Table 1 Evaluation measures

 Benchmarks

26 yes/no items (1 point for each done) measuring implementation progress at each site. Numeric summaries are also recorded for peer leaders trained, 7-session peer groups and boosters held (# completed, # of adult and youth participants). Scored every 6 months. Summaries for each community and the total are presented to the community and district leaders and integrated with related local health outcome data (e.g., # of HIV tests provided, % of pregnant women tested) routinely collected and provided by the District Health Officer.

 Peer group fidelity ratings

• Structured fidelity ratings by the Research team (trained to an inter-rater reliability of 0.85) [37]. At least one of the early sessions of all peer leaders plus randomly selected groups thereafter is observed.

• Peer leaders’ self-assessments and attendance sheets

 Observation notes on meetings

Study notes by the research team (semi-structured observations of all meetings attended) documenting type of meeting or observation, date, site, and discussion related to successes, challenges, if problem was solved; and personal interactions

 Annual focus group & Individual Interviews

7–12 focus group interviews with district and community implementation partners and 8–12 individual interviews with implementation partners including local “champions” and unsupportive individuals. Coded to identify successes, problems/barriers/ solutions, and appraisal of support and technical assistance

 Survey

• Demographics (e.g., Age; education; current partner status)

• Program time/dose: Date of last session attended; # of sessions attended

• Comprehensive HIV-related Knowledge: (5 HIV & 4 PMTCT items [35])

• Safer sex Practices over the last 2 months [10] (Primary Outcome) (a) Self-efficacy scale (4 items, # discussed: HIV prevention, condom use, being faithful, HIV test with partner; responses very, somewhat, not confident, range 3–12); (b) Partner communication index (4 yes/no items, # discussed: HIV prevention, condom use, being faithful, HIV test with partner); (c) If abstained; d) If sexually active, condom use (always, sometimes, never);

• Sexual Risk Index [10] (had unprotected sex, more than one partner, sex for money, sex while drinking, sex with a stranger; 5 items, # Yes)

• If used a condom at last sexual relations [35]

• If not sexually active at baseline, whether sexual debut occurred (youth only)

• HIV Test (Primary Outcome): If had HIV test in last year, yes-no; if yes, whether tested with partner [10]

 STIs

• Original Plan: Tests for clinical and sub-clinical gonorrhea, chlamydia, and syphilis using minimally invasive rapid tests appropriate for use in a community setting without electricity or access to laboratory or examination facilities. This plan had to be modified because tests were withdrawn from the market and/or found to be unreliable in new research [45, 46].

• Revised Plan: Detailed interview regarding STI symptoms experienced in the last week and in last 6 months (with referral for clinic follow-up when indicated)

 All above data

Within and cross-case matrices: implementation patterns and relationship to progress benchmarks, effectiveness, if sustained program