Michelson 2007.

Study characteristics
Methods	Double‐blind randomised controlled trial
Participants	Diagnosis: DSM‐IV major depressive disorder; HRSD‐17 score ≥ 18 N: 146 Age: atomoxetine group = 44.0 (SD = 12.3); placebo group = 45.5 (SD = 13.8) Sex: aAtomoxetine group 65% female; placebo group 66% female Baseline depression severity: atomoxetine group HRSD = 23.4 (SD = 3.5); placebo group MADRS = 23.1 (SD = 4.3)
Interventions	All patients first received 8 weeks of treatment with sertraline, initiated at 100 mg/day and increased in 50 mg increments to a maximum of 200 mg/day, based on efficacy and tolerability. Patients with a score > 4 on the Maier and Phillip core mood severity scale (MPS) of the HRSD were randomly assigned to 8 weeks of sertaline combined with atomoxetine or placebo Atomoxetine (N = 72) initiated at 40 mg/day, could be increased in 40 mg increments to a maximum of 120 mg/day, based on efficacy and tolerability Placebo (N = 74) The dose of sertraline during randomised treatment was fixed at 150 mg/day or 100 mg/day for patients who were unable to tolerate 150 mg/day during initial sertraline treatment Concomitant treatment: Unclear
Outcomes	Maier and Phillip core mood severity scale (MPS) of the HRSD HRSD CGI‐S Remission rate (MPS score ≤ 4 and no single item >1) Non response rate (< 30% reduction in MPS) Partial response rate (> 30% reduction in MPS but MPS > 4) Adverse events
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	No details given on random sequence generation, other than quote: ''were randomly assigned"
Allocation concealment (selection bias)	Unclear risk	No details given on random sequence generation, other than quote ''were randomly assigned under double‐blind conditions''
Blinding of participants and personnel (performance bias) All outcomes	Low risk	To minimise rating bias, investigators and patients were blind to the symptom severity threshold for randomisation. To preserve this blinding, patients who met the response criteria (MPS ≤ 4 and no single item > 1) after the initial 8 weeks continued sertraline monotherapy in the randomised phase but were not included in the analyses of results from the randomised phase of the trial
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	No details given on blinding of outcome assessment
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Responders data are per individual and only out of 26 and 28 rather than 72 and 74. Seems unclear
Selective reporting (reporting bias)	Unclear risk	Protocol unavailable
Other bias	Low risk	No other potential sources of bias identified