Implementing ecological momentary assessments to measure violence and adolescent HIV transmission risk: Lessons from Johannesburg, South Africa

Abstract

Ecological Momentary Assessment (EMA) is an important methodology to understand risky behaviour and holds promise for HIV research. EMA is still novel in sub-Saharan Africa. We describe challenges and lessons learned on a novel study implementing mobile phone EMAs with adolescent boys in South Africa. The Tsamaisano study was a longitudinal study from 2020-2023 to recruit adolescent boys aged 15-19 years; including those without HIV and those perinatally infected and living with HIV. Participants were prompted to complete 52 weekly mobile phone survey on emotional state, exposure to and perpetration of violence, and sexual risk behaviour. Surveys were delivered using a random algorithm to choose the day. We incorporated mechanisms to assess challenges and optimize survey completion: weekly team meetings with youth representation and real-time data monitoring. Additionally, 20 frequent vs infrequent survey submitters participated in qualitative interviews about barriers and recommendations. Real-time monitoring indicated low (defined as <50%) survey completion in the first months of study implementation. To ensure that both the adolescent participant and their caregiver understood the commitment required for successful EMA, we created and implemented a guided discussion around mobile phone access during the enrolment visit. We identified a need for increased and ongoing technical support; addressed by creating technical guides, implementing a standard two-week check-in call after enrolment, adding an automated request button for call-back assistance, creating a WhatsApp messaging stream, and reaching out to all participants failing to submit two sequential surveys. Entry-level smartphones, including those initially distributed by the study, did not have capacity for certain updates and had to be replaced with more expensive models. Participants struggled with randomly allocated survey days; completion improved with set completion days and targeted reminder messages. Together, these steps improved survey completion from 40% in December 2020 to 65% in April 2022. We describe key lessons learned to inform future study designs with mobile phone EMAs, drawing on our experience implementing such among adolescent boys, including persons living with HIV, in a low-and-middle income setting. The key lessons learned through the Tsamaisano study are important to inform future study designs with EMA utilizing mobile phone, electronic data collection among adolescent boys in low-and-middle-income settings.