Shiroma 2020.

Study characteristics
Methods	Randomised, double‐blind, active placebo‐controlled trial
Participants	Diagnosis: DSM‐IV major depressive disorder N: 54 Age: ketamine group M = 54.4 (SD = 13.8); midazolam group M = 51.2 (SD = 12.5) Sex: ketamine group 12% female; midazolam group 17.2% female Baseline depression severity: Ketamine group MADRS M = 34.88 (SD=7.80); Midazolam group MADRS M = 33.82 (SD=5.02)
Interventions	Participants were randomised to receive 6 ketamine (0.5mmg/kg) infusions or 5 midazolam (0.045mmg/kg) plus 1 ketamine (0.5mmg/kg) infusion. They received six infusions over a 12‐day period. Concomitant medications: participants were allowed to continue concomitant psychiatric medication regimen on stable dosages for at least 6 weeks prior to study onset.
Outcomes	MADRS mean change Response (≥50% MADRS reduction from baseline) Adverse events
Notes	Authors provided additional data for 5 ketamine infusions vs. 5 midazolam infusions.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: “Randomization was conducted using permutated blocks of 4 and 1:1 assignment between treatments”.
Allocation concealment (selection bias)	Low risk	Quote: "Group assignments for each participant was concealed in sequentially numbered, sealed, opaque envelopes.”
Blinding of participants and personnel (performance bias) All outcomes	High risk	Blinding tested with limited effectiveness for both personnel and participants. Quote: "At the end of the last infusion (midazolam plus single ketamine versus six ketamine),10.7% among midazolam cases and 50.0% among ketamine cases guessed incorrectly about assigned treatment (X2 1df = 9.12; P = 0.002). Raters, who were different during infusions days from those rating antidepressant outcomes at 24 h., incorrectly guessed 20.8% of midazolam cases and 7.1% of repeated ketamine cases prior to the last infusion (X2 1df = 1.25; P = 0.26). After the last infusion, 6.9 and 26% raters were incorrect about treatment assignment among midazolam plus single ketamine and repeated ketamine cases, respectively (X2 1df = 3.62; P = 0.06)."
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Blinding tested with limited effectiveness for both personnel and participants.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Participant recruitment and withdrawals documented. No participants withdrew after baseline assessment.
Selective reporting (reporting bias)	Unclear risk	Protocol unavailable.
Other bias	Low risk	None identified.