Abstract
Population-based disease prevalence surveys raise ethical questions, including whether participants should be routinely told their test results. Ethical guidelines call for informing survey participants of any clinically relevant finding to enable appropriate management. However, in anonymous surveys of human immunodeficiency virus (HIV) infection, participants can "opt out" of being given their test results or are offered the chance to undergo voluntary HIV testing in local counselling and testing services. This is aimed at minimizing survey participation bias. Those who opt out of being given their HIV test results and who do not seek their results miss the opportunity to receive life-saving antiretroviral therapy. The justification for HIV surveys without routine feedback of results to participants is based on a public health utility argument: that the benefits of more rigorous survey methods - reduced participation bias - outweigh the benefits to individuals of knowing their HIV status. However, people with HIV infection have a strong immediate interest in knowing their HIV status. In consideration of the ethical value of showing respect for people and thereby alleviating suffering, an argument based on public health utility is not an appropriate justification. In anonymous HIV surveys as well as other prevalence surveys of treatable conditions in any setting, participation should be on the basis of routine individual feedback of results as an integral part of fully informed participation. Ensuring that surveys are ethically sound may stimulate participation, increase a broader uptake of HIV testing and reduce stigmatization of people who are HIV-positive.