Abstract
OBJECTIVE: To examine the potential impact of deferral of blood donors at high risk of HIV infection in a west African city where blood is screened for HIV antibodies but no other special measures are taken to protect the blood supply. DESIGN: Cross sectional study. SETTING: National Blood Transfusion Centre and Project RETRO-CI, an international collaborative AIDS research project, Abidjan, Côte d'Ivoire. SUBJECTS: 1257 male first time blood donors. INTERVENTIONS: Blood donors were interviewed about demographic and behavioural characteristics and tested for HIV antibodies by enzyme immunoassay and, if positive, synthetic peptide based tests. MAIN OUTCOME MEASURES: HIV antibody status in relation to presence of behavioural risk factors; calculation of sensitivity, specificity, and predictive values of specific criteria for excluding HIV infected donors. RESULTS: The overall prevalence of HIV infection was 11.4%. The most important risk factors for HIV positivity were prostitute contact and being aged 30-39 years. For identifying seropositive donors individual criteria had sensitivity, specificity, and positive predictive values ranging from 15% to 98%, 38% to 91%, and 17% to 30% respectively. Prostitute contact in the past five years would have excluded 31% of all donors and 73% of HIV infected donors. 27% of those excluded would have been HIV positive. CONCLUSIONS: The widespread assumption that donor deferral is not feasible in sub-Saharan Africa needs reassessment. In Abidjan this approach was well accepted and potentially effective. Donor deferral requires evaluation as a strategy for improving blood safety in resource poor areas with high rates of HIV infection.