Finding the recently-infected to enhance medical care and prevention of transmission: A network approach and study in Athens, Greece

Background: Case finding and treatment as prevention should find additional recently-HIV-infected people. Implications for prevention remain unclear.

Methods: HIV+ “seeds” were referred from an Athenian study of people who inject drugs (PWID). Participants were considered recently infected if they had documented HIV negativity within the prior 6 months or if HIV-1 Limiting Antigen Avidity EIA (LAg) ODn ≤ 1.5 with no evidence of either late-stage infection or antiretroviral treatment. Other HIV+ participants were classified chronically infected. Seeds were interviewed to elicit and recruit network members who injected drugs or had sex with either seeds or others attending the same social spaces. Statistics use Poisson distribution for yield rates and binomial distribution for proportions.

Results: 162 people were recruited from 23 recently-infected seeds’ networks and social spaces (N&SS); 64 from N&SS of 18 chronically infected seeds. We located 11 recently-infected people in recently-infected seeds’ N&SS; yield rate and proportions were 0.478 (95% CI 0.239, 0.856) per seed and 0.068 (CI 0.034, 0.118) per N&SS member respectively. One recently-infected person was found in N&SS of chronically infected seeds: 0.056/seed (CI 0.001, 0.310) and 0.016/N&SS member (CI 0.001, 0.084). Rate and odds ratios in recents’ N&SS to chronically infecteds’ N&SS were 8.61 (p=0.039) and 4.58 (p=0.114), respectively.

Conclusions: Tracing N&SS of recently-infected PWID locates recently-infected people more efficiently than tracing N&SS of longer-term infected or, by extension, of all HIV-positives. The higher density of recently-infected in N&SS of recently-infected supports the hypothesis that the recently-infected account for an enhanced proportion of new infections.