Israel's ban on use of Ethiopians' blood: how many infectious donations were prevented?

The false-negative rate depends largely on the window period between HIV-1-infection and the development of antibodies, and the incidence of HIV-1 infection. 7–9 Second, the HIV-1-risk arguments were relative rather than absolute. The question is not whether the prevalence of infection among Ethiopian immigrants is 50 times greater than among other Israelis, but whether the incremental number of infectious donations prevented by exclusion of Ethiopian donors is sufficiently large to justify such a ban. This latter quantity depends on the number of discarded Ethiopian donations in addition to the per-donation risk of contamination. This paper estimates the number of infectious donations to Israel's blood supply originating from non-Ethiopian Israeli donors, as well as the number of infectious donations prevented by the ban on Ethiopian blood. In Israel, all donated blood has been screened for HIV-1 antibody since 1986. 10 The annual number of infectious donations emanating from a particular donor pool thus equals the product of the annual number of antibody-negative donations and the fraction of these donations that are actually infected (the false-reassurance rate). The false-reassurance rate is the product of the HIV-1-incidence rate among blood donors and the duration of the window period, currently estimated at 25 days (95% CI 9–41). 7,8 For both non-Ethiopian and Ethiopian Israelis, I assume that HIV-1 incidence among potential blood donors is equivalent to that in the population aged 15 years or greater, although the infection rate among potential donors is lower than in the general population owing to self-selection, and donors can also be excluded for reasons other than positive HIV-1 antibody. 8 Assuming that all infections occur in the population aged 15 or above further inflates the presumed incidence rate among donors. Aggregate HIV-1 incidence among adult non-Ethiopian Israelis was estimated by backcalculation. 6 From 1990 to 1993, an estimated 61 (SD 12) new infections occurred annually in this group. Of the 3·9 million non-Ethiopian Israelis aged 15 years or more, 11 an estimated 960 were infected. 3,6 Division yields a presumed incidence of 1·6 per 100 000 uninfected people per year for non-Ethiopian Israeli blood donors. Multiplication by 25 days gives a false-reassurance rate of 1·1 per million antibody-negative donations. By comparison, in the USA, it has been estimated that 2·3 per million antibody-negative donations fall in the window. 7 Virtually all HIV-1 infections among Ethiopian immigrants have occurred among adults who immigrated to Israel since 1991. 3,12 The …