Streamlining Global and Local Data on HIV: Underscoring Role of Institutions and Ethics in Improving Quality of HIV Research

There are inconsistencies in the South Africa HIV mortality data reported by Institute of Health Metrics and Evaluation (IHME), Joint United Nations Programme on HIV/AIDS (UNAIDS), and Statistics South Africa (StatsSA) platforms. Between 2006 and 2016, these global data sets (IHME and UNAIDS) show that HIV-related mortalities were improving in South Africa, whereas StatsSA argues the opposite. We explain the causes of this differing stands and highlight areas that may be improved to address such inconsistencies. This observational analysis uses data from IHME, UNAIDS, and StatsSA platforms. We demonstrate that IHME and UNAIDS data sets are based on a mathematical compartmental model, which is not dynamic to all HIV epidemiological aspects. Such limitation may cause inflated improvement in HIV mortality outcomes that are not in line with HIV mortality evidence recorded at the household level as demonstrated by StatsSA. There is a need to streamline the IHME, UNAIDS, and StatsSA data on HIV to improve the quality of HIV research and programming in South Africa. •Global HIV mathematical models are inconsistent in predicting HIV-related mortalities.•These global HIV mathematical models show inflated HIV mortalities and contradict local data sets.•Streamlining global HIV mathematical models with local data will improve the quality of HIV research.