Background and setting: The Botswana Combination Prevention Project demonstrated a 30% reduction in community HIV incidence through expanded HIV testing, enhanced linkage to care, and universal antiretroviral treatment and exceeded the Joint United Nations Programme on HIV/AIDS 90-90-90 targets. We report rates and characteristics of incident HIV infections.

Methods: The Botswana Combination Prevention Project was a community-randomized controlled trial conducted in 30 rural/periurban Botswana communities from 2013 to 2017. Home-based and mobile HIV-testing campaigns were conducted in 15 intervention communities, with 39% of participants testing at least twice. We assessed the HIV incidence rate [IR; number of new HIV infections per 100 person-years (py) at risk] among repeat testers and risk factors with a Cox proportional hazards regression model.

Results: During 27,517 py, 195 (women: 79%) of 18,597 became HIV-infected (0.71/100 py). Women had a higher IR (1.01/100 py; 95% confidence interval: 0.99 to 1.02) than men (0.34/100 py; 95% confidence interval: 0.33 to 0.35). The highest IRs were among women aged 16-24 years (1.87/100 py) and men aged 25-34 years (0.56/100 py). The lowest IRs were among those aged 35-64 years (women: 0.41/100 py; men: 0.20/100 py). The hazard of incident infection was the highest among women aged 16-24 years (hazard ratio = 7.05). Sex and age were significantly associated with incidence (both P < 0.0001).

Conclusions: Despite an overall reduction in HIV incidence and approaching the United Nations Programme on HIV/AIDS 95-95-95 targets, high HIV incidence was observed in adolescent girls and young women. These findings highlight the need for additional prevention services (pre-exposure prophylaxis and DREAMS) to achieve epidemic control in this subpopulation and increased efforts with men with undiagnosed HIV.