OBJECTIVES : Test & Treat (T&T) policy (WHO) recommends universal cART initiation regardless of CD4 T cell count. We modelled the impact of RI and T&T on the French HIV epidemiology, accounting for faster treatment initiation and virological control as well as reinforced treatment adherence and retention to care.

METHODS : We developed a Markov transmission model that simulates the HIV-infected population at a cohort level. Different subgroups of interest have been defined: MSM, heterosexual French and non-French, PWID. Each year, infectious patients with detectable viral loads transmit HIV a number of new subjects based on a “propagation factor” calibrated on the number of new infections every year, thus reflecting the number of at risk acts per infectious patient, condom use and efficacy and HIV status of partners.

Newly infected and undiagnosed patients then transit between states based on the time interval in the cascade of HIV care observed in France. Mortality is applied each year by sub-group.

Three scenarios based on the RAPID study results were tested: S1 accounts for RI of ART (day 1 vs. 33), S2 accounts for RI of ART and impact on lost-to-follow-up rates (RR=0.69, Rapid Start Study), S3 adds optimized testing and diagnosis of the hidden epidemic (90% by 2020) to these two effects.

RESULTS : The current situation (CS) results in a stable incidence of HIV infections between 5,655 and 5,684 new cases annually (years 2018 and 2030, respectively). S1 led to 5,573 new infections in 2030 (-111 [-2.0%] vs. CS). S2 led to 5,470 new infections in 2030 (-214 [-3.8%] vs. CS). S3 led to 3,908 new infections in 2030 (-1,776 [-31.2%] vs. CS).

CONCLUSIONS : By reducing the time during which HIV-positive patients are infectious and improving treatment initiation and adherence rates, rapid initiation of ART could prevent over 31% of new HIV infections in France by 2030.