The health system’s support of optimal engagement with HIV care has been placed at the “top of the list of policy priorities”: optimal engagement could facilitate the benefits of accessible treatment in controlling the HIV epidemic and maintain the public health gains that scaling antiretroviral therapy has delivered. Attempts to improve engagement are often targeted at ‘high risk’ individuals – such as men and youth under 25 years. However, most people who struggle to stay engaged are low to moderate risk, and middle-aged women make up the majority of the patients both in and out of care. Thus, better ways to segment the population are needed to target interventions efficiently.
This study aims to explore engagement behaviour through the patterns of measurable, longitudinal indicators such as retention and adherence, using survival analysis, multi-state analysis and group based trajectory modelling in a cohort from Khayelitsha and Gugulethu, South Africa. This could gain insight into
• Engagement behaviour and its dynamics in this population
• Patient phenotypes based on behavioural patterns as well as demographics
• The impact of longitudinal measurement and latent class methods on inferences from routine data