Barbosa‐Cesnik 2011.

Methods	Study design: parallel RCT Power calculation: yes Intention‐to‐treat analysis: no
Participants	Inclusion criteria Setting: Women presenting to a health service with symptoms of UTI Country: USA Women 18‐40 years, with UTI symptoms, residing in Ann Arbor next 6 months Number: 419 randomised; 319 analysed Average age: 21 years Previous UTIs: 3‐4 previously; 1 in previous year Exclusion criteria Antibiotics in past 48 hours; hospitalisation or catheterisation within past 2 weeks; kidney stones; diabetes; pregnancy; cranberry allergy; negative urine culture
Interventions	Treatment group Low calorie cranberry cocktail: 240 mL (8 oz) twice a day Mean PAC: 112 mg/240 mL Control group Placebo drink: same volume matched for flavour and colour Treatment duration: 6 months
Outcomes	Primary outcome: UTI (≥ 10³ cfu/L of known pathogen) Secondary outcome: urinary symptoms and vaginal symptoms at day 3, 1‐2 weeks, and ≥ 1 month
Notes	Compliance measured by direct questioning
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer generated
Allocation concealment (selection bias)	Low risk	External, web based allocation
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Placebo drink matched, participants and clinicians blinded
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	NS
Incomplete outcome data (attrition bias) All outcomes	High risk	100 participants randomised but no outcomes reported for them, they were actually not eligible to be randomised since they were culture negative
Selective reporting (reporting bias)	Low risk	UTI is most appropriate outcome
Other bias	High risk	Selection bias, representative nature of consenters is questionable Source of funding: National centre for alternative medicine at NIH