetwork.com/article.aspx?articleid=2548275)

It has become apparent after screening a number of patients that ‘detoxifications’ is ambiguous therefore we have added ‘inpatient’ to disambiguate this phrase. The criterion was included to exclude the very severe, higher risk of relapse and evidence suggest that these are those that have undertaken a very high number of inpatient detoxification programmes.

We are expanding our recruitment to enlist the help of GPs to send invitations to participate to eligible patients across CCGs in the South West and London. The amendment to add these PICs has been approved (amendment number: 15/SW/0312/AM03 – PIC sites) Additionally, we have designed a recruitment flyer which can be given to staff working in PICs to help them when trying to identify potential participants and to aid discussion which such participants.

The screening visit should originally have been within 14 days of the baseline visit. This has been updated so that visits could be 28 days apart. We will still endeavour to have these two visits within 14 days however this provides flexibility if participants have other commitments e.g. work, which get in the way of this specific timeline.

A refresher screening session has been added to the protocol for those instances where more than 28 days elapses between screening and baseline visit. Such a situation might arise if participants relapse in between visits, yet is a participant becomes abstinent again and wishes to continue with the trial then we would like to provide the option for such a situation to occur. We are therefore defining the procedure for what would happen if such an event was to arise. All measures which could have changed in the elapsed timeframe and proposed to be readminstered to the participant.

Further clarification to timepoints of data collection for the vital sign assessments on infusion days: vital signs will be continuously monitored through infusion but the eCRF will detail specific sets of readings at 3 time points, prior to infusion, after infusion and after recovery.

Updated to provide clarity to several criteria which should be judged by medically qualified personnel. • How to determine previous or current dependence is outlined based on trial personnel discussions, i.e. including this within the GP letter and self-reported seeking of help for a drug problem. • Only clinically relevant neuropsychological difficulties should influence eligibility for the trial. • How to clarify what classes as a seizure has been added based on trial team discussions, i.e. they should be medically witnessed by an appropriate clinician and there should be documented evidence from an EEG or a history consistent with a diagnosis of epileptiform illness. • Current suicidal ideation should be considered relevant to eligibility which would require discussion between the clinical team.

Clarification that any medications which are deemed to pose a risk when combined with ketamine can be grounds for exclusion from the trial, not just those medications listed.

Clarification that a current diagnosis of the physical health conditions outlined would be grounds for exclusion if the risk benefit ratio was not in favour of giving ketamine by medically qualified personnel. Criteria updated: • Patients receiving thyroid replacement has been removed as this is a duplication of the criteria outlined in the medication section listing thyroid hormone treatment as an exclusion criteria. • Diabetes has been removed from this list as this was a requirement when participants were wearing the SCRAM-X bracelet for 6 months, however they now wear this only during active treatment (visits 1-8). The SCRAM bracelet is checked weekly at these visits and so this criterion is no longer relevant based on the current protocol. • Neurological condition/brain damage has been removed from this list as this is duplication as this is covered by points e and h in this criterion. • Clarification of what exactly presence of a head injury means

Participants in the KARE trial are judged as ‘street ready’ by an anaesthetist after their infusion. However, it is not always possible to arrange for a responsible adult to collect a participant after a visit therefore wording has been added to indicate that if a responsible adult is not able to pick up the participant after the visit then transportation will be arranged for that participant.

Updated definition of expected adverse event to reflect that this isn’t a serious event as the classification of seriousness is used throughout this protocol to indicate how an adverse event should be reported. Expected AEs are reported in source whereas Unexpected AEs are reported in the eCRF as well as source.

The protocol currently states that visits 3, 5 and 7 should always be one day after visits 2, 4 and 6 respectively, and that failure to attend one of these visits leads to exclusion from the trial. Every effort should be made to maintain this rule, however if a participant has a valid reason for being unable to attend one of these visits then some flexibility can be given, up to 5 days, to avoid exclusion. However, if a participant fails to attend this re-arranged appointment then they will be excluded from the trial.

Update to reflect that the bracelet can be removed if deemed necessary by the study team.

SCRAM-X bracelets cannot always be attached to participants at screening visit due to uncertainty around participant eligibility and difficulty in getting the bracelets back. Where possible a bracelet will be attached at screening, however anytime within the window of screening to baseline visit is acceptable.

We will not exclude participants from the trial if they start taking anti-depressants during follow-up therefore we are clarifying that only during active treatment i.e. visit 2-8, would this be an issue for eligibility. This is being updated as ethically we don’t feel that we can ask participants not to take anti-depressants during the follow up phase of this trial if they are deemed by a medical professional to be needed for the participant’s mental health.

Update to the protocol to allow re-consenting of patients by trial psychologists (including research assistants and postdoctoral research associates.

Update for possibility to repeat some of the screening period assessments if initially they did not meet criteria for eligibility. Some of the eligibility criteria in this trial relate to physical variables which can change with time. If participants prior to randomisation are judged to be ineligible based on one such characteristic e.g. elevated blood pressure, elevated LFTs, BMI outside of protocol limits, or a positive urine drug test due to recreational use, they will be invited back to re-do one or more such tests to truly determine whether they should be excluded based on this characteristic. For example, blood pressure can be elevated by situational anxiety or recent alcohol withdrawal, both of which are not indicative of uncontrolled hypertension. If a participant came back and blood pressure was found to be within trial limits prior to randomisation then we would include this participant. Additionally, due to the nature of alcoholism LFTs can be elevated outside of protocol limits. However the liver regenerates when alcohol use is stopped. We will repeat LFTs and if significant improvement has been seen putting the participant within protocol limits then they would be included. If a participant is too overweight or underweight to participate then those who lose or gain weight respectively so that their BMI is within trial limits will be included in the trial. Participants who only recreationally take other substances, without any evidence of current or prior dependence, who test positive for a substance will be informed this is an exclusion factor and given the chance to repeat such tests at a later date. If found to be negative for such substances in the future then participants would be included in the study.

Patient information sheet updated to reflect GDPR compatibility.

Updates to the protocol to clarify the purpose of the breathalyser test in the study and the follow-up visit schedule for participants that are withdrawn/discontinued from treatment.