HIV infection among adolescents is an area of increasing concern in AIDS prevention efforts.  By the end of March, 1993, 55,120 cases of AIDS in individuals aged 20 to 29 were reported to the Centers for Disease Control and Prevention (CDC).  Given that roughly 10 years separate HIV infection and AIDS-related symptoms, most of these individuals became infected as teenagers (Flora & Thoresen, 1988; National Commission on AIDS, 1994).  Significant numbers of adolescents engage in a combination of high risk behaviors which may result in HIV infection.

In order to decrease the likelihood of contracting AIDS, changes in behavior regarding sexual activity and drug use are essential.  A substantial amount of research, however, indicates that achieving such behavior change is difficult.  Fisher and Fisher (1992), in a review of the literature on changing AIDS-risk behavior, found several studies showing improvement in adolescents' knowledge about AIDS and in intentions to practice AIDS prevention following various types of interventions.  They found only three studies, however, which demonstrated AIDS-risk reduction behavior change in adolescents (Jemmott, Jemmott, & Fong, 1992; Rotheram-Borus, Koopman, Haignere, & Davies, 1991; Winett, et al., 1990).

This poster presents results from a study of the effectiveness of the intensive AIDS prevention program developed by Rotheram-Borus et al. (1991) in reducing high risk behaviors in runaway youth. This "Stay Safe" program was modified for use in school, and was taught by the regular teachers and specially trained peer educators, who were themselves trained by a San Joaquin County Public Health Educator.

Participants

The participants were 330 (95 female, M age = 16.3; 235 male, M age = 16.1) students from eight alternative high schools in Central California, two experimental and two control schools each semester. Results of the pretest on this group were reported by Hannon, Hall, Berke and Coronel (1997a), looking primarily at the sex differences, and overall sexual behavior.

A post-test was administered at the end of the intervention, and the follow-up testing was five months later. Participants with both pre-tests and post-tests numbered 212 (122 experimental group and 90 controls). After the 5-month follow-up, there were 123 participants who completed all three tests (76 experimental and 47 control).

Measures

The questionnaire developed for this study first asks 14 True/False AIDS knowledge questions. Next is a series of questions about sexual behavior, condom use, and other risk behaviors in the last 3 months.  The questions in the next two sections of the questionnaire are substantially identical, with the first section referring to regular partners, the next to occasional partners, and the last to future intentions. The Sexual History Scale consists of eight Yes/No questions one should ask a partner about past sexual behavior prior to being sexual with her/him.  The High Risk Behavior Scale consists of six questions on condom use, drug use and multiple partners rated on a 5-point Likert scale from 1 (never) to 5 (every time). The same test was used for all three test periods.

An area of primary concern was the California legal requirement to obtain parental consent for teaching and asking questions about sexual behavior.  Since all students in alternative schools are required to register, with their parents, at a district run registration session, a consent form for parents was included in every district student's packet.  Most parents signed the forms, and 92% of eligible participants either had parental consent or were over 18.

On the pretesting day the students were given instructions for filling out the questionnaire, assured of their anonymity and confidentiality, and informed that they would receive $5 upon completion of the questionnaire.  Before filling out the questionnaire, the students were required to sign a student informed consent form.  Each student was then given a questionnaire with a pre-assigned identification number, and therefore students were not required to put their name on any part of the questionnaire.

Pre-test/Post-test by Experimental/Control Group Analysis.

Categorical variables were analyzed using chi-square tests for group differences on pre-test and on post-test.  Quantitative variables were analyzed using 2 (experimental/ control) X 2 (pre-test/post-test) split-plot factorial ANOVAs.

Analysis of demographic variables showed a significant age difference, F (1, 208) = 4.34, p < .05; with experimental participants slightly older (M age = 16.32 years, SD = 1.36) than control participants (M age = 15.92 years, SD = 1.16).  This is probably due to the inclusion of a school for students about ready to graduate in the experimental group.  Groups did not differ significantly on any other demographic variables (place lived for last 3 months, ethnic background, language spoken most at home, comfort speaking and reading English).

On the AIDS knowledge test (see Table 1), there was a nearly significant interaction between pre/post and groups, F (1, 210) = 3.72, p < .06.  Scores increased significantly for the experimental group, F (1, 121) = 20.25, p < .001; but not for the control group.

Variables describing characteristics of sexual behavior in the past 3 months are shown in Table 1.  Percent of participants who  were sexually active did not differ between experimental and control groups for pre-test or post-test.  Of the sexually active participants, experimental participants had significantly more partners than control participants, F (1, 131) = 6.01, p < .05, both at pre-test and post-test.  Experimentals also had sex significantly more often than controls, F (1, 114) = 4.45, p < .05, both at pre-test and post-test.  Groups did not differ in frequency of condom use.  Table 1 also shows the percentage of participants reporting regular partners and occasional partners.  There were no significant differences between groups or from pre-test to post-test for either measure.

Mean ratings of six items designed to assess high risk behavior using a 5-point Likert scale are shown in Table 2 for future partners.  For regular and occasional partners, neither group changed significantly from pre-test to post-test, so these data are not shown in the Table.  For future partners, controls did not change significantly from pre-test to post-test, while experimentals changed significantly in a more desirable direction on planning to have condoms when needed, talk about condoms before having sex, use condoms during sex, and ask their partners if they were having sex with anyone else.

The mean number of intervention sessions attended by experimental group participants was 14.87 (4.79) out of a total of 20.  Appropriate post-test measures (e.g., number of partners, number of times used condoms, etc.) were correlated with number of intervention sessions attended, and none of the correlations were significant.

 Pre-test/Post-test/5-Month Follow-up by Exp./Control Group Analysis.

Percent of participants who were sexually active did not differ between experimental and control groups for 5-month follow-up-test.  Of the sexually active participants, there were no significant differences between groups for number of partners, number of times had sex, or number of times used condoms.

Mean ratings of items designed to assess high risk behavior using a 5-point Likert scale showed no significant effects for regular or occasional partners.  For ratings regarding future intentions, experimentals showed significantly more desirable ratings than controls on 4 of 6 items, but this was true across all three testing periods and therefore did not show that the intervention was effective for this sample.

Appropriate 5-month follow-up measures (e.g., number of partners, number of times used condoms, etc.) were correlated with number of intervention sessions attended, and only one of the correlations was significant; number of times had sex correlated with sessions at r = -.26, p < .05.

The pre-test/post-test comparison showed positive effects of the intervention on future intentions regarding high risk behaviors, but not on current behavior with either regular or occasional partners. In addition, few 5-month follow-up comparisons indicated that the intervention was effective.

Possible reasons for the limited effectiveness of the intervention include differential interest and enthusiasm on the part of teachers and peer-educators who presented the intervention, prior education regarding safer sex before this study began, and relatively low overall rates of sexual behavior in these students.  The cohort in the Rotheram-Borus study was a more sexually active and drug using population, and we should have had a process for modifying our study to more directly relate to our urban/rural population. In our participants, alcohol seems to be the drug most commonly used. We also had schools that only met for half a day, and the sessions were too long to effectively fit into the demands of the rest of the curriculum. The teachers would have preferred one session a week, which would have been less disruptive and require less intense preparation.

Three of the control schools had AIDS presentations from the AIDS Foundation during this year, and there was a city-wide AIDS awareness day, in which some of our students participated. It is both impossible and unethical to keep the controls from the available AIDS educational activities, and this may account for some of the lack of difference in our study.

Return to Program
Return to Abstract

 Fisher, J. D., & Fisher, W. A. (1992). Changing AIDS-Risk behavior. Psychological Bulletin, 111, 455-474.

 Flora, J. A., & Thoresen, C. E. (1988). Reducing the risk of AIDS in adolescents. American Psychologist, 43, 965-970.

Jemmott III, J. B., Jemmott, L. S., & Fong, G. T. (1992). Reductions in HIV risk-associated sexual behaviors among black male adolescents: Effects of an AIDS prevention intervention. American Journal of Public Health, 82, 372-377.

Hannon, J.R., Hall, D.S., Berke, S., & Coronel, J. (1997a). Aids knowledge and risk behavior in alternative high school students. Poster presented at the 1997 meeting of the Western Psychological Association, Seattle, WA.

Hannon, J.R., Hall, D.S., Bell-Sanford, G., & Tirado, M. (1997b) Final Report: An analysis of HIV/AIDS intervention with high risk youth. UCSF, Center for Aids Prevention Studies, Contract Number 94-21130.

Rotheram-Borus, M. J., Koopman, C., Haignere, C., & Davies, M., (1991). Reducing HIV sexual risk behaviors among runaway adolescents. Journal of the American Medical Association, 266,1237-1241.

Winett, R. A., Anderson, E. S., Moore, J. F., Sikkema, K. J., Hook, R., Webster, D. A., Taylor, C. D., Dalton, J. E., Ollendick, T. H., & Eisler, R. M. (1990). Family/media approach to HIV prevention: Results with a home-based, parent-teen video program. Unpublished manuscript.