Part 10: Dr. Jeff Sheldon on an Adolescent HIV Prevention Program for Geita District, Mwanza Region, Northern Tanzania as Proposed to the NIH

Potential Challenges and Mitigation Efforts

Beyond challenges to prevention program development in Tanzania identified by Hartwig et al. (2005, 2006) - the silence of those in authority who perpetuate stigma, women’s changing values, and the low social status that genders risk-reducing behaviors - three intervention programs implemented in four developing sub-Saharan African countries (e.g., South Africa, Zambia, Ghana, and Nigeria) exemplify self-imposed challenges that come from poor design.?More importantly, they serve as a reminder that inherently flawed designs typically do not lead to intended outcomes no matter how sophisticated the analytical techniques employed to bolster their weaknesses.?Each program will be briefly explained, followed by a summative discussion of how the proposed program will avoid these problems.?In the first example, Visser and Schoeman (2004) implemented a community intervention to reduce adolescent risks from getting HIV.?The design of their study was pre-test, post-test with no control/comparison group.?In brief, the authors trained teachers from a number of different educational districts during eight two hour workshops in the afternoon after school. The training dealt with the three themes: information about sexuality and HIV/AIDS, life skills, and development of positive attitudes about people who have HIV/AIDS. After the training, the trained teachers were expected to develop a context-specific intervention for their respective schools according to the needs of the learners and the values of the community, using the material provided; implement the program for the learners in their schools (n = 667); and act as change agents in the schools by involving the other teachers, the principal, and parents in a change process to integrate HIV/AIDS education and prevention into the school curriculum.?The results showed that limited change occurred in the school system, but the learners gained knowledge about HIV/AIDS?(p < .001) and protective behavior (p < .01) and their attitudes toward people with HIV were more positive (p < .01). There were no significant changes with regard to learners’ psychological well-being and feelings of personal control.?Against all expectations there was an increase in the reported high-risk behavior of the learners over the period of a year.?More learners gained sexual experienced (33 percent at pre-test and 43 percent at post-test) likely due to maturation into an age of experimentation, but there were no meaningful changes in their condom use and involvement with multiple sexual partners.?The differences between the pre- and post-test means and frequencies were small and of little practical value.?Also the multiple-measures design of the study did not unequivocally lead to the conclusion that the changes that occurred were caused by the intervention. No control for extraneous factors was built into the design.???

In the second example, Agha and Van Rossem (2004) conducted a school-based peer sexual health intervention on normative beliefs, risk perceptions, and sexual behavior of Zambia adolescents.?The design of their study was quasi-experimental with pre-test and repeated post-test measures and unmatched controls.?The intervention was narrow in scope, focusing on promoting abstinence and condom use among male and female school children. Peer educators consisted of people aged 18 - 22 who were trained by a professional peer education trainer to convey their messages through a mixture of techniques. Because there is considerable confusion about HIV/AIDS in the minds of Zambian adolescents, peer educators were trained to impart correct factual information about HIV prevention and transmission through a single discussion-oriented session lasting 1 hour 45 minutes.?To minimize contamination by mass media AIDS prevention campaigns, boarding schools were selected as the sites for the project as students’ access to mass media information is limited.?The authors stated that students at boarding schools are likely to be representative of Zambian students, although they are likely to be from somewhat wealthier backgrounds and not representative of the majority of Zambian youth.?Five schools were included in the study. Of these five schools, three were randomly assigned to the intervention group and two to the control group.?The sample used in the analysis contained 416 students who were interviewed at baseline and during both follow-up assessments; 162 were in the control group and 254 were in the intervention condition.?There were also significant socio – demographic differences between students in the intervention and control groups. The control group consisted of a higher proportion of males than the intervention group; students in the control group were significantly more likely than students in the intervention group to report that they were sexually active at baseline; on average they were a year older than those in the intervention group; and students in the intervention group tended to come from higher socioeconomic backgrounds than students in the control group.?Their findings showed several positive effects of the peer education program immediately after the intervention (on approval of condom use and intention to use condoms) but these effects wore off within 6 months.?There were delayed effects regarding normative condom use beliefs, but they could not be attributed to the intervention.

In the final example, Brieger and colleagues (2001) conducted a reproductive health education program in Ghana and Nigeria (separated by the country of Togo) in which they used a quasi-experimental design with matched controls and single pre-test/post-test measures.?Grassroots youth serving organizations (YSOs) in both countries adopted a standard peer education intervention to increase knowledge of reproductive health issues. Given local responsibility for project implementation each YSO selected its own target population of youth, either secondary school, postsecondary school or out-of-school settings, and developed their projects through a site-specific mix of educational strategies, using peer education as the foundation.?Association for Reproductive and Family Health staff at each project site trained the YSO staff who in turn trained the peer educators. The peer educators were trained to reach their target populations with information, education and counseling on reproductive health through one-on-one and group activities, to create awareness of available services, and to make appropriate referrals for the services that the peer educators could not provide (e.g., prescriptive contraceptives, sexually transmitted infection diagnosis and treatment, or in-depth counseling). The YSO partners represented a fairly wide geographical and cultural spread within mostly urban communities in the two countries and had a broad range of experience. Some were newly formed grassroots organizations, whereas others were local chapters of well-established organizations with wide networks.?Youth involvement in all aspects of the project was stressed (e.g., selection of peer educators, and program planning, implementation, and evaluation).?It was observed that although out of-school youth had higher scores on reproductive health knowledge and contraceptive self-efficacy than in-school youth, only the in-school intervention youth group displayed significantly higher scores than control students.

To increase the likelihood that the proposed intervention will achieve its intended outcomes a number of design features will be put into place that control for confounds presented in these studies.?First, all facilitators will receive a minimum of 24 hours of training on, and will implement the same evidence-based curriculum.?Second, to obtain complete buy-in and cooperation from schools no less than two –weeks will be spent at the beginning of the program to build personal relationships and set expectations for participation in the study in terms of costs and benefits.?Third, two unannounced fidelity checks will be made at each site to ensure that that component is being implemented as intended.?If not, then a correction can be made to bring the component into compliance.?Fourth, to ensure that any effects detected during analysis can be at least 95 percent attributable to the intervention, communities, schools, and students will be matched as closely as possible on key outcome variables.?Fifth, only primary school students between the ages of 12 and 18 will be included in the intervention.?Sixth, the proposed study is based on sound, empirically – tested social science theory allowing it to focus on, education, skills, self-esteem, and self-efficacy in opposition to programs that are narrow in scope.?Seventh, to ensure there isn’t decay in intervention effects over time dosage will include weekly two-hour sessions and no less than 32 hours of direct work with students.?To instill and inculcate desired knowledge, attitudes, and behaviors all program participants will 1) receive the same intervention, and 2) be heavily involved in their own learning and skill development.?Methods will include but are not limited to: active participation, debate, role-playing, skills training, narratives, games, facts, and exercises, questions and discussion, and both large and small-group activities for males, females, and males and females.?To address social pressures on desired behaviors culturally appropriate skill rehearsal opportunities will be provided.?Eighth, unlike the three studies discussed earlier which suffered from inadequate statistical power due to small sample sizes, the sample size of both comparison and intervention groups will be large enough (234 schools, 120,000 + students) to supply the power necessary to detect any effects that result from the intervention.?Ninth, the only schools included in the study will be public given that boarding schools have a qualitatively different population which likely predisposes them to higher levels of the desired knowledge, attitudes, and behaviors at baseline increasing the likelihood of a ceiling effect.?Last, the proposed program will likely be the only one taking place in Geita District other than HIV/AIDS prevention mass-media campaigns. Unlike the last study mentioned statistical controls will be used to counteract the effects of extra-project messages.

Timeline

Two timelines are presented in Tables 4 and 5 which are in two additional posts due to LinkedIn spacing restrictions; if you would like the complete tables please InMail me.?Table 4 shows the estimated timeline for the program start-up year, and Table 5 shows the estimated timeline for the implementation of program components in years two though five.?These timelines were developed on the fiscal calendar year and not the federal calendar year, but are designed such that adjustments can be made.?The only aspect of the timeline we cannot change is when we deliver the program components to students as we need a four month window during the academic year that is neither at the beginning nor at the end of the school year.?In countries that lie below the Equator the academic year is opposite of that in Western countries.?In Tanzania students begin the first term of the school year in January and go through early June.?The second term begins in mid-July and runs to November. Much of the ground work will take place at the end of the previous program year prior to the start of the next program year.?

Summary of Potential Challenges

• The silence of those in authority.
• Traditions that discourage open discussions of sexuality.
• Overt and covert resistance to change.
• Poverty.
• Women’s low social status.
• Inaccurate and incomplete information on HIV transmission.
• Fears about HIV.
• Stigma.
• Changing female values.
• 4 different ethnic groups.
• Working through multiple layers of bureaucracy.
• Potential for increases in risk behaviors.
• Contamination from mass media campaigns.
• Matching intervention groups with controls.
• Inconsistent school attendance.

Summary of Mitigation Efforts for Potential Challenges

• All facilitators will receive a minimum of 24 hours of training on, and will implement the same evidence-based curriculum.?
• No less than two –weeks will be spent at the beginning of the program to build personal relationships and set expectations in terms of costs and benefits.?
• Two unannounced fidelity checks will be made at each site to ensure that that component is being implemented as intended.?
• Communities, schools, and students will be matched as closely as possible on key outcome variables.?
• Only primary school students between the ages of 12 and 18 will be included in the intervention.?
• The proposed study is based on sound, empirically – tested social science theory.?
• To ensure there isn’t decay in intervention effects over time dosage will include weekly two-hour sessions and no less than 32 hours of direct work with students.?
• The sample size of both comparison and intervention groups will be large enough (234 schools, 120,000 + students) to supply the power necessary to detect any effects that result from the intervention.?
• The only schools included in the study will be public.?
• The proposed program will likely be the only one taking place in Geita District other than HIV/AIDS prevention mass-media campaigns.

References

Agha, S., and Van Rossem, R. (2004). Impact of a school-based peer sexual health intervention on normative beliefs, risk perceptions, and sexual behavior of Zambian adolescents. Journal of Adolescent Health, 34, 441 – 452.

Brieger, W. R., Delano, G. E., Lane, C. G., Oladepo, O., and Oyedrian, K. (2001). West African youth initiative: Outcome of a reproductive health education program. Journal of Adolescent Health, 29, 436 – 446.

Hartwig,, K. A., Engb, E., Daniel, M., Ricketts, T., and Quninne, S. C. (2005). AIDS and ‘‘shared sovereignty’’ in Tanzania from 1987 to 2000: a case study. Social Science and Medicine, 60, 1613–1624.

Hartwig, K. A., Kissiok, S., and Hartwig, C. D. (2006). Church leaders confront HIV/AIDS and stigma: A case study from Tanzania. Journal of Community and Applied Social Psychology, 16, 492–497.