The construction of self-assessed HIV/AIDS infection likelihood: an examination from rural Malawi

Philip A. Anglewicz, University of Pennsylvania
Hans-Peter Kohler, University of Pennsylvania

How accurately do rural Malawians assess their likelihood of HIV infection? As elsewhere in sub-Saharan Africa, few people are tested for HIV in rural Malawi, so they assess their current status in other ways. However, these self-assessments can be inaccurate. To investigate the accuracy of self-assessed HIV infection, we use longitudinal survey data from rural Malawi (from 2001 and 2004), which cover approximately 4000 respondents and include HIV biomarkers, collected in 2004. In this analysis, we first identify factors that influence the subjective probability of HIV infection for men and women. To measure the accuracy of these self-assessments, they are compared to objective HIV status. Respondents who correctly and inaccurately assess their probability of infection are compared, and we then identify the determinants of accuracy in the self-assessed probability of HIV infection. Finally, the implications of these assessments on sexual behaviour and demographic aspects of rural Malawi are discussed.