Human immunodeficiency virus-1 and hepatitis C virus RNA among South African blood donors: estimation of residual transfusion risk and yield of nucleic acid testing

Abstract

Background and objectives: South Africa is an endemic area for human immunodeficiency virus 1 (HIV-1) infection, which has an impact on the safety of the blood supply. We studied the presence of HIV-1 and hepatitis C virus (HCV) RNA, and recent HIV seroconversion, in blood donors in order to estimate transfusion risk and to determine whether nucleic acid testing (NAT) could effectively improve blood safety.

Materials and methods: Unlinked samples collected in 1999 from 9077 HIV-low prevalence (LP) and 10,632 HIV-high prevalence (HP) donors were studied. Donor demographic information and serology results were collected prior to breaking the linkage. All samples were individually tested using a multiplex NAT assay for HIV-1 and HCV RNA. HIV antibody-positive samples were further tested using a 'detuned' (less sensitive) enzyme immunoassay (EIA) procedure to determine whether a donor had recently acquired infection. Data were used to estimate the residual transfusion risk and to project NAT yield.

Results: HIV was 45 times more prevalent in the HP- than in the LP donor group; and among the HP group, female donors had a significantly higher prevalence of HIV than male donors. All seven HIV-1 p24 antigen-positive samples in the study were also HIV NAT positive. Two HIV NAT-positive samples were anti-HIV negative; both of these samples were from HP donors. Assuming that 10% of the 900,000 annual donations in South Africa are from the HP group, we projected an annual NAT yield of 8.5 cases over the current screening of antibody and p24 antigen. However, if p24 antigen testing were to be eliminated, this number would be increased to 17 cases per year. Based on 'detuned' EIA results, the incidence rate for HIV infection was estimated at 1.29 and 51.12 per 10,000 per year for the LP and HP donor groups, respectively. Assuming a 15-day earlier detection by HIV NAT compared with antibody tests, these incidence rates project that NAT may intercept an additional 23 (95% confidence interval: 15-33) HIV-positive donations per year. For HCV, two viral RNA and antibody-positive samples (one from the LP group and one from the HP group) and no NAT yield cases were found in the study.

Conclusions: Implementation of routine NAT blood screening would allow elimination of HIV-1 p24 antigen testing and improve the safety of the blood supply in South Africa. However, the cost-benefit ratio of introducing such an expensive technology in a country with a limited health budget will have to be carefully considered.