HIV and Rational risky behaviors: a systematic review of published empirical literature (1990-2013)

Risky health behaviors have the negative effect - negative externality - of the individual being able to spread the disease to others. They thus represent a threat for the society and a tragedy for public health. The objective of this study is to inquire into the nature, extent and strength of the evidence for such risky behaviors for HIV/AIDS from an economic perspective. We aim at investigating the concept of risk or prevalence-elasticity of health behaviors in the case of HIV. We did an exhaustive review of published articles in French and English indexed in the databases PubMed, ScienceDirect and Jstor between 1 January 1990 to 31 December 2013. We searched for publications empirically investigating the risk or prevalence-elasticity of behaviors in the case of HIV/AIDS and performed a bibliometric and descriptive analysis of the dataset. Of the 12,545 articles that were screened, 189 (1.5%) full-text publications studied the risk-elasticity of health behaviors that are related to HIV/AIDS. Of these 189 articles, 167 (88.4%) were quantitative studies that empirically estimated the risk-elasticity, and 22 (11.6%) were qualitative studies. We found that 55.7% of the quantitative studies included at least a correlation between HIV risk and health behaviors that supports the concept of risk or prevalence-elasticity. Moreover, we identified articles that address the reverse causality problem between HIV risk and health behaviors, by using indirect HIV risk measures, to demonstrate the existence of a responsiveness of risk/preventive behaviors to HIV risk. Finally, an in-depth analysis showed seven out of ten articles using an objective measure of risk for HIV/AIDS gave strong support to prevalence-elasticity. However, only one of the ten articles established a direct measure of prevalence-elasticity while appropriately dealing with the reverse causation problem between objective HIV risk and preventive/risk behaviors. These results stress out the need to carefully monitor programs of risk behaviors' surveillance in the context of HIV becoming chronic, especially in sub-Saharan Africa where large scale HIV treatment policies are being implemented. More evidence is needed on the strength of rational risky behaviors to maximize the public health and economic impact of large scale HIV treatment or preventive policies. With this purpose, epidemiological surveillance programs could be paired with specific behavioral surveillance programs to better inform policy makers.