BACKGROUND AND PURPOSE
The Investment Framework (IF) for HIV prevention, treatment, and care, developed by Joint United Nations Programme on AIDS UNAIDS in 2011 calls for rapid scaling up of key HIV interventions that directly affect HIV transmission and mortality and support interventions that do not, themselves, directly affect transmission but support the scale-up of those that do. Previous modeling found that scaling up to universal coverage by 2015 and maintaining these targets beyond 2015 would avert a cumulative 12.2 million new infections and 7.4 million AIDS deaths by 2020. The cost of these investments peaks in 2015 at about $22 billion annually, then gradually declines, reflecting more efficient treatment and changes to more cost-effective approaches to prevention.1 This article examines the impact of the IF on adolescents who are 10–19 years old.
About one-fifth of all new HIV infections occur in the adolescent years. Moreover, adolescence is when most people first have sex, and many develop behaviors and undergo experiences that set them up for future vulnerability to infection. HIV/AIDS is also a major health problem in adolescence. In 2010, the disease ranked third as a cause of death in 10- to 14-year-olds (8.1% of total deaths) and ranked eighth among 15- to 19-year-olds (3.4%).2
Recent trends show some success in battling the epidemic among adolescents. HIV prevalence in pregnant women aged 15–24 years has shown a statistically significantly decline in recent years in 12 of 24 high-prevalence countries. Yet, progress remains tenuous in many countries, and incidence levels are rising in some. Currently, of the 2.1 million HIV-infected adolescents aged 10–19 years living with HIV, 1.3 million are girls and 870,000 are boys.3 Some adolescents are infected as children through mother-to-child transmission and grow into their adolescent years carrying the virus. Others are infected through unprotected sex or are injecting drug users who get the infection by sharing contaminated needles. Because prevention of mother-to-child transmission (PMTCT) programs started scaling up only about 10 years ago, in many countries in sub-Saharan Africa, about half of all current infections among 15- to 19-year-olds are in children who were infected at birth.
To address the needs of these adolescents, researchers and programmers have attempted to identify effective programming for adolescents.4–8 Effective approaches include interventions whose focus is adolescents; these include school-based sexuality education and support for orphans and vulnerable children and customizing existing interventions, such as mass media and clinical health services, to meet the specific needs of young people.
The IF was developed with the entire population in mind. The purpose of this article is to show the impact of implementing it on adolescent HIV outcomes. Thus, this article does not present an optimal response to adolescent HIV, but, rather, it focuses on the benefits to adolescents of the IF as it currently exists.
ADAPTING THE INVESTMENT FRAMEWORK TO ADOLESCENTS
About the Investment Framework
The IF groups interventions into 3 main categories based on their contribution to overall program impact: Basic Programs for which there is sound evidence of direct impact, Critical Enablers that support Basic Programs and may, themselves, have some direct impact, and Development Synergies that support the broader development agenda and have indirect effects on HIV transmission and mortality (see Figure S1, Supplemental Digital Content, http://links.lww.com/QAI/A521, for a schematic of the IF and Table S1, Supplemental Digital Content, http://links.lww.com/QAI/A521, for a description of each intervention).
To measure impact, we used the Goals model to conduct a detailed modeling for 26 countries, which account for 78% of the total HIV infections (see Table S6, Supplemental Digital Content, http://links.lww.com/QAI/A521). The Goals model simulates HIV incidence on the basis of behaviors of key population groups (stable couples, those engaging in casual sex, sex workers and clients, men who have sex with men, and people who inject drugs) and tracks new infections over time as they progress through primary infection, chronic infection, late stage infection, eligibility for treatment, and AIDS death. The model estimates incidence for the 15- to 49-year age group, distributes new infections by age, and tracks demographic and epidemics processes by age (see Supplemental Digital Content—The Goals Model, http://links.lww.com/QAI/A521, for more details on the model). We then scaled up the results to 157 low- and middle-income countries. The Goals model combines information on coverage, intervention effectiveness, and number of adolescents in need of services to produce estimates of impact. To estimate the number of adolescents requiring each intervention, we calculated 10- to 19-year-olds as a proportion of all people eligible to receive that intervention.
The coverage targets from the IF assume a scale-up to universal access to all the interventions by 2015 (SDC Table 1). Baseline coverage of the interventions was determined using the data available in 2011. We constructed 4 alternative coverage scenarios:
1. Base: The base scenario assumes constant coverage for prevention and treatment.
2. IF: In the IF scenario, all countries scale up services to reach target levels by 2015.
3. IF Delayed: Under the IF Delayed scenario, countries do not achieve coverage targets until 2020.
4. IF Low: The Pessimistic scenario models the consequences of not investing adequately in that adolescent markup to ensure that countries appropriately tailor programs to the 10–19 age group. It assumes lower coverage targets, a 5-year delay in reaching those targets, and low values of intervention effectiveness.
Estimates of the effectiveness of biomedical interventions [condoms, male circumcision, PMTCT, and antiretroviral therapy (ART)] are based on the findings from randomized control trials (see Table S7, Supplemental Digital Content, http://links.lww.com/QAI/A521). The effects of behavior change interventions are less certain. Estimates of these effects are drawn from an extensive literature review9 (see Table S4, Supplemental Digital Content, http://links.lww.com/QAI/A521). Effect values from the literature are used to provide a mean, low and high values for each behavior change intervention and each population group (stable couples, those with multiple partners, sex workers and clients, men who have sex with men, and people who inject drugs). For 3 of the 4 scenarios modeled in our analysis, we use the medium values for intervention effectiveness. The fourth scenario, the low scenario, assumes low values of intervention effectiveness.
Underlying the cost estimates of the IF is a matrix of unit costs by intervention drawn from hundreds of studies from around the globe. (The database and studies are available in an interactive database at www.FuturesInstitute.org—select Policy Tools -> HIV -> Unit Cost Database). For interventions that are adolescent focused, such as school-based programs and broader protection, care and support programs, we drew on the existing cost matrix for our estimates. For interventions that encompass a broad range of ages, however, there is no simple way to determine the cost of applying a particular intervention to an adolescent population. This is because almost no studies explicitly compare the cost of providing an intervention to adolescents versus that for adults. One approach is to assume that unit costs are the same, regardless of the target age group. This method, however, runs counter to adolescent program design principles, which posit that such programs have inherently different characteristics from programs for adults. Thus, we took an approach that determines the extent to which unit costs might differ for adolescent versus adult programming. For each intervention, we assessed 10 program elements and judged whether serving adolescents versus adults requires additional, the same, or fewer resources. This assessment produced a score that determined the cost of serving adolescents compared with that of serving adults (expressed a percentage markup over adult unit cost) (see the SDC Table 3 for more on how we determined the markup). We classified each intervention according to whether there would be no markup or whether there would be a low (10%), medium (25%), or high (40%) markup (Table 1). We set the upper bound based on recent work costing a package of adolescent health services.10 These markups are also consistent with those in the literature on adolescent-friendly services, which posits that attracting and serving an adolescent population requires additional programmatic effort.11
IMPACTS AND COSTS
New Infections Among Adolescents
Under the Base scenario, with constant coverage, the number of new HIV infections among adolescents each year will increase slightly (Fig. 1). This represents a reversal of the downward trend of the last 10 years because we assume no increase in coverage after 2012, in contrast to the recent rise of coverage of several key interventions. The result is that the incidence rate is essentially constant in this scenario and, because of continued population growth, the number of new infections increases.
Under the IF scenario, new infections to adolescents drop substantially, to 165,000 annually by 2020. The overall effect is to avert 2 million new adolescent infections by 2020, half of the new infections that would occur to adolescents if countries were to maintain current levels of investment. A delay in reaching the coverage targets (IF Delayed) reduces the impact by one-third, whereas the low coverage and low interventions effectiveness under the IF Low scenario produces even less impact (only a 20% decline in the annual number of new infections).
Each of the interventions classified as Basic Programs has been shown to prevent new HIV infections. The exact impact of each intervention depends on the context. The impact will be larger for interventions that currently have low coverage if they can be scaled up quickly than for those interventions that are already at high coverage. In some countries, most new infections occur among most-at-risk populations (sex workers and their clients, men who have sex with men, and people who inject drugs). In those settings, scaling up prevention interventions for those populations can have a large impact. In other settings where most new infections arise in other population groups (those who engage in casual sex, migrant workers, etc.), the impact may be less. The pattern of new infections by scenario is much the same in countries with hyperendemic, generalized, or concentrated epidemics, but the number of new infections are higher in the generalized epidemics than in the other 2 types (see Figure S2, Supplemental Digital Content, http://links.lww.com/QAI/A521). Because adolescent females have higher rates of new infections than adolescent males do, most of the impact of the IF is on reducing infections in females, halving them from 320,000 to about 160,000 annually (Fig. 2).
AIDS Deaths in Adolescents
The impact of the IF on HIV-related deaths in adolescents is complicated, as illustrated in Figure 3. In the short term, the IF scenario will actually result in more deaths than will the Base scenario in the 10- to 14-year age group because so many more children will survive from ages 5 to 9. So though the mortality rate will be less under the IF scenario, the number of deaths will be somewhat higher for a few years. Eventually, the effects of fewer new infections in the IF scenario will dramatically reduce the number of adolescent deaths. During 2012–2020, the IF scenario will result in 43,000 fewer AIDS deaths, and the IF-Delayed scenario will have 40,000 fewer AIDS deaths than the Base scenario (data not shown) will have. The IF Low scenario will have about 3500 more deaths than the Base scenario will have during this period.
Adolescent HIV-Positive Population Under the Investment Framework
Through the actions of the IF interventions to prevent new infections, the number of HIV+ adolescents gradually decreases, from about 1.9 million in 2012 to 1.6 million annually by 2020 (see Figure S3, Supplemental Digital Content, http://links.lww.com/QAI/A521). The biggest impact is on infections in females aged 15–19, which are reduced by 46% by 2020 under the IF scenario and 36% and 23% under the IF-Delayed and IF Low scenarios, respectively. For males, the reductions are slightly less, 43%, 33%, and 21%, respectively. The number of HIV+ children aged 10–14 years is determined by the number of perinatal infections and their survival to age 10, so the impact of the scenarios on this age group will not appear for 10 years. After 2023, the number of HIV-infected 10- to 14-year-olds would decline sharply because there would be fewer children infected at birth and therefore fewer reaching the age of 10–14 years.
Estimated Resource Needs for Adolescents
The resources needed in low- and middle-income countries for adolescent programs increase from about $3.8 billion a year in 2010 to $5.5 billion by 2014, then decrease as the impact of prevention efforts, and efficiencies begin to take effect (Fig. 4). The composition of the interventions changes markedly over time: basic programs increase from one-fourth to one-third the total cost; critical enablers go from one-half to one-fourth of the total cost; and development synergies go from about one-fourth to 40% of the total cost.
Programs to support protection, care, and support of vulnerable children affected by AIDS will be the single biggest component of costs by 20201 (17% of the total), followed by ART (11%), sexually transmitted infection management (7%), community mobilization (6%), and programs for injecting drug users (6%). It is important to note that counseling and testing programs are integrated into many of these interventions including voluntary counseling and testing, outreach to most-at-risk populations, PMTCT, community mobilization, and provider-initiated counseling and testing. New WHO guidelines on Antiretroviral Treatment were released in mid-2013, and national programs will consider and adopt these guidelines in the near future.12 It is likely that these changes will increase the resources needed for ART by about 50%.13
Putting Adolescents in Perspective
Our analysis shows that programs for adolescents account for about 24% of all the resources required for the IF from 2010 to 2020. Some key programs are mostly focused on adolescents (school-based and youth-focused programs, support for children affected by AIDS, male circumcision), whereas others, such as treatment, are mostly focused on older adults and on younger children. Our analysis indicates that programs for adolescents account for 24% of resources needed for prevention, 10% for treatment, and 37% for mitigation (from 2010 to 2020). (1 Interventions for the protection, care, and support of children made vulnerable by HIV and AIDS based on the UNICEF and Futures Institute 2013 resource estimates for children affected by AIDS include economic strengthening, education support, social care and community outreach, and program support).
The impact of the IF on adolescents is for both immediate and long terms. In terms of immediate impact, our analysis shows that programs focused on adolescents will account for 13% of infections averted by the IF. But the full impact of these programs can only be seen in the longer term and the safer behaviors adopted by adolescents also protect them in their later adult years. Programs such as male circumcision, school-based AIDS and comprehensive sexuality education, and community mobilization are investments in long-term prevention that return dividends far into the future. Support for protection, care, and support of children affected by AIDS accounts for a large share of resources required, but these programs can make a long-lasting difference to individual well-being and social and economic development.14
DISCUSSION AND CONCLUSION
Although the overall trends in battling the HIV/AIDS epidemic are positive, AIDS continues to be a major health problem, especially in the poorest countries. To maximize their use of limited resources, countries need effective programming that targets the special needs of adolescents. The IF approach is one way to bolster the strategic focus of the adolescent HIV and AIDS response. We show that applying the principles of the IF to the adolescent age group could avert 2 million new adolescent infections by 2020, half of the new infections that would occur in adolescents if countries were to maintain current levels of investment. A 5-year delay in implementing the full IF would reduce the impact by one-third. This delay, coupled with lower program effectiveness resulting from inattention to the specific needs of adolescents, would result in only a 20% decline in the annual number of new infections. Notably, the implementation of the IF package will have its largest direct impact on adolescent girls, the group that is most vulnerable to infection. To achieve the goals of the IF, resources will have to expand by 45% to $5.5 billion by 2014, then decline somewhat as new infections decrease.
The framework makes a critical distinction between the 3 types of interventions: basic programs would consume about one-third of resources, critical enablers about one-fourth, and development synergies would consume 40% of resources for adolescents. Because they have a direct impact on averting infections, all countries should scale up basic programs as quickly as possible. Investments in critical enablers and development synergies should be context specific. However, countries should take particular care in how they use the specific results for deciding how to invest. The IF helps countries take a first, critical look at what the broad strategic response for adolescents should involve in a particular country context. To take that a step further, however, and focus on the operational questions in a critical way, countries need to look closely at their epidemics and who is being reached and how, to make decisions on realignment of resources. The IF approach can be an important prerequisite to doing that kind of country process well.
It is important to reiterate that that analysis examines the impact on adolescents of the IF as proposed in 2011.1 It does not propose a new program specifically optimized for adolescents. Our approach faces many of the same limitations that exist for the IF more generally and that have been discussed extensively elsewhere.1 Perhaps the most critical limitation from the adolescent standpoint is the way the framework places interventions into the 3 program categories. Judged according to evidence for impact, the basic programs category excludes the school-based sexuality education interventions that have been among the most widely implemented of adolescent-focused programs. This was partly because they are behavioral interventions and are thus not easily amenable to measurement of effect. Their exclusion from the basic programs category may unwittingly send a message that they are not valued. On the other hand, the inclusion of such programs among development synergies recognizes that these sorts of interventions have impacts that go beyond HIV-specific outcomes and contribute on multiple fronts toward behavior change communication, community mobilization, and the broader synergistic effort. For now, however, an unavoidable weakness of the model is that it may not fully represent the range of HIV interventions, particularly those for which research has not rigorously proven impact.
Another limitation specific to the adolescent focus of the model is the estimate of the number of adolescents currently covered by specific interventions, some of which we based on relatively slim data. Improvements in the precision of these numbers will likely have a small effect on the global model results but could substantially alter the application of the model to a particular country.
Despite these limitations, this application of the IF principles to the adolescent age group is a valuable exercise. It shows that countries that put in place a holistic package of effective Prevention interventions can reduce new HIV infections (including PMTCT to reduce perinatal HIV infections and other interventions that affect sexual transmission of HIV during adolescence) and, through expansion of treatment, can reduce HIV-related illness and deaths in adolescents. It highlights the centrality of the adolescent age group in the fight against AIDS by showing that countries should target about a quarter of all their resources to this age group. Moreover, the social, health, and development effects of investing in adolescents—investing in preventing HIV infection, encouraging healthy behaviors, treating and caring for the infected, and mitigating the impact of the epidemic on the most vulnerable—will be felt far beyond the adolescent years. It is the responsibility of national and international organizations to fund and implement adolescent-focused interventions in the IF and to ensure that countries tailor interventions to the specific needs of people at this critical stage of their lives.
* Members of the Modeling Validation Group include
Delivette Castor (USAID), Charlotte Deogan (Karolinska Institute), Beverly Jane Ferguson (WHO), Gwyn Hainsworth (Pathfinder), Nina Hasen (OGAC), Bill Kapogiannis (NIH), Miles Kemplay (CIFF), Anita Krug (Youth Rise), Marjorie Opuni (UNAIDS), Mary Otieno (UNFPA), David Ross (LSHTM), Yong Feng Liu (UNESCO), Bettina Schunter (UNICEF), Amaya Gillespie (UNICEF), Ken Legins (UNICEF), Patricia Lim Ah Ken (UNICEF), Luong Nguyen (UNICEF), Rick Olson (UNICEF), Pierre Robert (UNICEF), Chiho Suzuki (UNICEF), and Rachel Yates (UNICEF).
1. Schwartländer B, Stover J, Hallett T, et al.. Towards an improved investment approach for an effective response to HIV/AIDS. Lancet. 2011;377:2031–2041.
2. Murray CJL, Vos T, Lozano R. Disability-adjusted life years (DALYs) for 291 diseases and injuries in 21 regions, 1990–2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet. 2012;380:2197–2223.
3. UNICEF. Progress for children: a report card on adolescents. 2012. UNICEF. Progress for Children: A report card for adolescents, Number 10. New York: United Nations Children's Fund, April 2012.
4. Ross D, Dick B, Ferguson J. Preventing HIV/AIDS in young people: a systematic review of the evidence from developing countries: UNAIDS interagency task team on HIV and young people. 2006.
5. London School of Hygiene and Tropical Medicine and the Mwanza Research Centre of the Tanzanian National Institute for Medical Research. HIV Prevention Among Young People in Sub-Saharan Africa the Way Forward Report from a Workshop Held 14–17th September 2009, White Sands Hotel. Dares Salaam, Tanzania. London: London School of Hygiene and Tropical Medicine; 2009.
6. FOCUS on Young Adults. Advancing young adult reproductive health: actions for the next decade. End of program report. 2001.
7. UNICEF. Opportunity in crisis: preventing HIV from early adolescence to young adulthood. 2011.
8. Napierala Mavedzenge S, Luecke E, Ross DA. Effective approaches for programming to reduce adolescent vulnerability to HIV infection, HIV risk, and HIV-related morbidity and mortality: A systematic review of systematic reviews. J Acquir Immune Defic Syndr. 2014;66(suppl 2):S154–S169.
9. Bollinger LA. How can we calculate the “E” in “CEA”? AIDS. 2008;22(suppl 1):S51–S57.
10. Deogan C, Ferguson J, Stenberg K. Resource needs for adolescent friendly health services: estimates for 74 Low- and middle-income countries. PLoS ONE. 2012;7:e51420.
11. Tylee A, Haller Dagmar M, Graham T, Churchill R, Sanci LA. “Youth-friendly primary-care services: how are we doing and what more needs to be done?” Lancet. 2007;369:1565–1573.
12. World Health Organization. Consolidated Guidelines on the Use of Antiretroviral Drugs for Treating and Preventing HIV Infection. Geneva, Switzerland: WHO; 2013.
13. WHO. Global Update on Treatment 2013: results, Impact and Opportunities. Geneva, Switzerland: WHO; 2013.
14. Bell C, Devarjan S, Gersbach H. The Long-Run Economic Costs of AIDS: theory and Application to South Africa. Washington, DC: World Bank; 2003.