Timmerman 1993.

Methods	Six‐week, double‐blind, randomised, parallel group, multicentre study.
Participants	Cooperative out‐patients of either sex with a reasonable knowledge of the Dutch language, who met the DSM‐III‐R criteria for "Major Depression, single episode", "Major Depression, recurrent", "Bipolar Disorder, depressed", with a score of a least 16 on the 17 items Hamilton Depression Rating Scale (HDRS). Age range: 18‐70 years. Exclusion criteria: patients who had been treated with MAO inhibitors or fluoxetine within the last 3 weeks or with other psychotropic drugs within the last week, with the exception of benzodiazepines. Patients with another primary psychiatric diagnosis than the above mentioned, or with a history of epilepsy, alcohol and/or drug abuse, pregnant or lactating women and women with childbearing potential failing to use standard birth control methods as well as patients with renal, hepatic, cardiovascular, neurological or somatic disorders, and/or significant abnormal laboratory findings.
Interventions	Citalopram: 108 participants. Fluvoxamine: 109 participants. Citalopram dose range: 20‐40 mg/day Fluvoxamine dose range: 100‐200 mg/day
Outcomes	Primary outcome: mean change on HDRS score from baseline to endpoint. Secondary outcomes: change in Clinical Global Impression‐Severity (CGI‐S) score, in the Zung Self‐rating Scale for depression score.
Notes	The study was funded by Lundbeck (citalopram manufacturer). One suicide completed in citalopram group, one fatal myocardial infarction in citalopram group, two suicide attempted in fluvoxamine group.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "patients were randomly assigned".
Allocation concealment (selection bias)	Unclear risk	No reliable information reported.
Blinding (performance bias and detection bias) All outcomes	Low risk	Quote: "patients were randomly assigned to double‐blind treatment".
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Quote: "the Intention‐to‐treat (ITT) population included all patients who had been allocated a randomization number to entry of double‐blind treatment".
Selective reporting (reporting bias)	Unclear risk	The study protocol is not available.
Other bias	Unclear risk	Sponsorship bias cannot be ruled out.