Allard 2004.

Methods	Twenty‐two‐week, double‐blind, randomised, parallel group study.
Participants	Outpatients meeting DSM‐IV criteria for major depression, having a minimum score of 20 on Montgomery and Asberg Depression Rating Scale (MADRS) and a ≤ 20% change in MADRS score between pre‐study and baseline visits, which were one‐week apart. Age‐range: 64‐89 years
Interventions	Venlafaxine: 76 participants. Citalopram: 75 participants Venlafaxine dose range: 75‐150 mg/day Citalopram dose range: 20‐40 mg/day Zopiclone (≤ 7.5 mg/day) or zolpidem (≤ 5 mg/day) for insomnia and medications for treatment of somatic disorders were allowed.
Outcomes	Primary outcome: change in MADRS score from baseline to week 8. Secondary outcomes: Clinical Global Impression (CGI), subscale Severity of Illness and Global Improvement. Geriatric Depression Scale (GDS‐20).
Notes	This study was funded by Wyeth (venlafaxine manufacturer). One death in the citalopram group (unknown cause of death).
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "the study was designed as a randomized". Probably done.
Allocation concealment (selection bias)	Unclear risk	No information provided.
Blinding (performance bias and detection bias) All outcomes	Low risk	Quote: "both venlafaxine and citalopram were administered orally in identically appearing capsules to maintain the double‐blind integrity of the study".
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Even though an Intention‐to‐treat (ITT) approach was used, no reliable information was provided in the paper to check the consistency between methods and results (for instance, see figures in Table 1 of the published paper). Quote: "Analyses of the efficacy variables were performed on an ITT patient population, defined as all randomised patients who had received at least one dose of study medication and with at least one efficacy evaluation while on treatment [...] In case of missing values at 8 or 22 weeks, the last prior on‐therapy value was carried forward (LOCF). Analyses of safety were performed on all patients who had received at least one dose of study medication."
Selective reporting (reporting bias)	High risk	No clear data about dropout rate in each group. Quote: "There were 33 withdrawals, nine of which due to adverse events (...). 118 patients completed the 6‐month study...".
Other bias	Unclear risk	Sponsorship bias cannot be ruled out.