Residual risk of transmission of HIV through blood transfusion in South Africa.

Abstract

Despite the ongoing review of donor recruitment criteria by local blood transfusion services and the development of highly sensitive and specific testing for the presence of antibodies to HIV in blood and blood products, there remains a residue of HIV in donated blood. This is because of donors who are in the 'window period' between acquisition of HIV and seroconversion, human errors and limits to the sensitivity and specificity of current tests. Data available from a national survey of HIV seroprevalence in South African blood donors allowed for the estimation of the number of units screened negative but likely to be infected with HIV. Assuming window periods of 4.8 and 14 weeks, a test sensitivity of 99.9%, a specificity of 98.5% and a human error rate of 0.1%, the likely rate of HIV-infected blood in the South African blood transfusion supply ranges from 1.1 to 3.9/100,000 units, with a likely estimate of 2.2/100,000 units. In the current South African blood transfusion setting, between 8.1 and 28.2 units of blood per annum will be HIV-positive with a likely estimate of 15.9 units. This corresponds to an odds ratio of between 1:90 909 and 1:25 641 units infected with HIV. These data are comparable with the risk in developed countries. The expected increase in the incidence and prevalence of HIV infection in all adult South African populations necessitates additional measures to ensure a blood supply which is as safe as possible. Some of these measures have already been taken by local blood transfusion services.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 15 citations exist upstream.