Home-based intervention to test and start (HITS) protocol: a cluster-randomized controlled trial to reduce HIV-related mortality in men and HIV incidence in women through increased coverage of HIV treatment

The trial is delivered in a four strata using a 2 × 2 factorial design to compare the effectiveness of two interventions, micro-incentives and male-targeted HIV-specific decision-support, called EPIC-HIV (Empowering People through Informed Choices for HIV), in 45 clusters (week blocks) using the Africa Health Research Institute (AHRI)‘s HIV surveillance platform (Fig. 1). The four strata are Arm 1 micro-incentives; Arm 2 both micro-incentives and EPIC-HIV (male-targeted HIV decision-support); Arm 3 EPIC-HIV (male-targeted HIV decision-support); and Arm 4 standard of care (SoC).The study protocol follows the Standard Protocol Items: Recommendations for Intervention Trials (SPIRIT) [17] and the Consolidated Standards of Reporting Trials (CONSORT) statement for reporting RCTs [18].

The study is registered at the National Institute for Health’s ClinicalTrials.​gov. Enrolment started in February 2018 and follow up will be completed in December 2021.

The trial is conducted in the Hlabisa sub-district of the uMkhanyakude district located in northern KwaZulu-Natal South Africa between 2018 and 2021 (Fig. 2). The area is predominantly rural and has recorded one of the highest population HIV prevalence estimates globally [20]. In this sub-district, AHRI has one of the largest population-based Demographic and HIV surveillance since 2000. The surveillance covers 432 km² geographic area over approximately 100 000 individuals and 60 000 residents at any given time. Trained field-workers visit all households in the surveillance area and interview a key resident informant annually. The household survey records demographic components of households including size, composition, fertility, migration, mortality, and nuptiality. The HIV surveillance is annually conducted on all resident individuals aged 15 years or older to collect sexual behaviour and general health data and dried blood spots (DBS) samples for anonymized HIV testing [21]. The fieldwork visits are carried out in systematic cycles of 45 "week blocks". A week block is a workload-equivalent area developed using a GPS based methodology [22]. AHRI works closely with the local Department of Health in a comprehensive HIV Treatment and Care Programme, decentralized to the 17 primary health care clinics in the sub-district. HIV testing and treatment are available at all 17 primary health care clinics and the district hospital. Since 2017, AHRI has introduced home-based rapid HIV testing as part of the annual HIV surveillance.

The two interventions, micro-incentives and the male-targeted HIV-specific decision support, are offered in a two-stage scheme. The first stage is aimed at encouraging HIV testing (Fig. 3), while the second stage is aimed at encouraging linkage to care if tested HIV positive (Fig. 4). Micro-incentives are in the form of a R50 food voucher (redeemable in a local supermarket) conditional on undergoing a home-based rapid HIV test. If diagnosed with HIV and present for HIV care in one of the local Department of Health clinics servicing the program area within 6 weeks of the HIV test, participants receive a second R50 food voucher.

The male-targeted HIV-specific decision-support is implemented via a tablet-based application called EPIC-HIV. EPIC-HIV development was guided by a combination of human behaviour change theory and person-based intervention design [23] and human computer interaction techniques. The development process (described elsewhere in more detail) included three phases: 1) a thorough review of literature was conducted to synthesize key facilitators and barriers of HIV care and management among men in South Africa; 2) the review informed content, design features and format of the application, and the app was iteratively prototyped; 3) the content was evaluated with the community advisory board, and usability testing and in-depth interviews were conducted with representative users at every iteration to establish understanding of content and usability issues. The content is rooted in local narratives and provided in a form of experiential information to increase risk perception, salience and likelihood of response [24, 25]. EPIC-HIV makes the decision to test or link to care explicit by providing experiential information about what various outcomes might be like (imagined future). It further supports three basic human ‘needs’ as defined by the self-determination theory - the need for autonomy, competence, and relatedness [26]. The app is available in two versions (EPIC HIV-1 and EPIC HIV-2). EPIC-HIV 1 is offered to men at the point of HIV test offer prior to HIV counselling and testing to support them  when choosing whether or not to test for HIV. If participants are diagnosed with HIV but do not present in the local Department of Health clinic within a month of the positive HIV test, a study tracker team re-visit participants at home and offer EPIC-HIV 2, which is designed to address barriers to seek HIV treatment and encourage them to link to HIV care.

Male residents in the clusters that receive both interventions are offered EPIC-HIV 1 prior to HIV testing, a food voucher conditional on undergoing a rapid HIV test, and a second food voucher when they link to care within 6 weeks of receiving an HIV positive test. If they do not link to care within a month of the positive HIV test they become eligible for EPIC-HIV 2. Female residents in these clusters are eligible to receive the food voucher only. Residents in the control clusters receive standard of care where they are offered the home-based HIV rapid test, and if diagnosed with HIV, referred to the clinic for linkage to HIV care.

The goal of the trial is to establish whether the provision of micro-incentives and a male-targeted HIV-specific decision-support app will increase uptake of HIV testing and linkage to HIV care thus, ultimately reducing HIV-related mortality in men and population-level HIV incidence in young women. The trial’s primary endpoints are to measure the impact of the interventions at both individual and population-levels as follows:

Specifically, the study compares differences in the percentage that received the rapid HIV test at home among those offered the test in men among the intervention arms compared to the SoC arm. The study will also compare the percentage who visited study clinics and initiated ART in men at 1-year after the home-based HIV-positive test, as well as prevalence of detectable viremia (detection limit of > 1550 copies/ml) obtained from the DBS in men. Changes in the number of HIV-related deaths and mortality rates in men and changes in the number of new HIV infections and HIV incidence rates in young women 3 years after the home-based HIV test will be compared across the arms. Secondary outcomes that will be measured in the study are shown in Table 1.

Participants are enrolled into the HITS trial during the ongoing annual individual HIV surveillance and routinely followed up between 2018 and 2021 (BREC Ref: BE290/16).

Trained fieldworkers visit participants at home once a year to conduct household and individual surveys using tablet-based computers. Data on demographic information, general health, and sexual behaviour (including contraceptive use, sexual activity and pregnancy history) are collected via the REDCap database [27]. Participating individuals are offered rapid HIV testing. Consenting individuals undergo confidential HIV pre- and post-test counselling according to the Department of Health procedures. A rapid HIV test is performed and test results are provided approximately 20 min after the testing. Participants who consent to the HIV surveillance have blood collected on a filter paper as DBS, which are stored at the AHRI Biorepository in Durban [28]. A 4th generation HIV ELISA (Genscreen™ ULTRA HIV Ag-Ab (Biorad, Marnes-la-Coquette, France) adopted and evaluated in-house by the AHRI Diagnostic Research Laboratory (unpublished data) is used for testing of all DBS samples. All HIV ELISA positive DBS samples are subjected to HIV viral load determination using the automated Nuclisens EasyMag® (Biomerieux, Marcy-l’Etiole, France). Viral load testing is performed on the same day as the extraction.

Individuals who test positive for HIV are referred to the nearby Department of Health clinics to receive HIV care within 10 days of the HIV test date. Data on linkage to care is collected from the Department of Health through TIER.​net (i.e. an electronic patient records management system that contains information on all clinic visits for people on ART, including data for all patients attending the 17 clinics in the Hlabisa health sub-district and the Hlabisa hospital) and clinic attendance data (where date and reason for attendance are collected for all consenting individuals in the programme area who attend one of the 11 clinics serving the programme area). All deaths are ascertained by verbal autopsies with closest care givers using a questionnaire based on the INDEPTH/World Health Organization (WHO) standard questionnaire [29].

Data is collected and managed by AHRI Research Data management within the routine AHRI HIV and demographic surveillance study databases as per AHRI comprehensive study procedures. The database has strictly restricted access via a data enclave on a secure server.

Randomisation was performed by the trial statistician prior to intervention rollout to ensure that each arm has similar baseline HIV incidence rates among women aged 15 and 30 years. First, all 45 clusters were stratified by HIV incidence in young women using data from 7605 young women aged 15 and 30 years old, who were residents in the PIPSA between 2006 and 2015 at least for 1 year. A total of 34551 person years and 1636 seroconversion events have been recorded, yielding overall 4.7 new infections per 100 person years. Then the 45 clusters were divided into four strata according to their HIV incidence rate such that stratum 1 comprises clusters with the lowest incidence rate, while stratum 4 comprises clusters with the highest incidence rate. Each of Arm 1, 2, and 3 consists of two clusters from each of the four incidence stratum (thus a total of eight clusters per arm). Arm 4 (SoC) consists of five clusters from each of the incidence stratum 1, 2, and 4, and six clusters from incidence stratum 3 (thus a total of 25-clusters). The study is an open label trial.

In order to determine how the implementation of the HITS intervention is achieved in our setting as well as the processes through which the interventions affect outcomes, we will gather stakeholder views in a purposive sample of up to 150 stakeholders (HITS participants, fieldworkers and health professionals) using mixed methods of quantitative and qualitative techniques (e.g. surveys and in-depth interviews). The evaluation is shaped by the UK Medical Research Council (MRC) guidance on the process evaluation of complex interventions [34]. The guidelines identify three important pathways of understanding the process of achieving a particular outcome: a) Context – we will examine how the socio-cultural context of our setting influences the development process, delivery and functioning of the HITS interventions; b) Implementation of the HITS program - we will describe the resources and processes through which HITS implementation is achieved, and the extent to which HITS intervention would be delivered as intended or needed any program adaptations; c) Mechanisms of impact - we will examine how study participants respond to and interact with HITS interventions and whether these interventions promote behavioural change or not. In order to examine the ‘causal assumptions’ of the intervention, the evaluation is informed by the Normalisation Process Theory (NPT) [35]. The theory will help us to understand how the HITS intervention would be delivered, as well as any changes that could affect its fidelity. All interviews will be recorded and transcribed and subject to thematic analysis following an interpretivist approach.

We will conduct a cost-effectiveness analysis of the two interventions, the micro-incentives and EPIC-HIV, from the societal perspective. A time and motion study will be conducted to determine how much time fieldworkers spend on the HITS related activities including consenting for study participation, HIV testing, counselling, and provision of micro-incentives and/or EPIC-HIV. Detailed information on direct and indirect costs will be documented during the study implementation to determine the unit costs of providing the interventions. Effectiveness will be measured as the number of uptake of HIV testing, new HIV diagnosis, as well as linkage to care. We will also fit mathematical modelling to simulate dynamic HIV transmission calibrated to the data from the trial and the AHRI surveillance and estimate disability-adjusted life years (DALYs) over 10 years. We will calculate the incremental cost-effectiveness ratio (ICER) for adding the interventions to standard of care.

The study is ongoing.  Intervention delivery started in February 2018 and ended in December 2018. Participant follow-up will continue until the end of 2021. No articles containing the results of this study have been published or submitted for publication.