Welch 1985.

Study characteristics
Methods	Study design: randomised, double‐blind, placebo‐controlled Method of randomisation: not stated Exclusions post randomisation: none Losses to follow‐up: 7/147 (5%)
Participants	Country: Belgium Setting: hospital Number: 147 patients Age: mean 44.5 ± 14 years active group; mean 43.6 ± 14 years placebo group Gender: 26 M:119 F Inclusion criteria: CVI with oedema and ≥ 1 related symptom Exclusion criteria: elastic stockings or compressive bandages; leg oedema from another origin; arterial insufficiency; superficial thrombophlebitis; varicose eczema or ulcer; diuretics, analgesics, steroids, NSAIDs or other venous drugs; pregnancy
Interventions	Treatment: O‐(beta‐hydroxyethyl)‐rutoside 1000 mg per day Control: placebo Duration: 28 days Follow‐up: 28 days
Outcomes	Primary Symptoms ‐ pain, cramps, heavy legs, swelling, restlessness, itching and paraesthesia measured by a semiquantitative scale (0 to 3) Oedema ‐ pitting measured by a semiquantitative scale (0 to 3), circumference of ankle and calf Secondary Side effects Global opinion of investigators and participants
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Comment: method of randomisation not given
Allocation concealment (selection bias)	Unclear risk	Comment: method of allocation concealment not given
Blinding (patients)	Unclear risk	Comment: no information given about methods used for blinding
Blinding (study researchers)	Unclear risk	Comment: no information given about methods used for blinding
Blinding (outcome assessment)	Unclear risk	Comment: no information given about methods used for blinding
Incomplete outcome data	Low risk	Comment: number of participants in each group described, and inclusion and exclusion criteria reported as well for the most important characteristics. Number of participants who dropped out prematurely given, along with numbers of and reasons for adverse events
Selective reporting	Low risk	Comment: protocol identified and no differences identified between protocol and article
Other bias	Low risk	Comment: none detected