Sumner 2020.

Study characteristics
Methods	Randomised, double‐blind, active placebo‐ controlled cross‐over design.
Participants	Diagnosis: participants who met DSM‐IV criteria for major depressive disorder. N: 30 Age: 30.7 (8.85) Sex: 15 male (50%), 15 female (50%) Baseline depression severity: ketamine group MADRS M = 30.00 (SD = 5.21); Placebo MADRS M = 30.33 (SD = 4.65).
Interventions	Participants were randomised to receive one dose of racemic ketamine (0.25 mg/kg bolus, followed a 0.25 mg/kg/hour infusion for 45 minutes), or the active placebo remifentanil hydrochloride as a 9‐minute infusion using a target‐controlled infusion system to achieve 1.7 ng/mL plasma concentration using the Minto pharmacokinetic model.
Outcomes	MADRS Response
Notes	Authors kindly provided additional data for the first half of the cross‐over.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Order counterbalanced.
Allocation concealment (selection bias)	Unclear risk	Order randomised as stated, but it is unclear how or by what method.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Blinding measures were not effective & high risk of performance bias: quote: “most (88%) participants were able to correctly identify their ketamine session during a debrief. This is a known issue with psychoactive drug research and is not unique to current study."
Blinding of outcome assessment (detection bias) All outcomes	Low risk	To maintain study blinding, the MADRS assessor was never present during the acute phase of drug effects and patients were explicitly instructed not to share their experiences during the treatment sessions with the assessor.
Incomplete outcome data (attrition bias) All outcomes	Low risk	No data missing.
Selective reporting (reporting bias)	Unclear risk	No protocol available.
Other bias	Low risk	None identified.