The development and utility of a clinical algorithm to predict early HIV-1 infection

The association between self-reported clinical factors and recent HIV-1 seroconversion was evaluated in a prospective cohort of 4652 high-risk participants in the HIV Network for Prevention Trials (HIVNET) Vaccine Preparedness Study. Eighty-six individuals seroconverted, with an overall annual seroconversion rate of 1.3 per 100 person-years. Four self-reported clinical factors were significantly associated with HIV-1 seroconversion in multivariate analyses: recent history of chlamydia infection or gonorrhea, recent fever or night sweats, belief of recent HIV exposure, and recent illness lasting > or =3 days. Two scoring systems, based on the presence of either 4 or 11 clinical factors, were developed. Sensitivity ranged from 2.3% (with a positive predictive value of 12.5%) to 72.1% (with a positive predictive value of 1%). Seroconversion rates were directly associated with the number of these clinical factors. The use of scoring systems comprised of clinical factors may aid in detecting early and acute HIV-1 infection in vaccine and microbicide trials. Organizers can educate high-risk trial participants to return for testing during interim visits if they develop these clinical factors. Studying individuals during early and acute HIV-1 infection would allow scientists to investigate the impact of the intervention being studied on early transmission or pathogenesis of HIV-1 infection.