Do demographics alone predict HIV testing behavior among rural malawians?

Background: Many individuals in developing countries, including Malawi, are not aware of their HIV status. HIV testing is a pre-requisite to knowledge of HIV status, accessing HIV treatment services, and preventing HIV transmission. We evaluated CARE Malawi's SAFE food security program to examine self-reported HIV testing rates at baseline and 18 month follow-up for 870 randomly selected individuals (582 SAFE, 288 controls) living in 6 traditional authorities in rural Malawi. Methods: Mixed methods logistic regression with random effects was used to identify predictors of HIV testing. Results: HIV testing rates increased in both SAFE group and controls. HIV testing rates increased from 52% at baseline to 70% at follow-up in the SAFE group, and from 66% to 70% in controls. Testers were three times (p =.01) more likely to be women compared to men, 28% (p <.001) more likely to live in large households, and three times (p <.001) more likely to have a secondary education. Perceptions of current HIV risk and confidentiality in testing were each associated with 1% (p <.001) likelihood of testing, while future HIV risk with 2% (p =.01) less likelihood of testing. Surprisingly, Village Health Committee (VHC) was associated with 50% (p =.02) less likelihood of testing. Conclusions: HIV testing in rural Malawi has increased. HIV risk perceptions and confidentiality are important predictors of HIV testing behavior. Because of the significant relationship between confidentiality and HIV testing, qualitative research may shed light on how to increase HIV testing rates in rural communities in Malawi.