Hobart 2018b.

Study characteristics
Methods	Phase 3, multicentre, randomised, double‐blind, placebo‐controlled trial, fixed‐dose
Participants	Outpatients; men and women aged 18–65 years Key inclusion criteria: diagnosis of MDD and in a current MDE, as defined by DSM‐IV‐TR criteria and confirmed by both the MINI and an adequate clinical psychiatric evaluation; the current MDE ≥ 8 weeks in duration; history for the current MDE of an inadequate response to ≥ 1 and < 3 adequate antidepressant treatments (an inadequate response defined as < 50% reduction in depressive symptom severity, as assessed by the ATRQ); HAM‐D17 Total score ≥ 18 at screening and baseline visits. Key exclusion criteria: inadequate response (< 50% reduction in depressive symptom severity) to > 3 monotherapy antidepressant treatments during the current MDE; received ECT for current MDE; inadequate response to ECT; current Axis I (DSM‐IV‐TR) diagnosis; current Axis II (DSM‐IV‐TR) diagnosis of borderline, antisocial, paranoid, schizoid, schizotypal or histrionic personality disorder.
Interventions	Brexpiprazole 2 mg + antidepressant, 192 participants Placebo + antidepressant, 202 participants Antidepressants: escitalopram tablets 10 mg/day or 20 mg/day: 35 (18.2%); fluoxetine capsules 20 mg/day or 40 mg/day: 38 (19.8%); paroxetine CR 25 mg/day, 37.5 mg/day or 50 mg/day: 29 (15.1%); sertraline tablets 50 mg/day, 100 mg/day, 150 mg/day or 200 mg/day: 24 (12.5%); duloxetine delayed‐release capsules 30 mg/day, 40 mg/day or 60 mg/day: 27 (14.1%); venlafaxine XR capsules 37.5 mg/day, 75 mg/day, 150 mg/day or 225 mg/day: 39 (20.3%)
Outcomes	Primary outcome: change in MADRS at week 6 Key secondary outcomes: change from end of Phase A to end of Phase B in MADRS Total score for the subpopulations with anxious distress; change from baseline on CGI‐S, HAM‐D17 scores, MADRS – Response, CGI‐I – Response; remission as MADRS ≥ 50% reduction from baseline and MARDS score ≤ 10; acceptability; tolerability; safety; change in SDS score.
Notes	Brexpiprazole + antidepressant: number of prior antidepressant failures: 1 antidepressant: 157 (81.8%) participants; 2 antidepressants: 31 (16.1%) participants; 3 antidepressants: 4 (2.1%) participants. Placebo + antidepressant: number of prior antidepressant failures: 1 antidepressant: 162 (80.2%) participants; 2 antidepressants: 35 (17.3%) participants; 3 antidepressants: 5 (2.5%): participants.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Treatments were assigned by the IWRS based fixed‐block, computer‐generated randomisation code provided by the study sponsor, stratified by study center."
Allocation concealment (selection bias)	Low risk	Quote: "Treatment assignments were blinded to patients, investigators, and sponsor personnel, including those involved in the analysis."
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote: "Brexpiprazole and matching placebo tablets were provided by the sponsor, packaged in numbered, weekly blister cards." Quote: "Masking: quadruple (participant, care provider, investigator, outcomes assessor)."
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: "Brexpiprazole and matching placebo tablets were provided by the sponsor, packaged in numbered, weekly blister cards." Quote: "Masking: quadruple (participant, care provider, investigator, outcomes assessor)."
Incomplete outcome data (attrition bias) All outcomes	Low risk	Low reported attrition rates according to protocol criteria (21/394 (5.3%) participants). Slightly unbalanced between groups (6/202 participants in placebo arm and 15/192 in brexpiprazole arm). Attrition bias was unlikely.
Selective reporting (reporting bias)	Low risk	Outcomes reported as listed in protocol.
Other bias	Unclear risk	No information on other bias available.