HIV Self-Testing Africa Zambia (STAR Zambia)

The study will use the OraQuick® HIV Self-Test (Orasure Technologies, Thailand). The kits contain the HIV Oral Fluid Test (HIVOFT) test kit, stand, buffer solution, locally-translated instructions for use (IFUs), materials on counselling and linkage to care, and primary and secondary packaging. The kits will be distributed for free by CBDAs. Leaflets will be provided in lieu of patient information sheets, but no other consent will be required for HIVST implementation.

Following baseline enumeration, clinic pairs will be randomly allocated to either the intervention or control arm at a public ceremony. An opaque bag containing 2 balls - one for HIVST and one for SOC - will be used to allocate the HIVST arm.

It will not be possible to blind participants, CBDAs, or their supervisors to the cluster intervention allocation, but all forms will be managed without reference to the intervention arm. Outcome data will not be analysed until completion of the trial, with the exception of data analysis by an independent statistician for presentation to the Trial Advisory Group.

CBDAs linked to intervention clinics will be provided with training in HIVST and Information, Education and Communication (IEC) materials including flipcharts, used kits to show clients how to interpret positive, negative and inconclusive results, a cotton bud and vial of water to demonstrate the mouth swabbing and development process, leaflets and a buffer stock of OraQuick® HIV Self-Test to be stored in a locked container in their own home.

Older adolescent and adult participants (age 15 - 17years or older) wishing to know their status will be provided with information on where to obtain further counselling and care together with the kit, and an envelope for return of the used kit, a self-completed questionnaire (SCQ), a self-referral slip for the nearest clinic in case of a positive result, and information on how to access VMMC for HIV-negative men.

Clients will be encouraged to return their used kits confidentially to the CBDA, either in person or by posting in the sealed envelope into an opaque locked "ballot box" container kept in/at the CBDA's/counsellor's house. Clients will also be encouraged to seek post-test advice from the CBDA, which can be "generic" or results-based.

Kits will be replaced by the clinic supervisors on presentation of used kits and following inspection of an HIV test logbook to confirm recording of names and addresses, but not results, of clients. Numbers of used kits, and re-read results, will be recorded by the CBDA/clinic supervisor. HIVST logbooks will be kept at the CBDA's home in a locked container. Couple's testing will be encouraged. In the case of couples wanting to self-test together, both partners will be asked to attend the IEC session. Clients will be allowed to take up to 2 kits home if the partner cannot attend. Testing of children (aged 15 years or less) will not be permitted as part of this trial, but can be arranged through special arrangement with SFH supervisors.

There are no sharps or hazardous materials that will be used in this study.

BASELINE SURVEY SAMPLING AND INCLUSION/EXCLUSION CRITERIA

Sample size calculations

Baseline enumeration will include approximately 2400 participants in the intervention communities (400 per community) and 2,400 participants in the control communities (n=~4,800). Within this sample, 20% of participants (~1,000) will be randomly selected for the extended baseline survey. The sample size justification for the extended questionnaire is detailed below.

The sample size was calculated to ensure 80% power to detect a 50% change in the proportion of individual tested for HIV over the last twelve months, with 95% confidence. Using 2013-2014 Demongraphic and Health Survey (DHS) data, baseline rates for individuals tested in the last 12 months are estimated to be between 28.6-57.1% (lower in men than women); for this sample calculation, we have assumed a baseline testing rate of 50%. For a two-sample comparison of matched proportions across 6 pairs of matched communities, we estimate that it will be necessary to recruit around 400 respondents per community, or 4,800 respondents in total.