Chen 2017.

Study characteristics
Methods	Randomised, placebo‐controlled, double‐blind trial
Participants	Diagnosis: DSM‐IV‐TR major depressive episode N: 132 randomised, 127 completed the study Age: ketamine 0.3 mg/kg group M = 40.94 (SD = 15.41); placebo saline group M = 37.44 (SD = 14.16) Sex: ketamine 0.3 mg/kg group 66.7% female; placebo saline group 64.1% female Baseline depression severity: ketamine 0.3mg/kg group HAM‐D score M = 38.05 (SD = 3.23); placebo saline group HAM‐D score M = 37.43 (SD = 2.64)
Interventions	Patients assigned to the study group were anesthetised using 1.5 mg/kg propofol and 0.3 mg/kg ketamine delivered intravenously. Patients assigned to the control group were anesthestised using 1.5 mg/kg propofol and normal saline. The ECT was administered 3 times a week (12 sessions of ECT in total). Ketamine 0.3 mg/kg group (N = 63) Placebo saline group (N = 64) Concomitant treatment: Not reported
Outcomes	Wechsler Memory Scale‐Chinese Revision (WMS‐RC) MMSE HAM‐D Remission rates (sustained HAM‐D score <10 after 2 consecutive ECTs) 4‐item positie symptom subscale of the Brief Psychatric Rating Scale (BPRS+)
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "All recruited participants were randomly assigned into 1 of 2 groups: the control group and the study group. Half the patients were placed in each group using a computer‐generated random number table." No further details given.
Allocation concealment (selection bias)	Low risk	Quote: "The patients, treatment teams, and the outcome assessors were all blinded to the intervention allocation."
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Blinding stated but not tested. Quote: "The anesthesiologists, patients, and outcome raters were all blind to the group assignments and the anesthetic regimen."
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Blinding stated but not tested. Ketamine and normal saline were both prepared before each ECT treatment, quote: "by a specialized nurse who was blind to the study design".
Incomplete outcome data (attrition bias) All outcomes	Low risk	Exclusion and withdrawal rates at each time point are recorded
Selective reporting (reporting bias)	Unclear risk	Protocol unavailable
Other bias	Low risk	No other potential sources of bias identified