) HIV remains a massive public health challenge in Latin America, with nearly 1.8 million adults living with HIV in the region's low- and middle-income settings. In order to identify key risk groups and deploy relevant interventions to end the HIV epidemic, researchers and public health organizations need high quality data regarding patient retention in clinical care, antiretroviral therapy (ART) use, and mortality estimates. Clinical cohorts are an excellent data source for measuring the quality and frequency of care, but can provide only imperfect measurements of sustained engagement and various clinical outcomes because of losses to follow- up (LTFU). Individuals lost to follow-up at particular clinical cohort sites appear to have ?disappeared? from the point of the data, but may in fact experience silent transfers (wherein patients switch clinics without notice), true gaps in care (wherein patients have truly discontinued medical care), or an unobserved fatal event. Groups like the World Health Organization (WHO) currently use post hoc analytic approaches to correct for LTFU, but no correction factors are available for Latin America. As a result, the WHO is currently applying estimates from sub-Saharan Africa to Latin America, where the populations may be very different. We propose to measure clinical and vital outcomes in the lost population through linkages between clinical data sources and well-established state/provincial or national registries. This tracing strategy of registry linkage has the benefit of being relatively low-cost and low-effort for clinical sites that may already be resource- constrained. The resulting revised estimates of LTFU could improve engagement outcome measurements, reduce potential selection bias and misclassification errors in research from these settings, and inform both researchers and public health stakeholders. The Caribbean, Central and South America network for HIV epidemiology (CCASAnet) is Latin America's largest observational HIV cohort and provides a rich data source in which to conduct this work. Several participating cohort sites are located in countries with robust public health infrastructure, including reliable pharmacy, laboratory, and vital status registries or surveillance systems. In fact, several of these sites are currently engaged in registry linkage activities with their respective health ministries. CCASAnet also has extensive experience conducting collaborative research projects addressing engagement in care and risk- population-defining correlates.
Aim 1 of this study will improve patient mortality and outcome ascertainment through linkage to pharmacy, laboratory, and vital status registries at CCASAnet sites with available regional and national registries.
Aim 2 of this study will be to replicate CCASAnet regional analyses using revised patient mortality and LTFU data and report findings along with optimal registry search strategies for Brazil, Mexico, and Peru.
) Patient care data from HIV clinics are our strongest sources of clinical research data in Latin America, but many people drop out of care so their health outcomes are unknown; these missing outcomes can lead to potential biases in how we interpret research studies, how we implement clinical solutions, and how we provide health care. Since no large-scale patient tracing studies have been done in Latin America, researchers and groups like the World Health Organization are currently applying estimates from Sub-Saharan Africa to Latin America, although these populations may be very different. We will fill this knowledge gap using a low-cost registry linkage technique, taking all lost to follow-up patients from Brazil, Mexico, and Peru that are included in Latin America's largest observational HIV research cohort (CCASAnet) and conducting a thorough search for them in the countries' national and regional death registries, prescription databases, and laboratory systems to determine their health outcomes and produce Latin America-specific findings that improve the quality of HIV research in the region.
