HIV / STI
Assessing Causal Effects of HIV Testing on HIV Incidence in Rural KwaZulu-Natal, South Africa Ke Zhang* Ke Zhang Zhang Zhang Zhang Zhang Zhang kzhang@uri.edu
Although there are effective strategies to control the HIV epidemic, it remains a significant individual and public health challenge in South Africa. HIV testing is the gateway to HIV treatment in those who have acquired HIV and HIV prevention in those who tested HIV negative. Existing studies have suggested that HIV testing has a significant effect in reducing HIV incidence. However, these studies have not fully assessed spillover effects, the effects of one’s HIV testing on HIV incidence among unexposed others. Assessing spillover can provide a more complete understanding of the impact of HIV testing. The data we used is from ANRS 12249 treatment as prevention (TasP) trial, conducted in a rural region of South Africa from March 2012 to July 2016. We grouped participants by homesteads and assumed partial interference limited to the homestead, estimated both the direct (i.e., the intervention effect under exposure versus no exposure while holding other factors constant) and spillover effects of altering the proportion of HIV testing in the homestead on subsequent HIV incidence. Estimation was carried out with a marginal structured model fit with inverse probability weights. The results suggested that, on average, (1) Risk of HIV infection decreased by 42.3% (95% confidence interval: [19.4%, 58.8%] ) under HIV testing versus no testing, with HIV testing for others in the homestead remained constant (i.e., direct effect); (2) No statistically significant spillover effect detected comparing two different proportions of exposed homestead members. Further research is needed to understand the underlying mechanisms.