3232 Translational Science 2019

OBJECTIVES/SPECIFIC AIMS: We hypothesize that VL testing varies by geographic sub-region, country, age, gender, mode of transmission, year of diagnosis, and country of origin; and also that a higher prevalence of VL testing may be associated with higher prevalence of population-level VL suppression. Our primary aim is to determine country- and regional-level factors that are associated with viral load testing amongst HIV patients. Our secondary aim is to explore the association between prevalence of viral load testing and viral load suppression at the population level. METHODS/STUDY POPULATION: This is a retrospective analysis of de-identified individual-level data reported to the European Surveillance System (TESSy). The TESSy is a database of communicable diseases (including HIV) for the ECDC and WHO European Regional Office. It captures data from 31 European Union/European Economic Area (EU/EEA) countries and 23 non-EU/EEA countries. Stored data is from year 2000. TESSy is used for data analysis and production of outputs for public health action. The patient cohort include adults older 18 years, whose last clinic attendance was reported in 2014 or later, or whose viral load test was reported in the year of the visit or the year before the year of their last reported clinic attendance. Patient demographic data include age, sex, mode of transmission, country of origin (migrants), country of diagnosis, geographic region, last clinic attendance, viral load and therapy status. Geographic region will be categorized into East, West and Centre as per WHO guidelines. Countries will be categorized and analyzed according to their European Union (EU)-, European Economic Area (EEA)- and income (GDP)-status, using current World Bank and International Monetary Fund (IMF) guidelines. All statistical analysis will be performed in R-Studio and R i386 3.0.2. Missing data will be characterized in terms of quantity (how much is missing) and pattern (random versus non-random) and impact on covariates to be tested. Multiple data imputations would be used in cases where missing data is found to be at random. Data from external sources like UNAIDS, World Bank and IMF will also be used for comparison and validation of TESSy data for imputation of missing data. Continuous variables will be analyzed through appropriate parametric and non-parametric tests while categorical variables will be analyzed through methods of proportion. Multivariate logistic regression methods will be used to explore the associations between VL testing and VL suppression separately with age, sex, year of diagnosis, country of origin (migrants), mode of transmission, in the total population, then at country- and regional-level. The same associations will be explored using a country’s EU and EEA status (EU versus EEA versus non-EU/EEA), and income status (high versus upper middle versus lower middle versus low). DISCUSSION/SIGNIFICANCE OF IMPACT: Even though this is a retrospective analysis of a database with likely significant missing data that may affect analysis of data and interpretation of results, our study will impact all levels of HIV policy across Europe. The strengths of this study likely outweigh the limitation imposed by missing data and include potential regional-, country- and demographic-specific public health, epidemiologic and ART program policy initiatives. Also our analysis of pattern of missing data may inform a more efficient and meaningful data collection and input into TESSy database.