DREAM Big: Emerging Results from a PEPFAR Partnership to Reduce HIV Among Adolescent Girls and Young Women

Jessie Lu
June 14, 2018

You can read a companion post authored by our Population Council colleagues here.

On World AIDS Day 2014, PEPFAR announced the DREAMS partnership, a multisector program to reduce HIV incidence among highest-risk adolescent girls and young women. It’s been three years since the rollout of the program began, and earlier this month in collaboration with the Population Council, CGD convened key players to discuss emerging results, what they mean for the future of DREAMS, and how we can ensure that the next years of programming go even farther to deliver the most effective services to those most at risk.

DREAMS is hugely important in pushing the boundaries of HIV programming—it spans 15 countries, aims to reach the highest risk populations, and undertakes a comprehensive approach to monitoring and evaluation to feed back into program design and delivery, an innovation important in the HIV world and for global health in general. Preliminary data suggest that HIV conditions have markedly improved for adolescent girls and young women over the past three years. But in order to understand the impact that DREAMS has had on HIV incidence in this target population so far, there’s still more to learn.

DREAM big—new HIV diagnoses drop dramatically in most participating facilities

Ambassador Birx presented preliminary results from a census of facility data from participating clinics showing that new HIV diagnoses among pregnant women are dropping in most places, most of the time—in 2017, 65 percent of the highest-burden districts in the original 10 DREAMS countries saw at least a 25 percent decrease in new HIV diagnoses as compared with levels before implementation of the program in the same clinics. The remaining 35 percent of districts did not see changes at the same scale, but almost all DREAMS intervention districts saw declines in new HIV diagnoses. These facility data are highly encouraging, but more analysis is needed to understand whether the data also reflect a decline in new HIV infections, specifically among adolescent girls and young women, the target group for DREAMS.

What do these monitoring numbers say about impact so far?

On their own, facility data on new diagnoses do not allow us to determine whether DREAMS is responsible for the observed changes as there are many factors that may influence this dynamic. And in the case of interventions to address risk factors among adolescent girls and young women, we don’t know enough about what works to reduce new infections—another reason why multiple data and evaluation strategies are so central to the DREAMS approach.  

The impact evaluation and implementation science research is underway—at this stage, there is not much to report about changes in HIV incidence or changes in HIV-related risk factors, as follow-up rounds of data collection with DREAMS beneficiaries are underway. And it’s worth noting a couple of issues. First, the DREAMS interventions vary from place to place depending on the constellation of implementing partners and the local context and challenges that the communities each face. Second, DREAMS packages encompass a wide range of programs that target multiple different community members and multiple different aspects of a girl’s life. With these varied interventions, it will be difficult to tease out which interventions have greater impact on outcomes than others.

Research presented by Isolde Birdthistle, PI for the impact evaluation based at the London School of Hygiene and Tropical Medicine, illustrates some of these difficulties associated with connecting results to impact. Data collected in two informal settlements in Nairobi in 2017 showed 90.4 percent of girls aged 15–17 enrolled in DREAMS were aware of their status versus 54.7 percent of girls not enrolled. These data alone are promising, suggesting DREAMS can reach girls earlier than their usual entrée into the healthcare system—before rather than after their first pregnancy, and before HIV risk typically peaks. But these positive results could be attributed to a variety of other factors, ranging from the environment around girls in informal settlements in Nairobi to the specific DREAMS package in Kenya, which incorporates HIV testing into enrollment and is run by a single implementing partner. The patterns are less promising so far in KwaZulu-Natal, South Africa, where HIV testing is often not included in DREAMS enrollment. Here, among young women aged 18–22 years, those receiving DREAMS interventions know about their status at about the same rates as those not receiving DREAMS. In the South African site, there are more than 10 implementing partners.

Initial results are also telling us about reach and coverage—and how to adjust implementation

Initial data collection is already pointing to issues in program uptake and coverage, which provides useful feedback for improving current program implementation as well as future efforts. The impact of DREAMS is still not reaching its full potential in very high-risk populations. A 2017 survey of young women who sell sex in Zimbabwe revealed that more than one year into program implementation, use of services in the DREAMS key prevention package remains low. Less than 5 percent of women in the high-risk population had received education or cash transfers, received vocational or job training, participated in a savings and loans program, or were taking PrEP.

Implementation science findings will be essential in complementing these data to ensure that the next years of DREAMS programming continue to innovate and reach the most vulnerable adolescent girls and young women. For example, research presented by Sanyukta Mathur, director of the Population Council’s DREAMS Implementation Science portfolio, aims to understand the vulnerability profiles among out-of-school adolescent girls and young women in Kisumu, Kenya. They found that although DREAMS programming is reaching many adolescent girls and young adults, those married and out of school are not yet well represented in DREAMS programming. Implementation science work by the Population Council also found that adolescent girls and young women experience more sexual violence than ever anticipated, which suggests an increased programmatic effort is needed. Work led by Julie Pulerwitz, director of the HIV and AIDS program at the Population Council, also pointed to the importance of reaching male partners of adolescent girls and young women. Council researchers are using innovative strategies to identify male partners and garner insights on how to best link them to HIV services.

All this information will feed back to program implementation to ensure that DREAMS reaches and delivers well-tailored packages to the most vulnerable.

These results, though generally positive, are still only preliminary. The true effects require more time to be realized and captured through data. Before finalizing the lessons learned from DREAMS and expanding the DREAMS approach, much work remains to tease out directly observable cause and effect linkages, which could be difficult given the multisector, varied, and long-term nature of these programs. In this context, keeping an eye toward data utilization and translation to programming is essential. This partnership has been groundbreaking in its three years thus far. Let’s see where it goes next.

More information about the DREAMS event, including a recording and full presentations, are available here.

With thanks to PEPFAR (Amb. Deborah Birx); the Population Council (Miriam Temin, Julie Pulerwitz, Sanyukta Mathur); World Education (Patience Ndlovu); LSHTM (Isolde Birdthistle); Gilead Life Sciences (Jirair Ratevosian); and UNICEF (Chewe Luo).


CGD blog posts reflect the views of the authors, drawing on prior research and experience in their areas of expertise. CGD is a nonpartisan, independent organization and does not take institutional positions.

Image credit for social media/web: Social media image by Kaveh Sardari