Abstract
Epidemiologic studies involving HIV (human immunodeficiency virus) antibody testing create ethical dilemmas, particularly about notifying asymptomatic seropositive subjects. Four study designs address this problem: mandatory notification, optional notification, anonymous testing, and blind testing. No single design consistently optimizes the trade-off between valid and ethical research. Each strategy differs substantially from the others in its effect on response rates, bias, ability to perform longitudinal studies, numbers of subjects who learn their test results, and the number of subjects counseled about HIV risk reduction. Both local institutional review boards and potential subjects of study (and their sexual partners) should participate in decisions regarding the conduct of sensitive AIDS (acquired immunodeficiency syndrome) research.