Garcia 2012.

Methods	Design: cluster‐RT (by city; n = 20) Groups: intervention group (management and prevention of STDs); control group (standard care)
Participants	Pharmacies: 773 Pharmacy workers: 2292 Median age: 34.6 years % female: 62% ‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐ Pharmacy user: targeted through pharmacies (data available from 12930 young adults) Setting: urban Country: Peru
Interventions	Pharmacy worker‐directed intervention: the intervention comprised 4 modalities: strengthened STD syndromic management by pharmacy workers and clinicians; mobile‐team outreach to female sex workers for sexually transmitted infection screening and pathogen‐specific treatment; periodic presumptive treatment of female sex workers for trichomoniasis; and condom promotion for female sex workers and the general population. Delivered by: pharmacist and midwife team Type: education materials, based on clinical practice guidelines; interactive meeting with role play, case studies Mode of delivery: group of 8 to 10 pharmacy workers; written materials Duration: 4 x 90‐minute luncheon training seminars on STD/HIV. Monthly follow‐up visits to discuss STD/HIV prevention and provide materials. Follow‐up: 1, 3 and 6 months after training Pharmacy worker control: usual treatment ‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐ Pharmacy user‐directed intervention: not reported
Outcomes	Pharmacy worker: Uptake: not reported Behavioural: simulated patients for management of STD ‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐ Pharmacy user: Clinical: infection with STD Psychological health: not assessed Behavioural: not assessed Quality of life: not assessed Process: not assessed Costs: not assessed
Notes	Study/intervention name: Peru‐PREVEN Funding Source: Wellcome Trust and Burroughs Wellcome Fund, National Institues of Health, Centre for AIDS Research, CIPRA, and USAID‐Peru. Garcia 2012 (cited under Garcia 2012) also reported on this study.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer randomisation
Allocation concealment (selection bias)	High risk	Participants in outcome surveys were recruited after city randomisation, precluding allocation concealment
Baseline outcome measures similar	Low risk	Adjusted for in analysis
Baseline characteristics similar	Low risk	Adjusted for in analysis
Incomplete outcome data (attrition bias) All outcomes	Low risk	Missing data replaced with classification of negative composite endpoint
Blinding of outcome assessment (detection bias) All outcomes	Low risk	US and UK investigators masked until testing complete
Protection against contamination	Low risk	Cluster randomised
Selective reporting (reporting bias)	Low risk	Primary and secondary outcomes reported
Other bias	Low risk	Not noted
Blinding of participants and personnel (performance bias) All outcomes	High risk	Fieldworkers and Peruvian study team could not be masked