Transition Dynamics and Treatment Outcomes Among HIV-Positive Patients of Ifakara, Tanzania: A Comprehensive Analysis for Optimizing the Existing Longitudinal Cohort Data

Abstract

The in-and-out of care movements at various stages of HIV treatment require a specialized approach for estimating transition probabilities between stages and the association between transitions and characteristics. The multistate models are appropriate for such a problem. However, their application in the HIV cascade and continuum of care is scarce. Furthermore, there are limited assessments of engagement in care in combination with treatment outcomes such as viral load (VL), and cascade assessments at the sub-national level in Tanzania are limited. Therefore, the goal of this project was to investigate the transition dynamics and virologic outcomes among PLHIV enrolled in care at the Chronic Disease Clinic of Ifakara (CDCI), Tanzania, from 2005 to 2024. The project had two parts: the first was the systematic scoping review of methods applied in evaluating the HIV cascade and continuum of care. In this review, descriptive statistics were used to summarize findings of the scoping review. The second part included the empirical analyses of existing data from CDCI. In these analyses, descriptive statistics, continuous-time non-parametric multistate models, and Cox proportional hazard models were used to evaluate engagement in HIV care and treatment outcomes. The review showed 93% (279/300) of the articles reviewed applied cross-sectional design methods. Of the 21 articles that used longitudinal design methods, only 6 articles applied multistate models that addressed the cyclical nature of engagement in HIV care. In empirical analyses, the proportion of people living with HIV (PLHIV) initiated on antiretroviral therapy (ART) increased over time from 67% in 2004-2009 to 96% in 2019-2022. Among PLHIV initiated on ART and retained in care, 95% were virally suppressed by September 2022. However, in the longitudinal assessment, at 12 months, 24% were lost to follow-up (LTFU). After being LTFU, 34% returned to care within 12 months. Furthermore, being on ART at baseline was associated with a reduced hazard of LTFU (HR=0.71; 95% CI: 0.67-0.75) and an increased hazard of return to care after being LTFU (HR=1.38; 95% CI: 1.31-1.45). Analyses focused on the test-and-treat period showed that in the first 12 months, 41% of the PLHIV achieved suppressed VL. However, after achieving suppressed VL, the proportion of PLHIV who remained in care with suppressed VL and without other events decreased over time from 73% at 6 months to 2% at 60 months. Cross-sectional analyses showed a high proportion of VL suppression among PLHIV retained in care, with the UNAIDS VL suppression target being achieved. However, longitudinal viii | P a g e models revealed that many PLHIV transitioned in-and-out of care, and the first 12 months after enrolment and LTFU are important time windows for maximizing retention. The use of longitudinal design methods is growing. Longitudinal HIV continuum of care methodological guidelines will increase the uptake of the advanced methods in different assessments