CEPHIA publishes a major paper in the journal AIDS, evaluating the leading candidate assays intended to discern ‘recently’ from ‘non-recently’ acquired HIV infection. 

The key finding of this work is that, in stand alone form, none of the evaluated assays are suitable  for general surveillance applications, as all are confounded by viral suppression (either through 'elite control' or anti retroviral therapy) which lead to 'recent' test results for specimens from long infected individuals. As noted in previous literature, test performance, and optimisation, is a complex context dependent issue, and involves trade offs to obtain a sufficiently long 'mean duration of recent infection' along with a sufficiently small 'false recent rate'. What is optimal in one context (defined largely by incidence and prevalence, but also other factors like populaiton level viral suppression) is not necessarily optimal in another.

CEPHIA will continue to investigate these assays, in particular by evaluating the use of a concurrently administered viral load test to reduce confounding. There is also ongoing discussion of improved 'target product profiles' which are being develoepd to support evaluation of candidate products.

Lead author on the paper, Reshma Kassanjee, is based at the South African Centre for Epidemiological Modelling and Analysis (SACEMA) which has recently produced a policy brief to provide and update of ‘recent infection’ testing as it applies to surveillance uses, which have until recently been driving the development of recent infection tests.


Read the article in AIDS

Read the SACEMA policy brief