The long-term impact of AIDS-preventive interventions for delinquent and abused adolescents.

Slonim-Nevo, Vered ; Auslander, Wendy F. ; Ozawa, Martha N. 等

In recent years, the rapid increase in AIDS cases among teens and young adults has been noted. There were 8,441 cases of AIDS among persons 13 to 24 years of age through August 1991, and the number increased to 14,127 by September 1993 (Centers for Disease Control, 1991, 1993). Moreover, reports suggest that this is just the tip of the iceberg; thousands more are likely to be infected with human immunodeficiency virus (HIV) and millions more are at risk (Hein, 1992).

Some characteristics of high-risk adolescents are already known - juvenile delinquents, abused and neglected children, and those who run away are at high risk of being infected with AIDS (DiClemente, 1991; Nader, Wexler, & Patterson, 1989; Rotheram-Borus & Koopman, 1991; Slonim-Nevo, Ozawa, & Auslander, 1991). In particular, Nader et al. (1989) reported that compared with public high school students, youth in a detention facility knew less about AIDS, perceived less personal threat of AIDS, felt less confident about preventing AIDS, and reported engagement in high-risk sexual activities. Their behaviors placed them at greater risk compared with those of high school students or gay adolescents. DiClemente (1991) also reported a high prevalence of unsafe sexual behavior among incarcerated youth. Slonim-Nevo et al. (1991) found that delinquent and abused youth in residential facilities were highly knowledgeable about AIDS, but had little positive attitudes toward prevention, and reported engagement in unsafe AIDS-related behaviors. A study by Rotheram-Borus, Becker, and Kaplan (1991) on runaway youth reported the same problem.

These studies clearly show that AIDS prevention programs are needed to change the beliefs and behaviors of troubled youth. A few researchers have conducted controlled programs and have evaluated their effectiveness. An exploratory study by Slonim-Nevo et al. (1991) with troubled youth indicated that a short-term information-only intervention was ineffective in increasing AIDS-related knowledge and reducing high-risk activities. The authors suggest that in addition to information, youth would need to acquire skills for dealing with high-risk situations and negotiating less risky activities with friends or partners. Several recent efforts have been made to prevent or reduce HIV among adolescents through skills training. Rotheram-Borus et al. (1991), instituted a program with runaway youth and found that as the number of intervention sessions increased, condom use increased and engaging in high-risk patterns of sexual behavior decreased at 3- and 6-month follow-up assessments. Jemmott, Jemmott, and Fong (1992) found that male adolescents who attended an AIDS prevention intervention program reported fewer occasions of coitus, fewer coital partners, greater use of condoms, and a lower incidence of anal intercourse than did their counterparts in the control group at a 3-month follow-up. Finally, Auslander et al. (1992) found that a cognitive/behavioral intervention is effective in changing AIDS-related knowledge, attitudes, and intentions to cope with AIDS-risk situations among delinquent and abused adolescents at the post-test assessments.

These findings suggest that AIDS prevention programs for youth at risk of HIV infection are likely to yield short-term changes in attitudes and behaviors. It is not clear, however, if such changes can be maintained over time. Therefore, it is important to follow troubled youth who received an AIDS prevention intervention over longer periods. This study examined whether cognitive/behavioral interventions that produced immediate changes in AIDS-related knowledge, attitudes and intentions among delinquents and abused adolescents (Auslander et al., 1992), are capable of producing long-term benefits assessed at the 9 to 12-months follow-up. Specifically, it was hypothesized that adolescents who participated in an intensive 9-session AIDS prevention program will (a) know more about AIDS, (b) hold more positive attitudes toward prevention, and (c) report engagement in fewer unsafe activities than would their counterparts in a control group at follow-up.

METHOD

Sample

We approached 359 adolescents who were placed in 15 residential centers affiliated with child residential welfare agencies and located within a 100-mile radius of St. Louis, Missouri. Youth had been referred to the centers because of juvenile delinquency, child abuse and neglect, or mental health problems. Participation in the study was voluntary and all but one of the adolescents provided informed consent for participation in the study. Of the 358 who completed pretest assessments, 268 youth remained in the residential centers long enough to complete posttest assessments. Follow-up was conducted 9-12 months after completion of the posttest assessments. Of the 268 who completed the posttest assignments, 218 completed the follow-up (81% of those who completed posttest assessments and 61% of the original sample).

Chi-square analyses and ANOVA were performed to determine any differences in socio-demographic characteristics and experimental (skills, discussion, and control group) between the study sample of 218 youth and those adolescents who were not included in the follow-up (N = 140). The two samples did not differ significantly with regard to race, religiosity, mother's and father's level of education, parents' marital status, number of siblings, and whether they were raised by their biological mother or father. However, a higher proportion of males than females were missing at the follow-up phase, [[Chi].sup.2] (1, N = 332) = 10.13, p [less than] .01; and the missing respondents tended to be older (F = 14.6, p [less than] .001, N = 326), belonged to Division of Youth Services (DYS), [[Chi].sup.2] (1, N = 350) = 26.7, p [less than] .0001, and were in the control or skills training group, [[Chi].sup.2] (2, N = 350) = 11.1, p [less than] .01.

The sample of 218 adolescents ranged in age from 12-18 years (Mean = 14.7 years, SD = 1.6). Fifty-six percent were males and 44% were females. The sample was 46% African-American, and 54% white. Religious backgrounds of the adolescents were 40% Protestant, 14% Catholic, 27% no religious affiliation, and 19% other.

The living arrangements of the youth were diverse; 64% lived with their biological mothers most of their lives, and 36% were raised by other mother figures (e.g., foster mothers, stepmothers); 33% were raised by their biological fathers for most of their lives, and 46% were raised by other men (e.g., foster fathers, stepfathers), and 21% reported that they had no father figure. With regard to parents' educational achievement, 39% of the mothers had a high school education, 36% had some college education, and 25% had less than a high school education; 48% of the fathers had a high school education, 34% had some college education, and 18% had less than a high school education. Adolescents' reports of parents' marital status indicated that 31% were married, 37% were divorced, and 32% were unknown to the youth.

Implementing the Intervention

Fifteen residential centers were randomly assigned to three groups: skills training, discussion-only, and control group. The two experimental groups participated in nine sessions delivered over a 3-week period, with each session lasting 1.5 to 2 hours. The substantive content of the skills training and discussion groups were identical and both were based on cognitive-behavioral principles. However, the methods of delivering the material were different. The skills training participants received technical and social skills through modeling, demonstrations, role plays, and practice as described by Kelly (1982). Participants in the discussion group talked about AIDS prevention using problem-solving techniques based on a cognitive-behavioral framework (for the topics of the sessions see Slonim-Nevo et al., 1993). Adolescents in the experimental groups were divided into same sex/age cohorts of 8 to 10 youths per group. Incentives such as snacks, lotteries, and prizes were given to the adolescents to encourage attendance. Participants in the control group received AIDS education on an individual basis after the study was completed.

Adolescents in the skills training, discussion, and control groups did not differ significantly from one another in terms of age, race, sex, and whether they have ever experienced physical abuse, rape, and sexual intercourse. Significant differences were found among the groups in terms of the agency affiliation of the residential center, with a greater percentage of youth from Division of Youth Services (versus Division of Family Services) in the control groups, [[Chi].sup.2] (2, N = 268) = 11.75, p [less than] .005.

Twenty facilitators, who participated in a minimum of 40 hours of AIDS prevention training, led the skills training and discussion groups (two facilitators per group). Training of the facilitators has been explained elsewhere (Slonim-Nevo et al., 1993). The intervention in both groups was monitored to ensure that the content was delivered to the participants. Review of session checklists and logbooks completed by the facilitators of the discussion and skills training groups indicated that the content detailed in the facilitator's manual had been delivered consistently to the adolescents in the two experimental groups.

Attendance at the sessions was monitored to determine any differences between the discussion and skills training group in participation. Seventy-five percent of the adolescents attended all nine sessions, 15% attended eight sessions, and 10% attended 5-7 sessions. No significant differences were found in the mean number of sessions attended by participants in the skills training and discussion only groups.

Data Collection

Data were collected by graduate students in social work who participated in extensive training in data collection and weekly group supervisory sessions in which problems related to the task were discussed (Slonim-Nevo et al., 1993).

The self-report data were collected in small groups of 5 to 10 adolescents each, divided according to gender and age, within each residential center. The interviewer read the questions for the group, clarifying terms as needed for the adolescents who each completed their own questionnaires. The interviewer was present to ensure the independence of responses. This procedure, a combination of self-report and interview, had been tried first in a pilot group of adolescents, and found to encourage honest responses.

Measures

The outcome variables included measurement of adolescents' knowledge, attitudes, reported intentions to cope with AIDS-risk situations, and behaviors related to AIDS(1). To assess Knowledge About AIDS, a modified version of the AIDS Information Survey was utilized (DiClemente, Boyer, Edwards, & Morales, 1988). Items were added that were specific to adolescents, such as "Teenagers cannot get AIDS," and others were deleted because they had less relevance to the prevention goals and to the population. Internal consistency analysis resulted in a Cronbach's alpha coefficient of .71. A test-retest reliability coefficient after a two-week interval was r = .56. (p [less than] .0001, N = 93), and r = .54 (p [less than] .0001, N = 67) after a four-week interval. Concurrent validity of the scale with a sample of delinquent and abused youth that is different from the cohort of the present study was evaluated by comparisons with the AIDS-Risk Behavior Knowledge Test (Kelly, St. Lawrence, Hood, & Brasfield, 1989), and the correlation was r = .50 (p [less than] .0001, N = 95).

To assess Attitudes for AIDS Prevention, adolescents completed a measure developed for this study. The initial questionnaire had been used previously in research with abused and neglected youth (Slonim-Nevo et al., 1991). The initial 28-item questionnaire was modified for the present study. A principal components analysis resulted in a large number of factors, with each factor explaining a very small percentage of the variance of the items. Thus, we chose to use a total scale score that included all the items. The total questionnaire was subjected to an alpha analysis; five items were dropped because they lowered the alpha. The final 23-item questionnaire revealed a Cronbach's alpha coefficient of .84. The test-retest stability coefficient of the 23-item scale after a four-week interval was r = .82 (p [less than] .0001, N = 65). This final questionnaire assessed attitudes in several areas: condom use, IV drug use, multiple sexual partners, personal susceptibility to AIDS, and self-efficacy. For each item, adolescents responded using a 4-point scale from "Strongly Agree" to "Strongly Disagree."

To assess reported Intentions to Cope with AIDS-Risk Situations, a questionnaire was developed. Adolescents completed the Coping with AIDS-Risk Situations (CARS) questionnaire which included situations or events that pose potential risk for engaging in unsafe sex or drug use. The initial questionnaire consisted of 12 short scenarios developed to measure the youth's intentions to resist peer influence and deal with problems related to the following general areas: unsafe sexual intercourse, intravenous drug use, and pressure to take drugs and drink alcohol. The scenarios were developed by the investigators after extensive pilot work with troubled youth in residential centers that included formal assessments (Slonim-Nevo et al., 1991) and informal discussions with the youth in the centers and with their counselors. Situations were developed that would be realistic for the adolescents' lifestyle and were worded to be easily understood.

A 3-point forced-choice scale was developed and respondents had a choice of 3 possible responses: one correct response that exhibited intention to cope with situations so as to reduce the risk of HIV infection, one incorrect response that indicated the youth intended to engage in unsafe sex or drug use, and one response indicating that the respondent did not know how to deal with the situation.

The 12 items were subjected to a principal components analysis to determine the factor structure of the questionnaire. The analysis resulted in a single factor solution, using 9 of 12 original items. Three items were dropped because they lowered the alpha for the total measure. Internal consistency reliability was computed using Cronbach's alpha coefficient. The 9-item measure yielded a reliability coefficient of .84 (N = 322). A stability coefficient was computed over a four-week interval with data from 57 troubled adolescents. The test-retest reliability of r = .86, p [less than] .0001 over a four-week interval indicated a high degree of stability for this measure.

To assess adolescents' report of engagement in unsafe activities, five types of risky sexual behaviors were examined: experiencing sex with someone the respondent did not know very well, combining sex with alcohol, combining sex with drugs, experiencing vaginal sex without condoms, and experiencing anal sex without condoms. For all of these behaviors, the respondents stated how often they have engaged in each of these activities in the last three months on a scale of "0" to "10 or more." No scale was created from these behaviors, and they were analyzed separately.

Data Analyses

This study used a three (skills, discussion, and control group) by two (pre-follow-up) mixed model factorial design. The main hypotheses of the study were analyzed through Analysis of Covariance (ANCOVA), with the pretests of the outcome variables as covariates. With continuous dependent variables we used general linear models, and with dichotomous dependent variables we used logistic models. No significant condition-by-covariate interactions was found. Given a significant main effect of the experimental conditions, post hoc t-tests were conducted on the adjusted means to determine the differential effects of the three conditions.

RESULTS

Knowledge, Attitude, and Intentions to Cope with AIDS-Risk Situations

Table 1 presents the means in AIDS-related knowledge, attitudes, and intentions to cope with AIDS-risk situations at the follow-up assessments, adjusting for the means at the pre-intervention assessments. It also shows the effect of the experimental group on these outcome variables, controlling for the results at the pre-intervention assessments.

Results indicate that the immediate impact of the intervention on AIDS-related attitudes was not maintained 9 to 12 months later. Regarding AIDS-related knowledge and intentions to cope with AIDS-risk situations, the results are mixed; on the one hand, the impact of group (skills, discussion, or control) on knowledge and coping, controlling for pre-intervention data, was significant only at a .07 level; on the other hand, the differences between the scores of the discussion group's and the control group's participants on the post hoc t-test analysis were significant at a .02 level. Specifically, pairwise comparisons of adjusted means using t-test analysis indicated that the discussion group demonstrated greater knowledge about AIDS than did the control group (t = 2.3, p = .02) (not shown in Table 1). Similarly, pairwise comparisons of adjusted means using t-test analysis showed that the participants in the discussion group demonstrated greater intentions to cope with AIDS-related high-risk situations than did those in the control group (t = 2.3, p = 0.2). The two treatment groups did not significantly differ from each other in their AIDS-related knowledge and coping skills.

Table 1. Adjusted Follow-up Intervention Means by Experimental
Condition on Knowledge, Attitudes, and Intentions to Cope With AIDS
Risk Situations (GLM Regression)

Outcome(1) Skills Discussion Control [F.sup.2] DF P

Knowledge(3) .93 .94 .92 2.7 2 .07
Attitudes(4) 3.09 3.14 3.07 .96 2 .4
Coping(5) 6.97 7.39 6.45 2.7 2 .07

(1) Higher numbers indicate greater knowledge about AIDS, more
positive attitudes towards AIDS prevention, and greater intention to
cope with AIDS-risk situations.

(2) The effect of Group (Skills, Discussion, or Control) on the
outcome variables at the follow-up assessments phase, controlling
for the scores at the pre-intervention assessments phase.

(3) Possible range: 0-1

(4) Possible range: 1-4

(5) Possible range: 0-9

Behavior

Table 2 presents the means in AIDS-related behaviors at the follow-up assessments for each of the study's groups, adjusting for the means at the pre-intervention assessments. It also shows the effect of the experimental condition on these outcome variables, controlling for the results at the pre-intervention assessments. The analyses were conducted [TABULAR DATA FOR TABLE 2 OMITTED] using two regression models: GLM when the behaviors were treated as continuous variables (possible range for each of the behaviors was 0-10 or more); and Logistic when the behaviors were treated as dichotomous variables ("0" if the respondent had not experienced the behavior at all during the last three months, or "1" if the respondent had experienced the behavior at least once).

Clearly, both of the interventions, skills training and discussion, were not effective in reducing engagement in high-risk activities related to AIDS, such as unprotected vaginal and anal sex, and combining sex with alcohol and drugs. For example, the adjusted mean of the frequency of engaging in unprotected vaginal sex in the last three months for the skills group was 1.8; for the discussion group it was 2.0; and for the control group it was 1.6. Similar results were observed regarding all five risky behaviors that were included in the study.

Nine to twelve months after completion of the intervention, the respondents who participated in the skills-training or the discussion groups did not significantly reduce their engagement in these behaviors relative to the respondents in the control group who received no treatment before all the data were collected.

DISCUSSION

This study examined whether cognitive/behavioral interventions that produced immediate changes in AIDS-related knowledge, attitudes, and intentions among delinquents and abused adolescents (Auslander et al., 1992), were capable of producing long-term benefits assessed at 9-12 months follow-up. Results showed that one intervention model, discussion groups, produced a long-term increase in knowledge about AIDS and higher reported capability to cope with high-risk situations related to AIDS. However, both interventions, skills-training and discussion groups, could not produce a long-term reduction in the level of engagement in high-risk behaviors.

There are several explanations for these results. First, because of the high mobility of the adolescents who left the residential centers and often the low motivation of their relatives to assist us in locating them, we lost a substantial number of adolescents at the follow-up phase, which limited our ability to draw more stable conclusions. Future studies may use the suggestions of Gwadz and Rotheram-Borus (1992) for tracking high-risk adolescents longitudinally.

Second, it may be that nine sessions are not sufficient to produce a long-term change. In fact, when working with runaway adolescents, Rotheram-Borus and her colleagues (1991) found that as the number of intervention sessions increased, runaways' reports of consistent condom use increased, and their reports of engaging in high-risk sexual behaviors decreased significantly. Therefore, it may be that a greater number of sessions, as well as booster sessions, are needed in order to produce a lasting impact on the behaviors of highly troubled adolescents.

Another possible explanation for these results is based on the theoretical perspective of life chances or life options (Dahrendorf, 1979; Sherraden, 1991; Weber, 1968; Wilson, 1987). This theory describes how life chances influence social and economic mobility, and how they affect individual behavior. According to this perspective, when individuals have opportunities, such as the ability to obtain adequate housing, higher education, and gainful employment, they internalize them, develop positive expectations about the future, and act accordingly. In contrast, those who have negative expectations make little attempt to preserve and improve their lives. (For a brief summary of the theory and its relationship to adolescents' high-risk behavior, see Slonim-Nevo et al., 1995.) Based on this perspective, it is expected that adolescents who perceive their future as worth protecting will make an attempt to stay away from high-risk AIDS-related activities, while their counterparts who perceive their future as a dead end, will make no such effort. Indeed, in a recent study based on the same sample as this study (Slonim-Nevo et al., 1995) it was found that after controlling for socio-demographic factors, adolescents who had higher educational aspirations were more knowledgeable about AIDS, had better attitudes toward prevention, and reported a higher level of intentions to cope with AIDS-risk situations. Moreover, those respondents who had less trouble in school were more likely to report engaging in AIDS-related risky activities. In other words, adolescents who had more chances, at least in their perception, to obtain a higher education, behaved in a way that did not jeopardize their ability to secure a more promising future, while their counterparts who had no such perception did not make an attempt to avoid the risk of contracting AIDS.

Thus, based on this perspective, cognitive/behavioral interventions that provide knowledge and skills are insufficient to motivate highly troubled adolescents to reduce, in the long run, their engagement in risky activities. Seemingly, social workers, psychologists, or educators who work with troubled youth should try to create a sense of meaning and purpose among troubled adolescents. Otherwise, faced with their current difficult and perhaps hopeless reality, they perceive no reason to forgo "a good time today." Perhaps, AIDS prevention for adolescents who have experienced delinquency, abuse, and the street life, should be provided in the context of more comprehensive care. Such care might include basic assistance such as financial help and educational opportunities, teaching how to plan for the future, avoiding risk, and securing external resources in order to achieve future goals, and family and individual counseling. It is hoped that when troubled adolescents have a sense of hope for their future, created by the care and counseling of professional practitioners, they will have a lasting motivation to preserve their lives.

1 Measures are available from the authors.

REFERENCES

Auslander, W.F., Slonim-Nevo, V., Ozawa, M.N., Shepard, S., & Gehlert, S. (1992). Comparison of skills training, discussion and control groups to prevent AIDS among abused and neglected youth. Poster presented at the VIII International Conference on AIDS/III STD World Congress, Amsterdam, The Netherlands, July 19-24.

Centers for Disease Control. (1991). HIV/AIDS surveillance report, September, 1991, 1-18.

Centers for Disease Control. (1993). HIV/AIDS surveillance report, third quarter edition, October, 1993, 1-19.

Dahrendorf, R. (1979). Life chances: Approaches to social and political theory. Chicago: University of Chicago Press.

DiClemente, R.J., Boyer, C.B., Edwards, S., & Morales, S. (1988). Minorities and AIDS: Knowledge, attitudes, and misconceptions among black and Latino adolescents. American Journal of Public Health, 78, 55-57.

DiClemente, R.J. (1991). Predictors of HIV-preventive sexual behavior in a high-risk adolescent population: The influence of perceived peer norms and sexual communication on incarcerated adolescents' consistent use of condoms. Journal of Adolescent Health, 12, 385-390.

Gwadz, M., Rotheram-Borus, M.J. (1992). Tracking high-risk adolescents longitudinally. AIDS Education and Prevention, Supplement (Fall), 69-82.

Heinz, K. (1992). Adolescents at risk for HIV infection. In R.J. DiClemente (Ed.), Adolescents and AIDS. Newbury Park, CA: Sage.

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

Kelly, J.A. (1982). Social-skills training: A practical guide for interventions. New York: Springer.

Kelly, J.A., St. Lawrence, J.S., Hood, H.V., & Brasfield, T.L. (1989). An objective test of AIDS risk behavior knowledge: Scale development, validation, and norms. Journal of Behavior Therapy and Experimental Psychiatry, 20, 227-234.

Nader, P.R., Wexler, D.B., & Patterson, T.L. (1989). Comparison of beliefs about AIDS among urban, suburban, incarcerated, and gay adolescents. Journal of Adolescent Health Care, 10, 413-418.

Rotheram-Borus, M.J., & Koopman, C. (1991). Sexual risk behaviors, AIDS knowledge, and beliefs about AIDS among runaways. American Journal of Public Health, 81, 208-210.

Rotheram-Borus, M.J., Koopman, C., Haigmere, C., & Davies, M. (1991). Reducing HIV sexual risk behaviors among runaway adolescents. JAMA, 266, 1237-1241.

Rotheram-Borus, M.J., Becker, C.K., & Kaplan, M. (1991). AIDS knowledge and beliefs, and sexual behavior of sexually delinquent and nondelinquent (runaway) adolescents. Journal of Adolescence, 14, 229-244.

Sherraden, M. (1991). Assets and the poor. Armonk, NY: Sharpe, Inc.

Slonim-Nevo, V., Qzawa, M.N., & Auslander, W.F. (1991). Knowledge, attitudes and behaviors related to AIDS among youth in residential centers: Results from an exploratory study. Journal of Adolescence, 14, 17-33.

Slonim-Nevo, V., Auslander, W.F., Ozawa, M.N., & Gehlert. S. (1993). AIDS prevention for adolescents at risk: Training social work students as leaders. Journal of Teaching in Social Work, 7, 89-106.

Slonim-Nevo, V., Auslander, W.F., & Qzawa, M.N. (1995). Educational options and AIDS-related behaviors among troubled adolescents. Journal of Pediatric Psychology, 20(1), 41-60.

Weber, M. (1968). Economy and society (two volumes). R. Guenther, & W. Claus (Eds.). Berkeley: University of California Press (Translated from the 4th edition, 1956).

Wilson, W.J. (1987). The truly disadvantaged: The inner city, the underclass, and public policy. Chicago: University of Chicago Press.

This research was supported by National Institute of Mental Health grant # MH45306. The authors thank Sarah Gilbert and Suzanne Shepard for their assistance in data collection.

Wendy F. Auslander, Ph.D., Associate Professor of Social Work; Martha N. Ozawa, Ph.D., Bettie Bofinger Brown Professor of Social Policy; Kenneth G. Jung, M.A., Statistical Consultant, George Warren Brown School of Social Work, Washington University, St. Louis, Missouri 63130.

Reprint requests to Vered Slonim-Nevo, DSW, Senior Lecturer, The Spitzer Department of Social Work, Ben Gurion University in the Negev, P.O.B. 653, Beer-Sheva 84105, Israel.