East Africa Point of Care Viral Load Study

Main Study Aims:

i) To determine the effectiveness of PoC VL monitoring in improving viral suppression among children, adolescents and young people living with HIV in East Africa.

ii) To evaluate feasibility and acceptability of using PoC VL monitoring among children, adolescents and young people living with HIV in East Africa.

iii) To understand the psychosocial life of children, adolescents and young people living with HIV in East Africa and how they predict adherence and viral load suppression

Objectives of Aim 1:

Primary objective i. To estimate the effect of PoC VL monitoring on viral load suppression among children, adolescents and young people living with HIV in East Africa at 6 and 12 months of follow-up.

Secondary objectives i. To describe the effect of PoC VL monitoring on the proportion of children, adolescents and young people living with HIV that experiences virological rebound after initial suppression within 6 and 12 months of follow-up.

ii. To describe the effect of PoC VL monitoring on time to initiation of intensive adherence counselling following virological failure among children, adolescents and young people living with HIV.

iii. To estimate the effect of PoC VL monitoring on the proportion of children, adolescents and young people living with HIV that experiences change of ART regimen within 6 and 12 months of follow-up.

iv. To determine the effect of PoC VL monitoring on the proportion of children, adolescents and young people living with HIV that is retained in care at 6 and 12 months.

Objectives of Aim 2 i. To assess the acceptability of the implementation and scale-up of PoC VL testing and monitoring from the perspective of children, adolescents, young people and their caregivers.

ii. To assess the critical determinants that may affect the implementation of PoC VL testing and monitoring from the perspective of healthcare workers and policy makers.

iii. To assess potential barriers and facilitators to implementation and scale-up of PoC VL testing and monitoring among children, adolescents and young people living with HIV.

iv. To assess the incremental cost-effectiveness of PoC VL from a modified societal perspective using established models, with data collected alongside the implementation of the intervention combined with data estimated based on existing studies.

Objectives of Aim 3 i. To psychometrically validate 'psychosocial life' questionnaires. ii. To iteratively develop and psychometrically validate a novel user-friendly tablet-based 'Psychosocial Life' game that predicts adherence and viral load.

iii. To develop and test a set of clinical micro-interventions for psychosocial support based on existing best practice.

iv. To measure how 'psychosocial life' mediates the PoC effect on adherence and viral load.

v. To test proof-of-concept of the use of the novel 'psychosocial life' game and associated micro-interventions in a clinical setting.

Design: A cluster randomized controlled trial with 14 intervention clusters and 14 control clusters.

Study Sites: Twenty (28) health facilities spread across 4 countries as follows: Uganda (8), Kenya (4), Tanzania (12), and Rwanda (4).

Population: The study will be conducted among three population sub-groups

• Study population 1: children, adolescents and young people (≤24 years) living with HIV in participating countries in East Africa.
• Study population 2: Care givers/ guardians and treatment supporters of children, adolescents and young people living with HIV in participating countries in East Africa
• Study population 3: Health care workers attending to children, adolescents and young people living with HIV, and policy makers.

Study Intervention: PoC VL monitoring using Cepheid Gene Xpert machine Control: Standard-of-care / centralized HIV VL monitoring

Duration: The whole project will take 48 months including 24 months of data collection. Each participant will be followed up for approximately 12 months.

Sample size The study will enroll 956 children, adolescents and young people (6months - 24years); with 476 in 14 intervention clusters and 476 in 14 control clusters (34 participants per cluster). The effect of the intervention and 95% CIs will be estimated using cluster-level summary methods and a power of 80%.

Study outcomes of Aim 1 Primary outcomes i. The proportion of children, adolescents and young people living with HIV that achieves viral suppression at 6 and 12 months of follow-up.

ii. The time between enrolment into the study and viral suppression.

Secondary outcomes i. The proportion of children, adolescents and young people living with HIV that experiences virological rebound after initial suppression within 6 and 12 months of follow-up ii. The time between enrolment into the study and initiation of intensive adherence counselling following virological failure.

iii. The proportion of children, adolescents and young people living with HIV that experiences change of ART regimen within 6 and 12 months of follow-up iv. The proportion of children, adolescents and young people living with HIV that is retained in care at 6 and 12 months.

Study outcomes of Aim 2 i. Assessed acceptability of the implementation and scale-up of PoC VL testing and monitoring from the perspective of children, adolescents and young people, and their caregivers.

ii. Identified critical determinants that may affect the implementation of PoC VL testing and monitoring from the perspective of healthcare workers and policy makers.

iii. Identified potential barriers and facilitators to implementation and scale-up of PoC VL testing and monitoring among children, adolescents and young people living with HIV.

iv. An assessment of the incremental cost-effectiveness of PoC VL from a modified societal perspective

Study outcomes of Aim 3 i. Quantified and psychometrically validated measurement of the psychosocial life of children, adolescents and young people living with HIV in East Africa ii. A novel psychometrically validated user-friendly tablet-based 'Psychosocial Life' game that predicts viral load based on psychosocial life.

iii. A list of proposed clinical micro-interventions for psychosocial support based on existing best practice.

iv. A model explaining how psychosocial life indicators mediate the PoC effect on adherence and viral load suppression.

v. A proof-of-concept of the use of the novel 'psychosocial life' game and associated micro-interventions in clinical settings.

Statistical Analysis Aim 1

To investigate the effect of the intervention on viral suppression our analysis will be carried out at the cluster level as follows:

• Reduce clusters to independent observations and provide summary statistics by comparing those in the intervention arm to those in the non-intervention arm.
• Use fixed effects regression to take into account unobserved time-invariant heterogeneity and the treatment effects across the clusters
• Analyse the main outcome (viral suppression) using time to event comparison of the two arms from enrolment to failure or completion of the follow-up time (2 years).

Mixed method Analysis Aim 2 Acceptability and feasibility will be investigated using a combined qualitative and quantitative approach, consisting of thematic content analyses of observations, in-depth interviews and focus group discussions, triangulated and generalized with descriptive statistics of survey data. As part of the feasibility study, the impact of a range of values in sensitivity analyses using each of the respective sites' 2020 per capita GDP as a benchmark cost-effectiveness threshold will be evaluated to be able to estimate the incremental cost-effectiveness of the PoC HIV VL compared to the standard of care. Quality adjusted life years (QALYS) will be calculated based on weights derived from the Global Burden of Disease.

Analysis for Aims 3 Step 1: Determining which psychosocial indicators will be used in the structure equation model on baseline data Pearson correlation analyses will be conducted to examine the strength of associations between all psychosocial life indicator variables and medication adherence. For that, a correlation table will be obtained for looking for multicollinearity. If any two explanatory variables have a Pearson's coefficient of 0.80 or greater one of the variables will be excluded from further analysis as they may measure the same underlying factor.

To determine statistically which psychosocial life indicators have the most predictive power and ascertain the extent to which selected indicators can predict adherence or viral load, a stepwise hierarchical multiple regression will be carried out. Potential predictors will be automatically added into the regression model in steps based on statistical algorithms. Only indicators that significantly improve the overall model fit (R square change) will be retained for the final model of children's psychosocial life.

Step 2: Building a structural equation model and testing how psychosocial life mediates the effect of PoC on adherence and viral load.

Following the RCT design, the goal of this analysis is to measure how psychosocial life indicators mediate the PoC effect on adherence and viral load. An initial hypothetical psychosocial life model will be developed based on variables proved to have the most predictive power obtained from phase one and a synthesis of the empirical literature. The outcome variable will be the change in adherence and viral load separately (difference between baseline and endline data). From the main study (WP1) we assume that PoC improves the medical adherence of children with HIV. We also hypothesize that PoC will improve psychosocial life indicators and that better psychosocial life improves medical adherence and viral load.

Three different types of mediational model will be verified to test the potential mediating role of psychosocial life on the relationship between PoC and medication adherence: (a) a single mediator model, (b) multiple parallel mediator model, and (c) SEM model considering psychosocial life as a latent variable indicated by selected indicators.

Inclusion criteria

• Age 6 months - 24 years
• Documented evidence of HIV infection
• Receiving ART for treatment of HIV infection for at least 6 months.
• Has had a detectable VL of >1000 copies/ml in the last 6 months.
• Guardian, parent or legal representative able and willing to give voluntary consent and sign/mark an informed consent document.
• Willing and able to comply with protocol requirements/study procedures.

Exclusion criteria

• Any medical conditions that require pausing of ART for more than three months.
• Potential participant already enrolled in another study which may interfere with the study outcome or participation as per investigator's judgement.
• Child, adolescent or young person already enrolled and completed follow up in the current study.
• Any medical or other condition in the potential participant or their parent/ guardian that precludes provision of informed consent/ assent or that may hinder achieving study objectives as per investigator's judgement.