Blood screening for human immunodeficiency virus: a new algorithm to reduce the false‐positive results

Abstract

The objectives to this study were to evaluate the performance of an anti-human immunodeficiency virus (HIV) blood screening test and propose a new screening algorithm for blood banks routinely using nucleic acid amplification testing (NAT) to reduce false-positive results. Most anti-HIV enzyme-linked immunosorbent assay (ELISA) results are false-positive because of the low prevalence of HIV infection and high sensitivity of the ELISAs. A total of 281 588 voluntary donations were collected and sera reactive on one or both anti-HIV ELISAs were confirmed by Western blot (WB) testing. All samples with nonreactive results for the two ELISAs underwent NAT. A confirmed HIV-1-positive result was defined by a reactive result on NAT or WB testing. Correlations between signal-to-cutoff ratios and the confirmed HIV-1 infection rate were analysed for each enzyme immunoassay and two-enzyme immunoassay combination. The positive predictive values (PPVs) of the current and proposed algorithms were calculated. Seventy-nine donations (13·9%) were positive on WB analysis and one donation negative for anti-HIV antibody was reactive on NAT and confirmed to be a window period donation on additional follow-up testing. The PPV of the 567 donations reactive on one or two ELISAs was 13·9%. However, using the new screening algorithm, 457 donations underwent NAT immediately instead of WB testing. Only 110 donations were tested with WB and the PPV was 71·8%. Screening for HIV is sensitive, specific and time saving for donors with this algorithm, which is suitable for HIV screening in low prevalence settings.