JSI Research and Training Institute, Inc.

This 86-page report is an evaluation of the African Youth Alliance (AYA), an organisation established by The Bill & Melinda Gates Foundation to be a collaborative, and comprehensive prevention programme for improving adolescent sexual and reproductive health (ASRH) among young people age 10–24 in Botswana, Ghana, Tanzania, and Uganda. Implemented between 2000 and 2006, the AYA programme strategy focused on implementing and scaling up a defined set of integrated and comprehensive ASRH interventions using existing local institutions. A variety of groups worked with AYA on the behaviour change communication (BCC) component, implementing strategies that included the spread of messages through mass media, group presentations, and one-on-one peer education. This report presents the results of the Tanzania evaluation. The research found that AYA’s integrated programme had a substantial impact on most of the desired outcomes, especially among female youths.

The report outlines the following key findings of the evaluation research:

• Female respondents who were exposed to AYA were significantly more likely to have better knowledge about HIV/AIDS and condoms (spontaneous response) than those who were not exposed to AYA. There was no definitive impact (corroborated by multiple analysis methods) on prompted knowledge. No impact was observed among males.
• Male and female respondents who were exposed to AYA were significantly more likely to feel confident insisting that a partner use condoms than were those not exposed to AYA. Females who were exposed to AYA were significantly more likely to be confident about putting on a condom correctly. No impact was found on other self-efficacy variables.
• Exposed females were less likely to abstain from sex than were unexposed females. No definitive impact was seen among males.
• Female respondents who were exposed to AYA were more likely to report fewer lifetime sexual partners than were those who were not exposed to AYA. No impact was detected among males.
• Male and female respondents who were exposed to AYA were significantly more likely to use condoms at first sex and to use condoms consistently with a current partner than those who were not exposed to AYA.
• Females who were exposed to AYA were significantly more likely to use contraceptives at first sex and at last sex. Males who were exposed to AYA were significantly more likely to use modern contraceptives at first sex than those who were not exposed to AYA.

The main conclusions of the evaluation are summarised as follows:

• Significant impact was observed on antecedents - A positive impact was observed for spontaneous HIV/AIDS knowledge, positive attitudes toward condom users, correct use of condoms among females, and confidence in insisting that a partner use a condom among both males and females.
• Either little impact or a negative impact was found on sexual debut and abstinence - The lack of any impact on sexual debut and abstinence is likely to stem from AYA’s focus on providing abstinence messages to younger age groups. The impact on all those behaviours may also take a longer time to emerge than was available during the AYA programme’s implementation.
• A strong impact was detected in certain areas - Those successes covered (a) condom and contraceptive use among females and males and (b) partner reduction among females. In terms of the programme’s overall impact, this change was considered the achievement of AYA in Tanzania.
• The impact differs among males and females - Although the findings represent an outstanding achievement of AYA, it is important to ask why there was a greater impact on the behaviour of females than on the behaviour of males.
• Lack of a statistically significant impact does not necessarily imply no benefit from AYA - This evaluation did not detect a positive impact on several outcomes such as the belief that condoms protect against HIV, confidence in obtaining a condom when needed, sexual debut, and abstinence. However, lack of a significant finding does not necessarily mean that the programme did not have any impact in those areas.
• The research design was effective - The impacts of AYA varied for some outcomes by method of analysis, but, overall, the results were consistently corroborated for both antecedents and behavioural outcomes. This convergence of findings through the different designs and analysis approaches provides confidence that the impact detected through the evaluation can indeed be attributed to AYA’s programmes.

According to the evaluation results, recommendations tend to fall into three general categories: (a) expanding interventions that appear to lead to a positive impact, (b) seeking creative alternatives to address areas that show a limited impact, and (c) carrying out further data analysis or additional research.

In conclusion, evidence from this impact evaluation suggests that multiple component programmes that combine (a) strategies such as behaviour change communication address risk behaviours, (b) youth-friendly services (for HIV/STI counseling and testing modern contraceptives, including condoms), and (c) outreach services such as peer education and other activities in the community can be an effective approach to addressing young people’s ASRH needs. The success of such programmes also depend on cultural adaptation of programming to the local context, sensitivity to the needs of young people, and efforts to build on the strengths of local institutions.

JSI website on March 31 2009.