In Zimbabwe, where 38% of the population is aged 10-24,1 the average age at first intercourse is 18 for both men and women,2 but many begin sexual experimentation far earlier.3 Unprotected sex puts young people at risk of unwanted pregnancies, which may contribute to their dropping out of school, marrying early, abandoning babies and obtaining abortions.4 Sexually active young people also face the risk of contracting HIV and other sexually transmitted infections (STIs). Zimbabwe has one of the highest AIDS prevalence rates in the world;5 HIV infection rates there are highest before age 25, and among teenagers, women are especially vulnerable.6
Most young people in Zimbabwe are aware of HIV, AIDS and the risk of pregnancy but still engage in unprotected sex.7 According to the 1994 Demographic and Health Survey (DHS), 98% of women aged 15-19 had heard of AIDS, but only one-third of those who were unmarried and sexually active were using modern contraceptives, and only 19% were using condoms.8 In Zimbabwe, as in other countries around the world, gender roles and social norms—along with a host of economic and legal factors—contribute to risky sexual behavior.9 Stereotyped sexual norms and peer pressure encourage young males to prove their manhood and enhance their social status by having sex. At the same time, young women are socialized to be submissive and not to discuss sex, which leaves them unable to refuse sex or insist on condom use. Women's economic dependence on men also leads young females to exchange sex for the opportunity of marriage or for gifts, sometimes with older "sugar daddies," who may be HIV-infected.10
Other societal influences have exacerbated this situation. Traditionally, aunts, uncles and other extended family members provided sexuality-related information to young people, but as urbanization increases the distance between family members, parents are taking greater responsibility in this area, and many feel uncomfortable in this unaccustomed role.11 Health care providers have not filled the void because they share the overall societal bias against adolescent sexuality, they lack the skills needed to communicate with young people about sensitive topics and they are barred by law from providing reproductive health services to individuals younger than 16. According to one study, 72% of service providers believe that contraceptives should not be offered to people aged 16 or younger.12 Furthermore, even after age 16, most teenagers except for married women with children are denied contraceptives.13 As a result, young people in Zimbabwe generally have inadequate information about reproductive health issues, lack the skills to negotiate with their partners about delaying sex and have limited access to reproductive health services.14
To address these problems, the Zimbabwe National Family Planning Council (ZNFPC) launched the Promotion of Youth Responsibility Project, with technical assistance from the Johns Hopkins University Population Communication Services. The project aimed to encourage young people to adopt behaviors that reduce the risk of pregnancy and STIs, including HIV. It encouraged abstinence for young people with no sexual experience, but promoted condom use and a reduction in the number of sexual partners for those already sexually active.
In this article, we present an assessment of the project's success at reaching its target audience and promoting responsible sexual behavior among young people.
At the heart of the project was a six-month multimedia campaign directed at young people in five pilot sites: one urban area (Mutare) and four growth points (Maphisa, Nemanwa, Nzvimbo and Tongogara). (Growth points are small towns at the center of rural districts.) The campaign was based on the Steps to Behavior Change framework, which synthesizes theories of communication and behavior change into a practical model to guide reproductive health communication programs.15 The framework describes five stages through which people pass as they change their behavior: knowledge, approval, intention, practice and advocacy. Effective communication campaigns determine the stage that their audience is at and focus their energies accordingly.
Given the situation in Zimbabwe, the youth campaign focused on the model's two earliest stages, when people learn key information and skills, then discuss campaign messages with others and find support for behavior change among their family, peers and community.16 International experience with pregnancy and HIV prevention programs for adolescents has found that outside approval is critical for two reasons. First, young people's decisions are strongly influenced by friends, family and social norms. Second, their access to reproductive health information, commodities and services is controlled by adults, including parents, service providers and political leaders. Therefore, it is not surprising that programs have found it easier to improve knowledge and attitudes than to prompt behavioral changes. Effective interventions have addressed gender roles, sexual norms, peer pressures and public policies, in addition to teaching basic information and skills.17
Youth Campaign Goals
The first objective of the youth campaign was to increase reproductive health knowledge, because young people in Zimbabwe had only a general awareness of HIV, AIDS and family planning. They did not appreciate the personal risks of unprotected sex and did not know how to negotiate with partners about sexual decisions.18 Two other objectives were to heighten approval of safer sexual behaviors and of the use of reproductive health services—first, by promoting discussion of sexual matters with family and friends, and second, by encouraging parents, leaders and policymakers to support reproductive health communication and services for young people. The final objective was to encourage young people to adopt safer sexual behaviors and attend service facilities.
The main themes of the youth campaign, which was launched during August and September 1997, were self- respect and self-control. These were expressed in three slogans, which were repeated in all materials and activities, in both English and native languages: "Have self-control," "Value your body" and "Respect yourself." All campaign materials and activities were designed to reinforce a single set of messages, emphasizing the consequences of unprotected sex; negotiation skills; discussion with friends, family and providers; and safer sexual behaviors. Young people helped design appealing materials and relevant messages, and local management committees helped plan and execute activities at each site.
ZNFPC wanted to reach an audience of 10-24-year-olds of both sexes and, secondarily, the adults who control young people's access to reproductive health information and services. The age range selected for the young people was chosen for two reasons: Studies suggest that sexual responsibility interventions have the greatest impact on young people before they initiate sexual activity, perhaps even before they reach puberty.19 On the other hand, by encouraging adolescents who are already or are on the verge of becoming sexually active to take preventive measures (e.g., use condoms and reduce their number of partners), safer sex interventions can make an immediate impact on older teenagers.
The campaign employed a mix of communication channels, since different media can reach different audiences. Also, by repeating and reinforcing messages, a strategy combining media and activities increases the likelihood that people will recall and act on campaign messages.20 Some of the campaign's components employed mass media, which reach large audiences at a low cost per capita, raise awareness, disseminate information and have the potential to change behavior.21 The remaining components employed interpersonal communication, which reaches fewer people but may be effective in motivating people to adopt new practices.22 The campaign's logo, a yellow triangle with a blue circular "youth-friendly" seal, was featured on all materials.
•Posters. A series of eight posters carried messages like "Value your body and a happy future lies ahead" and "You may think you are ready for sex, but are you ready for the consequences?" In campaign sites, 10,000 copies of each poster were distributed. Older ZNFPC posters addressing AIDS, peer pressure, parental communication, sexual responsibility, drugs and alcohol also were on display.
•Leaflets. Five leaflets—on abstinence, how to say no to sex, postponing sex, delaying parenthood and STIs—were produced, and 19,000 copies of each were distributed. In addition, the popular older ZNFPC booklet Facts About Growing Up continued to circulate.
•Newsletter. Peer educators and schools distributed 100,000 copies of Straight Talk, a four-page newsletter on reproductive health issues of importance to young people. Each of the three issues included advice columns and articles written by young people.
•Radio program. Radio is widely available in Zimbabwe: Ninety-four percent of urban and 87% of rural young people surveyed by this project had access to a working radio. During the campaign, 26 episodes of Youth for Real, a one-hour radio variety show, were broadcast nationwide. This weekly program, which has continued to air since the campaign ended, combines information and advice with entertainment such as music and minidramas. Listeners can phone the show to ask questions of a peer counselor and doctor.
•Launch events. To mobilize community support for the campaign, local committees spent months planning elaborate launch activities and garnered substantial support from local businesses. Popular musicians attracted large crowds to the launches, which featured different activities at each site, including speeches, dramatic performances, drum majorettes, soccer games, donkey parades and a parachute drop. Adults who influence youths—including chiefs, counselors, church leaders, parents, siblings, teachers and service providers—attended the launches, and novelty items bearing campaign messages were distributed.
•Dramas. During the first two months of the campaign, two community theater troupes performed daily at schools, churches and town centers, presenting interactive dramas on reproductive health issues. Peer educators accompanied the troupes and facilitated a discussion with members of the audience after each performance.
•Peer educators. Peer educators aged 18-24 were recruited from the community and trained to speak with groups of young people at schools, churches and town centers and with individuals at home. Four educators were assigned to each growth point and six to Mutare.
•Hot line. A hot line was established at the Mutare Youth Center. Staffed by peer educators, it operated eight hours a day and was used to disseminate campaign messages as well as to answer individual queries. Because the hot line was promoted nationally on Youth for Real, it received calls from all over the country. However, access to telephones is limited in Zimbabwe, especially in rural areas.
Like most youth programs, the campaign conducted some activities and disseminated some materials through the schools, although it did not employ conventional, classroom-based education. The drama troupes and peer educators spent much of their time in schools and distributed copies of the newsletter there. Nearly all 11-15-year-olds (87%) are enrolled in school in Zimbabwe, but this proportion drops precipitously after age 16, especially among women.23 Therefore, while this strategy is effective for reaching young adolescents, other interventions are needed to reach older, out-of-school youths.
To increase the reach and impact of the campaign, the launch events, radio program and dramas used an entertainment-education strategy. This strategy, which inserts educational content into entertainment media, has proven effective in disseminating development and health messages around the world. It attracts and holds the attention of large audiences, provides role models for social learning and generates an emotional response that can heighten the impact of messages.24 It is a proven and potent strategy for young people, who enjoy mass media entertainment but frequently receive misleading messages about sex from the media.25 In Tanzania, for example, a radio soap opera prompted behavioral changes to prevent HIV by increasing audience members' perceptions that they were at risk, their belief in their ability to prevent the disease, discussion with peers and modeling of characters' behavior.26
Budget limitations precluded producing the print materials and radio program in multiple languages. English was used because it is understood throughout the country (in contrast to Zimbabwe's main native languages, Shona and Ndebele). Most young people in Zimbabwe understand and read English, and English is the language commonly used on radio.
Finally, since a lack of reproductive health services for young people may reduce the effectiveness of media promotions of safer sex,27 ZNFPC designated 26 clinics in the campaign areas as "youth-friendly." During a one-week course using a newly developed manual, ZNFPC trained one family planning provider from each of these clinics in interpersonal communication and youth counseling skills. After these providers returned to their home clinics, they were expected to train their coworkers in youth counseling. Throughout the campaign, peer educators, drama groups and print materials referred young people who needed reproductive health services to the youth-friendly clinics.
Baseline and Follow-Up Surveys
The evaluation applied a quasi-experimental design with baseline and follow-up surveys to assess changes in reproductive health knowledge and attitudes in both experimental and comparison groups. The impact of the campaign on behavior was assessed by self-reported actions. Using retrospective and self-reported actions to measure behavior change runs the risk that respondents may not report their actions accurately because of lapses in memory, a desire to please the interviewer or discomfort with discussing sexual matters. However, evaluations of other adolescent reproductive health campaigns have used this approach successfully.28
Young people in the five campaign sites made up the experimental group. Young men and women in two other sites, Kwekwe (a city) and Mubaira (a growth point), where ZNFPC made no intentional effort to reach youths with the multimedia campaign, were the comparison group. This comparison group is not a true control group, however, because residents were exposed to some elements of the campaign—the nationally broadcast radio program, advertisements for the hot line on that program and a ZNFPC clinic calendar featuring the campaign's posters. Young people in both the experimental and the comparison groups also were exposed to other activities and materials directed to youths that were not part of the campaign, including ZNFPC print materials from an ongoing effort to reach young people, a condom marketing program, peer educators and youth dramas sponsored by an HIV prevention program, and family life education in the schools.
The baseline survey was conducted among 1,426 respondents in April and May 1997, three months before the campaign was launched. The follow-up survey, including 1,400 respondents, was conducted one year later; this was approximately three months after the end of the campaign, although some activities continued beyond the six-month campaign period. In addition to repeating the questions asked at baseline, the second survey asked respondents whether they had been exposed to various campaign components and what actions they had taken in response.
Four teams conducted the fieldwork; each consisted of a supervisor and four interviewers, most of whom had no prior connection to the project. In some cases, ZNFPC staff collected data, but they were assigned to sites where they had not worked before. The DHS sampling frame was used to select houses randomly within a 30-kilometer radius of the town center; within each household, one youth aged 10-24 of the same gender as the fieldworker was randomly selected for interview. Before conducting an interview, the fieldworker explained the reason for the research, described the content of the questionnaire and asked for the consent of the potential respondent or of a parent or guardian if the respondent was younger than 15.
Urban Survey of Radio Listeners
Youth for Real was broadcast nationally on Radio 3, a station whose English-language programs generally are more popular in urban than in rural areas. To assess the program's reach throughout Zimbabwe and to help determine its long-term prospects, ZNFPC commissioned a survey of 700 youths aged 10-24 who lived in six cities outside of the campaign areas (Bulawayo, Gweru, Harare, Kadoma, Kwekwe and Masvingo).
Characteristics of Respondents
Almost equal numbers of young men and young women were interviewed in campaign and comparison sites at baseline and follow-up (Table 1). Roughly 20-30% of respondents to both surveys were aged 10-14, about 50% were aged 15-19 and the remainder were aged 20-24; 13% of females had not begun menstruating (not shown). Virtually all youths had at least a primary education; about two-thirds had a secondary education. About 50-65% of respondents were currently enrolled in school. The vast majority of respondents were single, and no more than one-third reported being sexually experienced; about half had ever had a boyfriend or girlfriend (not shown). About three-quarters of sexually experienced respondents reported having intercourse within the past six months, and the same proportion reported one or two sexual partners in their lifetime (not shown).
Some differences emerged between the baseline and follow-up samples and between campaign and comparison sites. At baseline, respondents in campaign sites were younger, less well educated, less likely to be married and less likely to be sexually experienced than those in comparison sites (Table 1). In campaign sites, respondents to the follow-up survey were significantly older, better educated and more likely to be sexually experienced than participants in the baseline survey. The rural-urban composition of the campaign and comparison samples also differed: Four of five respondents at campaign sites lived in rural areas, compared with half at comparison sites. Given these differences, we performed multivariate logistic regression analyses to control for age, sex, education, sexual experience, marital status and urban-rural residence.
Significantly higher proportions of youths in campaign than in comparison sites were exposed to each campaign component; nevertheless, exposure levels to some components were substantial in comparison sites (Figure 1). Two frequent problems for experimental research designs in mass media campaigns may explain this finding. First, the posters, radio program and leaflets were available in comparison sites and had some impact on young people there. Second, if respondents were unable to distinguish between campaign and similar noncampaign activities, the follow-up survey may have measured exposure to dramas and peer educators sponsored by programs other than the youth campaign. The inclusion of noncampaign activities presumably boosted exposure levels by roughly equal amounts in campaign and comparison areas. Because of the considerable exposure levels in some comparison sites, all analyses include the comparison as well as campaign sites.
Ninety-seven percent of respondents in campaign areas were exposed to the campaign. Posters and launch day events reached the largest proportions of young people (92% and 87%, respectively), followed by the leaflets (70%) and dramas (46%). The hot line reached the smallest share of the target audience (7%). However, the likelihood of respondents' exposure to each component varied according to the youths' characteristics.
Students were exposed to significantly more campaign components than out-of-school youths (on average, 3.4 vs. 3.0), presumably because much of the campaign was school-based. When background characteristics are taken into account, students were twice as likely as other young people to have seen a poster, read a pamphlet, watched a youth drama, read the Straight Talk newsletter or talked to a peer educator. Respondents aged 10-14 were exposed to fewer campaign components than older respondents (3.0 vs. 3.3) and were significantly less likely to have been exposed to posters, campaign launches, Straight Talk, the radio program and peer educators. Sexually experienced youths were exposed to the same number of components (3.3) as sexually inexperienced respondents (3.2), but were more likely to have been exposed to the four components with the least reach: the newsletter, radio program, peer educators and hot line.
While gender and urban-rural residence also were related to campaign exposure, the direction of the association varied for different materials and activities. Some components reached a greater proportion of women than men (posters, pamphlets and the radio program), while others reached a higher fraction of men (dramas, newsletter and peer educators ). Similarly, launch events, youth dramas, Straight Talk and the peer educators had their greatest reach in rural campaign sites, while leaflets, the radio program and the hot line reached a higher proportion of urban than rural youth.
The six-city survey provides a fuller picture of listenership to Youth for Real. Nationwide, 41% of young people living in urban areas had heard the program—a substantial proportion, considering that the show had been on the air for just six months at the time of the survey. Five percent of listeners reported having called the show to discuss problems with boyfriends, gangs, drugs, STIs and other issues.
In the campaign sites, 67% of young people recognized the campaign's logo. By comparison, 98% recognized the Coca-Cola, Dairy Board and Bata Shoe logos, all of which have been in use for more than a decade. Ninety-four percent of respondents in campaign sites recognized the Shona and Ndebele slogans, compared with 52% in comparison sites. High rates of recall in comparison sites probably reflect that the slogans were phrases in common usage and appeared nationwide on clinic calendars.
Gains in Knowledge
The campaign did not produce new materials on contraceptive methods because such pamphlets and posters already existed. However, it provided an enabling environment for young people to learn about contraceptives. In campaign sites, respondents' knowledge of every contraceptive method except the implant increased significantly from baseline to follow-up, when background variables are controlled for: Teenagers in these areas were 2-4 times as likely to know of most methods after the campaign as they had been before, and about eight times as likely to know of the female condom (Table 2). In comparison areas, knowledge levels rose significantly for fewer methods, and the magnitude of the changes was smaller. The dramatic increases in awareness of the female condom in both campaign and comparison sites were due to a separate initiative to promote this method.
Young people's level of general reproductive health knowledge remained low after the campaign, especially on items regarding the safety and efficacy of family planning methods. In campaign sites, correct knowledge increased significantly for only one of six questions asked: whether family planning methods can cause deformities. In comparison sites, correct knowledge increased for whether a healthy-looking person can have HIV, but declined for whether family planning methods could cause infertility.
Approval and Attitudes
The campaign succeeded in generating discussion on a wide variety of topics, including sexual issues, HIV and AIDS, and physical growth and maturity. Analyses controlling for background variables reveal that in the period during and immediately after the campaign, respondents in campaign sites were significantly more likely than those in comparison sites to have a discussion with anyone about STIs and AIDS (78% vs. 67%), whether to have sex (77% vs. 69%), menstruation (56% vs. 47%), body changes associated with puberty (50% vs. 41%), the pressure to have sex (48% vs. 42%), sexual urges (43% vs. 34%), wet dreams (43% vs. 34%) and where to buy contraceptives (40% vs. 34%).
When asked if they had taken action as a result of the campaign, young people were most likely to report having discussed reproductive health issues with others (Table 3). Eighty percent of respondents in campaign areas said they had talked with someone—mostly friends (72%), but also siblings (49%), parents (44%), teachers (34%) and partners (28%). When background characteristics are accounted for, these youths were more likely than their peers in comparison areas to report such discussions (odds ratios, 3.5-5.7). While relatively few young people had spoken to their partners, many were not married or dating at the time of the campaign and thus did not have a partner to talk to.
The campaign had less success in shifting young people's thinking about gender roles. Respondents were asked whether the male, the female or both partners should be responsible for making sexual decisions. About four-fifths of young people in both campaign and comparison sites believed that the male should decide whether to have sex (not shown). Opinions did not differ by respondents' age or gender, and no significant change occurred between baseline and follow-up.
In theory, increased knowledge and heightened approval lead people to recognize that new behaviors can meet a personal need, to decide to take action and, eventually, to adopt new practices. The follow-up survey asked young people who were exposed to the campaign—regardless of whether they lived in campaign or comparison sites—if they had practiced certain safer sexual behaviors as a result. Saying no to sex was a major campaign message, but the phrase covers a wide range of possible behaviors and may best be interpreted as a sign of intention rather than actions taken. The odds that respondents reported that they had said no to sex in campaign sites were 2.5 times as great as the odds that youths in comparison sites said so. Young people at campaign sites also were somewhat more likely than youths in comparison sites to say that they were continuing to abstain from sex as a result of the campaign, which may indicate positive intentions.
According to the multiple regression analysis, young women were more likely than young men to report having said no to sex. In part, this reflects that young women are frequently pressured by boyfriends and older men to have sex (not shown). However, it also may suggest a positive change in women's attitudes about gender-appropriate behavior: As a result of the campaign, some young women may have come to believe that they had the right and responsibility to refuse unwanted sex.
The surveys also measured the campaign's effects on the likelihood that young people would seek services and use a modern family planning method. As a result of the campaign, young people in campaign sites were more likely to visit a health center (odds ratio, 4.7) and to visit a youth center (14.0) than were respondents in comparison sites (Table 3). Notably, the campaign encouraged groups that are historically less likely to seek services to visit a health center: males, single people and those who lack sexual experience (not shown). Among those exposed to the campaign, almost equal proportions of young men and women (29% vs. 28%) and of sexually inexperienced and experienced youths (27% v. 31%) visited a health center. The gap between single and married people also was smaller than expected (27% vs. 41%). In contrast, single people, sexually experienced youths and urban residents were more likely than their married, sexually inexperienced and rural peers to visit a youth center.
Use of modern contraceptives increased significantly in campaign sites between surveys: Among respondents who had had sex within the previous six months, the proportion who reportedly used a modern method during their last sexual encounter (not shown) rose from 56% at baseline to 67% at follow-up (odds ratio from multivariate analysis=1.7, p<.05). Use of modern methods did not change significantly in comparison areas.
By far the campaign's biggest effect was to convince sexually experienced young people to stick to one partner (Table 3): Sexually experienced respondents in campaign sites were much more likely than those in comparison sites to report taking this action as a result of the campaign (odds ratio, 26.1). They also were substantially more likely to start using condoms (5.7).
Effect of Different Components
To determine which campaign components were most effective, we assessed the impact of each independently. These analyses include all respondents in both campaign and comparison areas who were exposed to at least one component. Bivariate analyses found that nearly all campaign activities and materials had a significant impact on a broad range of respondents' self-reported actions. However, these analyses do not take into account that most respondents were exposed to multiple campaign activities and materials. Therefore, we conducted a multivariate analysis that controlled for respondents' exposure to all other campaign components (Table 4).
Although posters had the greatest reach, they had relatively little impact on respondents: Only discussion with friends and with teachers and intention to continue abstaining from sex were positively affected by posters, and the odds ratios were small. Launch events proved to have the strongest impact. Exposure to these events substantially increased the odds of youths' discussing reproductive health issues with others and of seeking services; it also significantly increased their likelihood of adopting safer sexual behaviors. (We cannot explain its disproportionate impact on avoiding sugar daddies.) Leaflets also proved surprisingly effective: Although the odds ratios were smaller than those for launch events, exposure to leaflets significantly increased the likelihood of every outcome except the intention to continue abstaining from sex. Dramas, which reached fewer than half of respondents, promoted discussion and health center visits.*
Along with more limited exposure, the remaining components had far less impact. These components may have had a stronger effect on sexually experienced young people than on respondents as a whole, but small sample sizes make it impossible to analyze this subgroup separately.
Sixty-one percent of respondents exposed to the campaign, in both campaign and comparison sites, saw or heard at least three campaign components. As the number of campaign materials and activities to which young people were exposed increased, their likelihood of discussing reproductive health issues, of adopting safer sexual behavior (except avoiding sugar daddies) and of seeking services increased (Table 5). The intensity of campaign exposure also had a positive influence on their knowledge of family planning methods, but it was not related to reproductive health knowledge or beliefs about which partner should make the decision to have sex (not shown).
Maximizing Campaign Exposure
Like several other multimedia campaigns promoting reproductive health among adolescents,29the Zimbabwe youth campaign reached more than 90% of its chosen audience, in most cases with multiple materials and activities. It succeeded in reaching young people of different ages and backgrounds because of the variety of activities and materials deployed. For example, launches proved especially popular in rural areas, where entertainment is limited, while the radio program and hot line had greater reach in urban areas, where young people are more receptive to English-language broadcasts and telephones are readily available.
Although it was harder to connect with 10-14-year-olds and sexually inexperienced youths than with others, the campaign did surprisingly well at reaching these groups, given the bias in Zimbabwe against teaching children that age about sexual issues and their lack of immediate need for reproductive health advice. Operating in the schools increased exposure among the youngest, least sexually active group. However, the best way to reach older, out-of-school youths proved to be activities that reach a general audience. Anecdotal evidence suggests that these activities also reached important secondary audiences among adults, including parents and providers.
High levels of campaign exposure and message recall were due to the appeal of the campaign components. This appeal, in turn, resulted from young people's participation in every aspect of designing and implementing campaign materials and activities. The entertainment-education strategy drew large audiences to launch events, but was not as successful (in terms of either exposure or impact) for the radio program. Language problems may explain the discrepancy: Rural youths prefer radio broadcasts in Shona and Ndebele rather than in English. Greater use of local languages in all components of the campaign might have increased its reach. (ZNFPC continued to air the radio show after the campaign ended, adding broadcasts in local languages and on other radio stations to reach rural youths.)
Compared with other multimedia campaigns promoting safer sexual behaviors among young people,30 the campaign had little impact on reproductive health knowledge and beliefs but generated high levels of interpersonal communication. A countercampaign run concurrently by a prolife organization may have contributed to young people's misconceptions about condoms, HIV and AIDS. The campaign's failure to emphasize basic facts about reproductive health may also explain its limited impact on knowledge in this area. Yet the campaign did prompt young people to discuss a range of reproductive health issues with friends and family, and early discussions about reproductive health issues may prompt more responsible decisions later in life.31 Indeed, a full assessment of the campaign's impact would follow up young people for years rather than months.
Given the campaign's brief duration and the preponderance of sexually inexperienced young people in its audience, it had a strong influence on behavior. While it is impossible to directly compare the impact of different adolescent health campaigns because various outcome measures are used, the proportions of respondents who reported changing their sexual behavior or seeking reproductive health services in response to the campaign in Zimbabwe are similar to those from other multimedia campaigns.32 However, the campaign did not increase contraceptive use as much as social marketing campaigns that have focused on promoting condoms.33
The use of multiple channels of communication contributed to the campaign's impact. The evaluation confirms a clear dose-response relationship between exposure and impact: The more materials and activities young people were exposed to, the more actions they took in response. Combining mass media and community events may have been particularly effective. An evaluation of the Safer Sex Campaign for young people in Uganda found that its featured radio program was most influential in districts that added local activities such as bicycle rallies and drama contests.34 Likewise, a comparison of four operations research projects in Sub-Saharan Africa found that the most effective adolescent sexual health campaigns combined mass media with interpersonal communication.35 In the Zimbabwe campaign, as elsewhere, local events ensured that messages were expressed in young people's own languages, in familiar contexts and with the endorsement of respected local figures. This finding confirms that mass media and interpersonal communication channels may play complementary roles in encouraging behavior change.36
Building Social Support
One of the campaign's greatest accomplishments was building support, in the community and within the health care system, for reproductive health interventions directed at young people. It achieved this by decentralizing management to local committees that included representatives from local government, religious, educational, health and business groups; by designing activities to reach a secondary audience of family, friends and teachers, and to prompt discussion of reproductive health issues; by training providers to overcome entrenched biases against offering reproductive health information and services to young people; and by involving providers in campaign preparations and launches.
Among the results of this strategy were unexpectedly high levels of parent-child discussion about sensitive reproductive health issues and increases in the number of young clients seeking reproductive health services, including STI treatment and family planning care, at youth-friendly clinics. Community support for the project also has enabled some key activities to continue, including training for peer educators, youth-friendly clinics and the Mutare hot line. Yet no direct, quantitative evidence for the project's impact on adults is available. Future evaluations should measure the campaign's effects on the knowledge, attitudes and behavior of key secondary audiences.
As part of the effort to build social support for sexually responsible behavior among young people, the project tried to address gender constraints on sexual behavior. The campaign called on young females as well as males to take charge of their lives and fostered individual self- esteem. However, it did not directly confront the problem of unequal gender relations and fell into a common trap by asking young women to take actions that contradict accepted female roles, such as refusing sex or insisting on condom use, without first empowering them.37 Gender attitudes have proven intractable in other adolescent health campaigns. For example, the Tsa Banana program in Botswana convinced young women of the health benefits of condoms, but the women still believed they would lose their partner's respect if they initiated condom use.38 About half of young women reported saying no to sex in response to the youth campaign. This is a hopeful sign, but future campaigns need to directly address the gender inequities that underlie risky sexual decisions by young people in Zimbabwe.