PEPFAR and adult mortality: a replication study of HIV development assistance effects in Sub-Saharan African countries
3ie Replication paper 15, 2018
In this paper, Nicholas Hein and colleagues conducted a replication of the 2012 publication HIV development assistance and adult mortality in Africa by Eran Bendavid and colleagues. The original study investigated the relationship between increased funding to sub-Saharan African countries through the President’s Emergency Plan for AIDS Relief (PEPFAR) and adult mortality. It concluded that during the years of PEPFAR implementation (2004-2008), the odds of adult mortality was significantly lower in PEPFAR sub-Saharan countries as compared to non-PEPFAR sub-Saharan countries.
A few minor discrepancies did not change the original authors’ conclusions of the study. The replication study reproduced results that supported the original authors’ published conclusions. Additionally, the team replicated both individual- and country-level modeling adjustments, to a 2010 study by Duber and colleagues that did not find a statistically significant effect of the PEPFAR program on all-cause mortality in Sub-Saharan countries between 2000 and 2006. They found that PEPFAR had a significant impact on all-cause adult mortality when examining the effects of PEPFAR using an unadjusted model, as well as after adjustment with country-level covariates
However, modeling adjustments with both country and individual level covariates resulted in borderline insignificant findings of association between PEPFAR funding and all-cause adult mortality. Therefore, the replication researchers cautiously support the findings by the original authors.