Temporal trend and characteristics of recent HIV-1 infections: application of an algorithm for the identification of recently acquired HIV-1 infections among newly diagnosed individuals over a 10-year period

Identification of recent infections (RI) may contribute to improve the quality of human immunodeficiency virus (HIV) surveillance, monitoring ongoing transmission and planning and evaluating prevention programs. Our study applied an algorithm combining clinical and serological information to identify RI in individuals newly diagnosed with HIV in Rome, during the years 1999-2008, in order to describe the trend and characteristics of recently infected individuals. RI were documented seroconverters, or people with an HIV avidity index (AI)<0.80. Individuals with advanced infection (CD4 count <200 cells/µL or AIDS-defining illness) or with AI ≥0.80 were considered long-standing infections. Overall, we observed 2,563 new HIV diagnoses. The algorithm was applied in 2124/2563 (82.9%). Of these, 355 were RI (16.7%). RI was found independently associated with calendar year (adjusted odds ratio [aOR]= 1.06, 95% confidence intervals [CI]=[CI 1.02-1.11], for every year of increase), HIV-risk category (men having sex with men: aOR=1.44, [CI 1.04-1.98]; injecting drug users: aOR=1.58, [CI 1.03-2.42] vs. heterosexuals), country of origin (foreign-born: vs Italians: aOR=0.46, [CI 0.33-0.62]), and recruitment site (inpatient vs outpatient clinic: aOR=0.49, [CI 0.37-0.66]). By the application of our algorithm we could characterize the pattern of ongoing HIV transmission, identifying groups needing more urgent prevention programs. SUMMARY