Background:  In France, the HIV National Reference Centre (NRC) has developed an immunoassay that discriminates infections lasting ≤ 180 days from those > 180 days. The use of this assay of recent infection on blood donations found HIV-1⁺ between 1992 and 2003 allowed to analyze the characteristics of recently infected donors, to estimate HIV incidence in blood donors and to validate this assay in the repeat blood donor population.

Method:  One third of HIV-1⁺ blood donations between 1992 and 1999 and all HIV-1⁺ donations since 2000 were sent to the NRC for serotyping. The assay for recent infection was retrospectively applied to all these donations. Results were analyzed according to donor status (repeat vs first-time), interdonation interval, sex, age, geographic origin, mode of transmission and HIV subtype. Incidence rates were calculated for first-time donors (recent infections/donations, 180/365) and for repeat donors (recent infections/donations, 180/365; mean interdonation interval/365). They were compared to classical estimates among repeat donors (seroconversions/person-years).

Results:  Between 1992 and 2003, 26.0% (95% CI 21.5 to 30.8) of the 366 HIV-1⁺ donors were identified as recently infected. This proportion was higher in repeat donors (41.4%) than in first-time donors (12.7%) (p < 10^–4) and for non-B subtypes (32.4%) than for B subtype (20.0%) (p = 0.03). The proportion of recent infections decreased when the interval between the positive and the last negative donation increased; from 91% for intervals ≤ 180 days to 6% for intervals >36 months. Trend analysis showed a significant increase in recent infections among homosexual donors, from 10% in 1992 to 1994 to 44% in 2001 to 2003 (p = 0.009). Incidence decreased from 4.3 per 10⁵ (95% CI 1.9 to 9.4) in 1992 to 1994 to 1.6 (0.8 to 3.3) in 2001 to 2003 (p = 0.004) in first-time donors and from 3.2 (95% CI 2.0 to 5.0) to 1.1 per 10⁵ (0.7 to 1.8) (p = 0.003) in repeat donors. These rates were similar to those derived from the classical method in repeat donors.

Conclusions:  The consistency of the results of the assay for recent infection with the interval between the positive and the last negative donation on one hand, and the similarity of incidence rates with the classical method on the other hand, validate the relevance of the immunoassay for recent HIV infection. This new assay allowed us to estimate the HIV-1 incidence in first time donors showing that incidence is slightly higher in this population than in repeat donors and that it has decreased overtime in both populations.

Keywords: HIV incidence; Blood donors; Immunoassay for recent infection