How to assess HIV prevention trials—and design better ones
When a landmark randomized trial in Botswana aimed at preventing HIV via test-and-treat interventions showed variable reductions in new infections in trial communities, experts wondered why. Now, a new study led by researchers at Harvard T.H. Chan School of Public Health has shown that new infections were stemming largely from communities outside the trial area—and that if universal test-and-treat interventions were implemented nationwide, HIV transmission would be significantly lower.
The study was published March 14 in Nature Communications. The co-corresponding authors included Lerato Magosi, department associate in the Department of Epidemiology; Shahin Lockman, associate professor in the Department of Immunology and Infectious Diseases; Max Essex, Mary Woodard Lasker Professor of Health Sciences, Emeritus; Victor De Gruttola, professor of biostatistics, emeritus; and Marc Lipsitch, adjunct professor of epidemiology.
The goal of the study was to aid in interpreting complex HIV trial results and inform the design of such studies in the future, as well as to inform public health policy decisions about effective HIV prevention strategies.
The researchers analyzed the impact of the Botswana Combination Prevention Project (BCPP), a 2013-2018 randomized trial that tested the effectiveness of an HIV prevention and control intervention in communities throughout Botswana. The BCPP found that in trial communities—those that received the test-and-treat intervention, which entailed expanded testing, strengthened linkages to care, and universal access to antiretroviral treatment—HIV transmission declined by 30%, even though the decline had been expected to be higher.
In the new study, the researchers sought to understand this lower-than-expected decline. They developed a statistical model that enabled them to estimate how HIV was being transmitted—whether new infections in trial communities were occurring from individuals belonging to the same or different communities, and whether these communities were trial communities, control communities (those not receiving the intervention), or non-trial communities (those outside the trial area).
The study found that 90% of new HIV infections in trial communities originated from individuals living in non-trial or control communities. The researchers estimated that if every community nationwide received the intervention, HIV transmission in trial communities would be 59% lower—a substantial improvement from the 30% the BCPP actually observed.
According to the researchers, the findings highlight what’s missing from how HIV prevention and control programs are monitored, analyzed, and used to inform public health policy. “Our findings elucidate the potential impact of a nationwide intervention and provide insight on the extent to which the BCPP intervention was diluted by spillover infections from control communities and from communities outside the trial area,” they wrote. “We recommend that the impact of sources of transmission beyond the reach of the intervention be considered when designing and evaluating interventions to inform public health programs.”
Read the study:
Unpacking sources of transmission in HIV prevention trials with deep-sequence pathogen data
