NCT01848210.

Study characteristics
Methods	Study design: randomised, double‐blind, placebo‐controlled Method of randomisation: "by chance, like flipping a coin" Exclusions post randomisation: 166 Losses to follow‐up: 36
Participants	Country: Brazil Setting: not specified Number: 829 (411 experimental group and 418 placebo group) included and 711 analysed (383 experimental group and 388 placebo group) Age (mean): 56 (18 to 75 years old) Gender: 83 men and 688 women Inclusion criteria: consent of subject or legal representative. Men or women of any ethnicity, aged between 18 and 75 years, and BMI equal or less than 40. CVI in the reference leg with the clinical classification C3, or C4a or C4b or C5, stable edema. Scoring in "Severity Score of Local Complaints" equal to or higher than 5 total points. Women who are using an effective birth control or who are postmenopausal Exclusion criteria: Deep vein insufficiency or venous obstruction and/or DVT and/or presence of phlebitis in lower limbs during the last 3 months.Surgery at the venous system or sclerotherapy or any treatment for CVI during the last 3 months. History of known or suspected allergy or intolerance to any of the ingredients of the medicinal product under investigation. Serious systemic disease. Hepatitis A, hepatitis B, or C or any liver disease. Use of diuretics. Diabetes insulin‐dependent. History of alcoholism, drug abuse, psychological or emotional problems
Interventions	Treatment: Coumarin 30 mg, troxerutin 180 mg fixed‐dose combination tablets (Venalot), orally, 3 times daily Control: placebo Duration: 16 weeks Follow‐up: 18 weeks
Outcomes	Primary: Mean change from baseline in volume of reference leg at week 16 using a water plethysmometer Secondary: Change from baseline in local complaint severity (eight symptoms assessed by a lickert scale) Overal assessment by the investigator Number of participants with adverse events
Notes	Sponsor: Takeda. Results published in clinicaltrials.gov
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Participants will be randomly assigned (by chance, like flipping a coin) to one of the two treatment groups"
Allocation concealment (selection bias)	Unclear risk	Comment: not specified.
Blinding (patients)	Low risk	Quote: "Placebo (dummy inactive pill) ‐ this is a tablet that looks like the study drug but has no active ingredient. All participants will be asked to take two tablets three times a day throughout the study"
Blinding (study researchers)	Low risk	Quote: "Placebo (dummy inactive pill) ‐ this is a tablet that looks like the study drug but has no active ingredient. All participants will be asked to take two tablets three times a day throughout the study"
Blinding (outcome assessment)	Low risk	Quote: "Placebo (dummy inactive pill) ‐ this is a tablet that looks like the study drug but has no active ingredient. All participants will be asked to take two tablets three times a day throughout the study"
Incomplete outcome data	Low risk	Comment: there was a IIT analysis (patient that taken the treatment at least 28 days) and a PP analysis. The total losses were 166 (20%)
Selective reporting	Low risk	Comment: all results are published.
Other bias	Low risk	Comment: none detected