Public health practitioners frequently must make programmatic decisions with data that are less than perfect. The main reason to use such data is because few well-conducted, high-quality studies exist that can answer an everyday public health question directly, and often we rely instead on imperfect surveillance data to guide our actions. This lack of data has been of concern for the human immunodeficiency virus (HIV) epidemic in Rwanda, where the scale-up of international resources has required rapid responses in planning and targeting prevention and treatment programs. In recent years in Rwanda, there has been a proliferation of HIV surveillance data, including sentinel surveillance in pregnant women, demographic and health surveys, and behavioral surveillance in youth, truck drivers, and sex workers. There also has been an increase in the number of quantitative and qualitative research studies, as well as an accumulation of local and national expertise. Despite this increase in available information, no method for synthesizing the data and no consistent way to analyze or interpret diverse data has been available. In the absence of time and resources to conduct the perfect study to guide programmatic and policy decision making, the important question of how can we use existing data to achieve similar purposes remains.
This study was an HIV data triangulation exercise to answer high priority questions using existing data sources in Rwanda. This exercise, funded by the Presidential Emergency Fund for AIDS Relief (PEPFAR), aimed to better define at-risk populations by geography and risk behavior and to determine the reach and intensity of corresponding programmatic responses.