Bowden 2010.

Methods	Study design: an international, multicentre, randomised, open‐label, parallel‐group, equivalence study
Participants	Diagnosis: bipolar I participants experiencing a manic or a mixed episode Method of diagnosis: according to the DSM IV Age: for lithium, median = 38.2 (SD = 13.1) years; for valproate, median =38.8 (SD = 12.0) years; range = (18 ‐ 75 years accepted) Sex: lithium 85 women; 50 men, valproate 66 women; 56 men Location: the study was conducted in 21 centres in six countries (Bulgaria, Hong Kong, Malaysia, Russia, Taiwan, and Thailand) Co‐morbidities: only minor anxiety disorders, no axis 2 or substance misuse problems Adjunctive therapy: none Adjunctive medication: concomitant administration of lorazepam (up to 8 mg/day from day 0 to day 7 and up to 4 mg/day after day 7) or equivalent (diazepam or other benzodiazepine) was permitted to manage agitation, irritability, restlessness, insomnia, or hostility. In addition, nonbenzodiazepine hypnotics (zolpidem 10 mg or zopiclone 7.5 mg per night) or antidepressants other than fluoxetine could be given if needed for the management of insomnia or emergent depression.
Interventions	Participants were randomly assigned to either: Experimental arm ‐ lithium N = 132 Duration: 12 weeks Treatment protocol: "After a screening visit at which inclusion criteria were assessed, a 3‐day wash‐out period was initiated for all psychotropic drugs, except benzodiazepines at doses of 8 mg/day lorazepam equivalents. This could be reduced to 1 day in case of worsening severity of mania. lithium was provided as scored tablets of 250, 300, 400, and 500 mg of lithium carbonate. A sustained release form was recommended when this was available. lithium was started at the nearest dose to 800 mg/day orally (600–900 mg/day depending on the available formulations in individual countries), divided into two daily doses, for the first 5 days, after which dose adjustment was permitted. Target serum concentrations were 0.8–1.2 mmol/L for lithium and 70–125 mg/mL for valproate. If major side effects occurred, dose reduction to 15 mg/kg/day for valproate and to achieve serum concentrations of 0.6–0.8 mmol/L for lithium was allowed". Therapist/face‐to‐face contact: not described Comparator arm ‐ valproate N = 122 Duration: 12 weeks Treatment protocol: "After a screening visit at which inclusion criteria were assessed, a 3‐day wash‐out period was initiated for all psychotropic drugs, except benzodiazepines at doses of 8 mg/day lorazepam equivalents. This could be reduced to 1 day in case of worsening severity of mania. valproate was provided as a sustained‐release formulation in 200, 250, 300, and 500 mg tablets depending on the country. valproate was started at the nearest dose to 20 mg/kg/day orally for the first 5 days, after which the dose was adjusted as a function of response and serum concentration of valproic acid at day 5 and at any subsequent study visit at the discretion of the investigator. If the total dose did not exceed 1000 mg/day, valproate was administered once a day, otherwise a twice daily regimen was implemented".
Outcomes	Timepoints for assessment: baseline, 5 , 10, 21, 56, 84 days Primary outcome: Mean change in YMRS score from baseline to study end. Secondary outcome: Completion of study Discontinuation Change from baseline to study end in the YMRS score in the ITT population Response rates (defined as a reduction of at least 50% in YMRS scores) CGI‐BP bipolar severity scores MADRS Score Remission rate: YMRS ≤ 12 and no increase in MADRS Remission rate: YMRS ≤ 12 remission rate and decrease in CGI‐BP ≥ 2 YMRS item scores Use of anxiolytics Adverse events Blood tests Change in body weight
Notes	Notes Date of study: January 2004‐February 2006 Funding source: "This study was supported, sponsored, and funded by sanofi‐aventis, manufacturer of sodium valproate. The study sponsor, sanofi‐aventis, initiated the study, chose the study investigators, provided study medication, coordinated the data analysis, and provided financial support for the conduct of the study and the preparation of a draft version of this manuscript by a medical communications agency (SARL FOXYMED, Fresnes, France). THERAMIS/MEDISCIS, a contract research organization, was responsible for data management, statis‐ tical analyses and production of the study report". Declarations of interest among the primary researchers: "The authors of this manuscript constituted the Steering Committee of the study. Sergey Mosolov, Luchezar Hranov, Eric Chen, Hussain Habil, Ronnachai Kongsakon, and Hsin‐Nan lithiumn also participated in the study as investigators. All academic authors have received honoraria from the sponsor for participation in the study as well as, in some cases, consultancy fees in the previous 3 years. Robert Manfredi is an employee of the study sponsor. The corresponding author, Charles L. Bowden, chairman of the Steering Committee, had full access to all data from the study, and was responsible for the decision to submit the finalized manuscript for publication".
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"The randomization code was generated centrally. Investigators were provided with randomization numbers in sealed envelopes in blocks of four".
Allocation concealment (selection bias)	Unclear risk	Not described
Blinding (performance bias and detection bias) All outcomes	Low risk	"Double‐blind. Wherever possible, raters were to be blinded to treatment. Although it was intended that patient assessment would be conducted by blinded raters, this was not always possible given the pragmatic, clinical practice base of the study, which could introduce a source of evaluation bias".
Incomplete outcome data (attrition bias) All outcomes	Low risk	One hundred and eighty‐five participants completed the study as planned (69.0% of randomised participants).
Selective reporting (reporting bias)	Low risk	Good reporting
Other bias	Low risk	None identified