2.0 Aims The overarching aim of this study is to evaluate the impact and effectiveness of the scale up of the DREAMS HIV prevention package of biological, behavioural and social interventions in reducing HIV incidence in adolescent girls and young women residing in the uMkhanyakude district of KwaZulu-Natal. 1. To measure the impact of combined HIV prevention package for adolescent girls and young women (DREAMS) package, including community, facility, and school-based interventions, on HIV incidence and other key outcomes among adolescent girls and young women (AGYW) and their male partners 2. To identify pathways of protection for DREAMS effect on health, education and social well-being of AGYW 3. To explore the process and experience of DREAMS on the community 2.2 Primary and Secondary Objectives of the Evaluation The primary objective of the evaluation is to measure whether HIV incidence and HIV-related outcomes change due to DREAMS at a population level. To investigate this, the main approach will be to assess changes in outcomes over time in relation to DREAMS roll-out (before, during and after). Where data are available, we will also assess: • Change by dose, i.e., assessing dose-response relationships between DREAMS exposure and outcomes at the individual level and small-area level. ‘Layering’ of different DREAMS interventions will be a component of dose. • Change by place, where data are available from both DREAMS and non-DREAMS sites. The second objective is to explain the observed changes, specifically to explore the pathways of protection by which DREAMS interventions influence the lives of young women and ultimately their HIV status. Pathway analyses will be guided by the Theory of Change described below. 2.3 Study Hypotheses and Theory of Change Illustrated in Figure 1, we hypothesise that DREAMS will work through three main pathways of protection: a. Behaviour as prevention – this pathway reduces acquisition of HIV by promoting safer sexual behaviours among AGYW and their male partners including the consistent use of condoms. b. Social protection – this pathway reduces the social and economic vulnerability of AGYW by keeping them in school longer and enabling financial independence to offer socio-economic alternatives to early marriage and transactional sex – all of which are associated with HIV risk. c. Biological protection – through this pathway, the likelihood that AGYW will acquire HIV is reduced through biomedical modes that prevent transmission of the virus, e.g., by reducing the viral load (uptake of antiretroviral therapy; ART) and acquisition of HIV (uptake of voluntary medical male circumcision; VMMC) among male sexual partners, as well as pre-exposure prophylaxis (PrEP) offered to AGYW themselves. All three pathways can be influenced by ‘mediators of change’ such as aspiration, empowerment and resilience; the role of such mediators on outcomes and impact will be explored. 2.4 Study Endpoints Table 1 summarises the primary and secondary outcomes, for measures of impact to be captured among the total population over time before, during and after DREAMS roll-out (Evaluation Objective 1). The primary endpoint is HIV incidence among young women aged 15-24 years. We will also measure the effect of DREAMS on HIV-related outcomes that lie on the three pathways of interest: biological, behavioural and social protection. 2.5 Study Settings uMkhanyakude, South Africa: The Africa Health Resaerch Institute (AHRI) in uMkhanyakude has followed over 160,000 members (92,000 at any one time) from 11,000 geocoded households in a 428km2 surveillance area through 4-monthly demographic surveys (will be annually from 2017) and annual collection of individual behavioural and HIV biomarkers for on-going longitudinal HIV incidence (a total of 2700 sero-converters since 2003). Since 2007, all residents aged 15 years and above are eligible for HIV testing, along with a stratified, random sample of 12.5 percent of the non-resident population. This includes a dried blood spot for HIV testing and HIV viral load in those who are eligible, with the capacity to do HIV genome sequencing. The AHRI has a memorandum of understanding with the Department of Health which enables linkage of the population surveillance data at individual level to the electronic record systems in the local health care facilities (2010 onwards; 29,000 patients and 34,000 admissions) and TIER.Net for HIV treatment (2004 onwards and 39,000 patients). The AHRI also has access to all clinical laboratory test results of patients in the sub-district through linkage with the National Health Laboratory Systems database (since 2004). Moving forward the AHRI (a combination of AC with KwaZulu-Natal Research Institute for tuberculosis and HIV –K|RITH) will support the population surveillance platform on which this study is based.