Anderson 2017.

Study characteristics
Methods	Multicentre, double‐blind, randomised, parallel‐group, superiority trial
Participants	Diagnosis: DSM‐IV moderate‐to‐severe unipolar or bipolar depressive episode N: 79 randomised, 70 included in modified intention‐to‐treat analysis Age: ketamine 0.5 mg/kg group M = 52.5 (SD = 11.9); placebo saline group M = 56.4 (SD = 12.4) Sex: ketamine 0.5 mg/kg group 67% female; placebo saline group 60% female Baseline depression severity: ketamine 0.5 mg/kg group MADRS score M = 31.8 (SD = 7.4); placebo saline group MADRS score M = 35.2 (SD = 8.4)
Interventions	Ketamine (0.5 mg/kg) or saline was given as a slow intravenous bolus directly before anaesthetic induction at each treatment. ECT treatments were administered twice weekly. Ketamine 0.5mg/kg group (N = 33) Placebo saline group (N = 37) Concomitant treatment: yes, oral psychotropic medication was continued by the treating clinical team during the trial
Outcomes	MADRS EQ‐5D‐3L Response rates (≥50% decrease in MADRS score from baseline) Remission rates (MADRS score ≤ 10)
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Patients were randomly assigned (1:1) to receive either ketamine or saline as anaesthetic for their ECT treatment. Randomisation was done by the Christie Hospital Clinical Trials Co‐ordination Unit (CTU) by use of permuted block randomisation, which varied randomly between four and eight, and was stratified by the NHS Trust."
Allocation concealment (selection bias)	Low risk	"Both patients and assessment and ECT treatment teams were masked to treatment allocation, although the anaesthetists administering the study medication were not. The anaesthetist broke the seal away from the psychiatric ECT team, drew up the trial medication into a syringe, and disposed of the ampoule without revealing which drug was being administered. Researchers responsible for outcome assessment did not attend ECT sessions. To assess success of masking, patients and assessors were invited to complete a questionnaire to indicate suspected treatment group after four ECTs and at the end of the ECT course, and the results were sent directly to the CTU in a sealed envelope for data entry.”
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Blinding was tested and deemed effective. Quote: "Masking of treatment allocation was successful as assessed by questionnaire; patient guesses were correct in 28 (48%) of 58 patients who completed the questionnaire at mid‐ECT and 30 (56%) of 54 patient guesses at the end of treatment, while 35 (56%) of 63 assessor guesses at mid‐ECT and 28 (51%) of 55 assessor guesses at the end of treatment were correct, which did not differ significantly from chance."
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Assessor blinding was assessed to be effective. Quote: "Masking of treatment allocation was successful as assessed by questionnaire... 35 (56%) of 63 assessor guesses at mid‐ECT and 28 (51%) of 55 assessor guesses at the end of treatment were correct, which did not differ significantly from chance."
Incomplete outcome data (attrition bias) All outcomes	Low risk	Exclusion, withdrawal and drop out rates at each time point are recorded
Selective reporting (reporting bias)	Low risk	Protocol available and data for all time points reported.
Other bias	Low risk	No other potential sources of bias identified