Feasibility of Measuring HIV-related Mortality During Population-based Surveys in Africa

Associate Name: Stephane Helleringer

Funding Source/Period of the Grant: NICHD R01 06/23/16-05/31/18

Description: 

Every year, large amounts of financial and human resources are invested in addressing HIV epidemics worldwide. Reducing HIV/AIDS deaths is one of the clearest indicators of success in these HIV prevention and treatment programs. It is thus crucial to collect the best possible data on HIV/AIDS mortality to 1) monitor progress towards this target and 2) evaluate the impact of HIV programs. Unfortunately, the vast majority of countries most affected by HIV epidemics, in particular those located in Sub-Saharan Africa (SSA), often lack well-functioning civil registration and disease surveillance systems (CRS). Instead, reliable data on causes of death in SSA are currently only available for small and highly selected populations. We propose to address this lack of representative data by assessing the feasibility of collecting information on HIV mortality during household surveys such as the Demographic and Health Surveys (DHS). These data constitute the primary input used by international organizations to estimate all-cause mortality in SSA countries with limited CRS. They measure mortality by asking a random sample of respondents to report the date of birth and (possibly) death of each of their close relatives (e.g., siblings). They frequently also include an assessment of other causes of death such as maternal mortality and accidents, but they do not currently collect data on HIV mortality. We will thus enhance the adult mortality questionnaire used by the DHS with 3-5 direct questions on i) the HIV status and ii) the engagement in HIV care and treatment of each reported deceased relative. We will develop this enhanced questionnaire through a participatory process of qualitative research (aim 1). We will then measure the accuracy of this questionnaire in recording HIV deaths during a retrospective validation study against reference data on HIV mortality from a health and demographic surveillance system in Malawi (aims 2 & 3). A nested randomized sub-study will also determine whether the use of audio computer-assisted self- interviewing (ACASI) can help further improve the reporting of HIV deaths during mortality surveys. If accurate, this enhanced questionnaire will yield direct estimates of the excess mortality faced by persons living with HIV in a country. This promises to significantly strengthen the evaluation of HIV care and treatment programs in SSA, and to improve our understanding of the HIV care cascade (From NIH Report).