Effective targeted and community HIV/STD prevention programs.

Targeting interventions to change sexual behaviors which cause sexually transmissible diseases (STDs) at individuals in large numbers usually involves community-level interventions, and such community-level interventions utilize both individual-level and community-level theoretical approaches. Renewed interest in the development of community interventions in the United States and Western Europe has occurred since the spread of the Human Immunodeficiency Virus (HIV), the causative organism of Acquired Immunodeficiency Syndrome (AIDS). In contrast to face-to-face programs that involve work with individuals, community-level HIV approaches attempt to bring about reductions in the level of risk behavior within entire populations or particular population subsets. To achieve their objective of promoting population-level risk behavior reduction, community interventions frequently attempt to bring about changes in safer sex knowledge, attitudes, intentions, and peer norms among members of the entire target population. The definition of community, however, whether defined geographically or subculturally, and definitions of peers need to be clarified, since often these terms are used without clear definition.

COMMUNITY INTERVENTIONS.

There are several major community interventions intended to reduce STDs and increase contraception. Arnold and Cogswell (1971) distributed condoms in an eastern United States city in 1970 to young men between the ages of 12 and 24 over a period of 3 months. The investigators chose nine sites for free distribution of condoms, including five barber-shops, two grocery stores, a pool hall, and a restaurant. Measures, based on a random time sample and a brief questionnaire handed out to condom recipients at randomly chosen distribution sites included the numbers of men who had used a condom in the past week, and the numbers who had never used a condom. After 4 weeks, 60% of recipients said they had used a condom in the past week (< 20% at baseline). No recipients had not used a condom by weeks 12 and 13. Eighty percent of condom recipients lived within six blocks, and half within three blocks, of the distribution sites, and about one quarter said that their choice to use condoms was influenced by other boys, or by girls. While there is some question whether this program instigated condom use or prompted users to shift from commercial sources, the rise in those reporting that they had used a condom previously (from 62% to 100%) is highly significant. In the test period, over 18,000 condoms were distributed from the nine sites, and those who had used a condom on their last sexual intercourse increased from 20% to 91%. Of particular importance is the fact that the consumer acceptability of the distributors (small local businesses) and the enthusiasm of the shopkeepers provided a community control component which Arnold and Cogswell believe enhanced the overall effectiveness and efficiency of the service. These data indicate both that condoms are acceptable to adolescents in a magnitude not previously appreciated and lead to an impressive increase in protected sex, and that small neighborhood commercial outlets have a potentially important role in contraception/STD prevention.

In France, however, a randomized trial to reduce genital discharge in matched counties with an average size of nearly 1 million population in both women and men showed no significant results after 5 months. Genital discharge is usually caused by sexually transmissible infections (STIs) such as chlamydia and gonorrhea. The prevention campaign (Job-Spira, Meyer, Bouvet, Janaud, & Spria, 1988; Meyer, Job-Spira, Bouyer, Bouvet, & Spira, 1991) consisted of installation of automatic condom vending machines and free condom distribution by health centers, university medical centers, and youth clubs. Public information messages were developed specifically for the study: advertisements on local radio and in newspapers; debates; a free hotline with recorded message; posters in medical offices and public areas; a videotape used to introduce debates; a 2-minute cartoon shown during advertisements in local cinemas; a pamphlet distributed through family physicians, pharmacists, family planning centers, and school health services; and three different strip cartoon books distributed through youth clubs, schools, and cinemas. Health professionals' education consisted of a pamphlet and a technical guide widely circulated in the experimental areas, along with an audio cassette specifically developed for family physicians. Finally, conference debates bringing together family physicians, pharmacists, and other health professionals were organized. The main message focused on chlamydia and the need for treatment of all the partners of infected patients.

Before the study, 100 family physicians per area were randomly drawn and invited to participate. Participation rates were similar across the six areas. Each physician had to fill out an anonymous questionnaire for all women with genital discharge and all men with urethral discharge. While the data showed some decrease in cases of genital discharge, this was not significant between the two areas of each pair. Each family physician saw an average of 2.6 cases of discharge in the 4 weeks of the post-intervention evaluation. In women, those presenting following the campaign were younger and presented earlier: There were no changes among men. While this community campaign over 5 months did not achieve significant results, this may have been due both to the dependent variable (genital discharge) and its low incidence, and to differences between the areas selected. Further, the focus on health practitioners and treatment for symptomatic disease in conjunction with population prevention may have undermined the prevention message. Nevertheless, younger women did respond to symptoms earlier.

The HIV prevention projects implemented by Kelly et al. (1991, 1992) targeting men who have sex with men (MSMs) provided convincing evidence that theory-driven interventions can reduce unsafe sexual behavior. These earliest published examples of controlled community-level HIV prevention research interventions directed toward gay and bisexual men were reported by Kelly et al. (1991, 1992), and involved interventions undertaken with men patronizing gay bars in three small cities in the southern United States. Men entering bars were surveyed concerning their sexual practices in the past 2 months to establish baseline levels of risk in each city. Following baseline data collection on population risk characteristics, an intervention was initiated in one of the three cities, with the other two cities serving as controls. The intervention, based on Rogers' diffusion of innovation theory (Rogers, 1983), relied on bartenders and other key informants to observe and identify persons who appeared to be opinion leaders within their social network in the bars. Diffusion of innovation theory postulates that opinion leaders exist in all community populations and, because they are popular and well-liked, these opinion leaders can model and endorse behavior standards that diffuse through the population and create new peer norms. Opinion leaders were defined as persons who were popular, were well-liked, and who frequently interacted with MSM in the bars.

Persons nominated by bartenders and others as key opinion leaders were then contacted; told that, because of their popularity, they could help others make behavior changes to reduce the threat of AIDS in their community; and invited to attend a series of four group meetings to learn how to communicate effective HIV prevention conversational messages to others with whom they interacted. Between 8% and 15% of the persons present in a bar on a typical weekend were opinion leaders recruited for the training.

In the four-session program, key opinion leaders were taught characteristics of effective health communication messages, were taught how to have conversations in which they recommended and personally endorsed the desirability of making specific behavior changes to reduce HIV risk, and were asked to practice role plays of conversations they could have with friends or acquaintances. Following each weekly training session, each opinion leader agreed to have conversations with between 4 and 10 men, and outcomes of the conversations were always reviewed in the next week's group session. In effect, the intervention recruited and engaged persons already popular in their social networks to become active and visible risk reduction behavior change advocates to their own friends and acquaintances.

An initial study examined the impact of the intervention by repeating risk behavior surveys of all men entering the intervention city's bars 3 and 6 months following the intervention, and compared population behavior changes with the behavior found among men in the two control cities surveyed at the same points (Kelly et al., 1991). This initial study revealed that the percentage of male bar patrons in the intervention city who reported engaging in any unprotected anal intercourse in the past 2 months declined from 37% to 28%, a reduction of 25% from baseline levels. There was a 30% decrease in the percentage of men reporting any unprotected receptive anal intercourse. No change was found over time in the risk behavior patterns of men in the control city bars. In a follow-up study, the intervention was then extended to each of the two former control cities (Kelly et al., 1992). Following the points when the intervention was implemented in each of these cities, risk behavior surveys were repeated, with results that replicated the population risk behavior shifts found in the original intervention city.

More recently, a larger-scale trial of the same intervention was completed in eight small U.S. cities, four of which received the intervention and four of which were control cities (Kelly et al., 1997). This study extended the earlier research because it included more cities in different regions of the country, determined population risk behavior levels 1 year following intervention, compared the effects of the popular opinion leader intervention against a traditional AIDS education campaign undertaken in the control cities, and obtained data on condom-taking patterns in study city bars to corroborate population members' risk behavior change self-reports. Between baseline and 1-year follow-up, significant reductions were found in mean frequency of unprotected anal intercourse episodes reported by men in the intervention cities (from 1.7 occurrences in the past 2 months at baseline to 0.6 occurrences at follow-up), and the percentage of population members' anal intercourse occasions protected by condoms increased from 45% to 67%. Rates of condom taking from free dispensers in intervention city bars increased by 65%. No significant changes occurred in the study's four control cities over the same period of time (Kelly et al., 1997). Kelly et al. (1997) also raise the possibility that the impact of interventions may decline over time, and that periodical booster sessions may be required to maintain the impact of interventions, given that their effect may decline over time.

Kegeles, Hays, and Coates (1996) have also examined the impact of a community-level risk reduction intervention focused on young gay men. Two small cities were studied in this trial, one of which received an intervention and one of which served as a control. In the intervention city, a core group of about 15 men were recruited to provide leadership to a program that entailed three components: (a) peer outreach in which safer sex messages were diffused to others in a manner similar to that used in the Kelly et al. (1991, 1992, 1997) projects just described; (b) small-group, 3-hour risk reduction workshops; and (c) a safer sex educational material distribution focused in venues that served young gay men in the community. The entire program lasted for 8 months. The Kegeles et al. study was evaluated by examining changes in risk behavior of cohorts of men longitudinally followed from baseline to a 1-year postintervention. Reductions were found in the proportion of intervention-city cohort members who reported any unprotected anal intercourse in the past 2 months, from 41% at baseline to 30% at the follow-up point. Changes in behavior were found for men who had sex in both primary and nonprimary relationships, although rates of unprotected anal intercourse declined more among men with casual partners than among men who had steady boyfriends.

The Centers for Disease Control (CDC, 1999) AIDS community-level HIV intervention demonstration project was carried out in five U.S. cities using 10 intervention-comparison community pairs. The intervention was based on the use of small media and role model stories (Bandura, 1986; Corby, Enguidanos, & Kay, 1996), and the mobilization of community members to distribute and verbally reinforce prevention. The small media materials featured theory-based HIV/STD prevention messages and were distributed along with condom and bleach kits. The small media (community newsletters, pamphlets, baseball cards) contained authentic stories about people from the community who were changing or preparing to change their HIV risk behaviors, and based on the stages of change of the transtheoretical model (Prochaska, DiClemente, & Norcross, 1992; Prochaska & Velicer, 1997). These stages are precontemplation, contemplation, action, maintenance, and sometimes relapse. Each role model story described the model's progress toward consistent practice of safer sex or safer injection drug use. Data indicated that exposure to the intervention increased from 5% in month 2 to 54% in month 27. There were statistically significant increases in condom carriage and stages of change scores for condom use for both main and nonmain sexual partners, not just in the individuals reached by the intervention but in the communities as a whole, suggesting that the intervention had diffused. With nonmain partners, the increase was largely due to people moving to consistent condom use (i.e., into action and maintenance stages). In the intervention communities, the proportion of individuals reporting consistent condom use with main partners doubled from 8% at baseline to 17% at the last wave of data collection. It is clear that this community intervention can have a significant public health impact at both individual and community levels.

The AIDS Prevention for Pediatric Life Enrichment (APPLE) project (Santelli et al., 1995), a community-based program to prevent perinatal HIV infection by preventing infection in women, tested a primary prevention model developed from principles of cognitive social learning theory which used street outreach and community-targeted small media materials to increase the use of condoms. The program was evaluated using a two-community, time series, quasiexperimental design with annual street surveys. Norms reflecting social acceptability of condoms were more positive among women in the intervention community, and although condom use at last sexual encounter rose in both communities, it was significantly higher in the intervention community, and was higher among women exposed to either small media or small media and street outreach. Other self-reported HIV-prevention behaviors did not show change in the initial period.

Community-level programs have shown success in promoting norm and risk behavior changes within larger gay community populations. Although the magnitude of risk reduction produced by community-level interventions is often somewhat smaller than that produced by the very intensive face-to-face programs described earlier, community-level approaches can reach a much greater number of persons. By creating peer norms and supports favoring safer sex, community interventions may be able to help persons maintain behavior change better than approaches that work with the individuals in isolation but do not change peer group norms. In population segments where safer sex is not yet an accepted peer norm, community approaches can have considerable potential impact. Further, the interventions reviewed here all engaged members of the target population to deliver intervention activities to their peers. Such approaches are likely to be culturally appropriate in the community and may be more sustainable than HIV preventions carried out by external agents.

MEDICAL SERVICE-BASED INTERVENTIONS.

Outside of the gay community, community-based programs have also used changes in medical infrastructure to influence STI and HIV rates. In East Africa, the Mwanza study was designed to test syndromic management of STD in rural communities in Tanzania (Grosskurth et al., 1995). The design matched six pairs of villages randomly assigned as intervention or control communities and undertook a 2-year intervention which consisted of training staff in the treatment of STDs, including provision of the appropriate drugs and dose supervision. After 2 years, HIV seroconversion rates were 1.2% in the intervention community and 1.9% in the control community. After correcting for baseline prevalence, this represents a 38% decrease in HIV incidence. This program illustrated that provision of the infrastructure and treatment facilities for STDs itself will lead to reductions in STDs and subsequently HIV.

The Mwanza study is usually compared with a similar trial in Rakai, Uganda. Both trials included STD treatment and behavioral interventions as essential components of the trials, and each used a community trial format. Results were found to be different in each of the trials. In Mwanza there was a significant reduction in the incidence of HIV after the trial was implemented. In Rakai, there was no significant difference in HIV incidence, and the trial was stopped after 3 rounds. Grosskurth, Gray, Hayes, Mabey, and Wawer (2000) provide an explanation for these contradictory results. The Mwanza trial (1991-1994) compared six pairs of communities, one of which received the intervention. To compare the results, a cohort of 1000 randomly selected adults was followed for 2 years. For ethical reasons, any individual who tested positive for STDs was treated. After adjustment for confounding factors, incidence decreased by 38% after the intervention. Sexual behavior showed no difference between the two groups. In Rakai (1994-1998), the trial tested the hypothesis that repeated rounds of mass treatment for STDs would lead to reduced incidence of HIV. The trial was implemented in 56 communities, grouped into 10 clusters, of which 5 were selected to receive the intervention. Individuals in the intervention communities were tested and treated every 10 months. Effects were measured for entire communities, where everyone from ages 15 to 59 in the study communities were tested every 10 months. Although there was no difference in HIV incidence, syphilis was 20% lower and trichomoniasis 41% lower in the treatment communities. The authors suggest that the difference in results was due to the different stage of the epidemic in the two areas. Mwanza was at a lower stage (approximately 1% prevalence) compared to Rakai, which had a stable, mature epidemic (16% prevalence rate). The authors hypothesize that high levels of ulcerative STDs may be more necessary early in the HIV epidemic. After HIV prevalence rates reach a threshold level, there is no longer a cofactor effect of STDs as HIV infections are much more likely to occur among stable sexual partners with repeated exposures to the virus. Rakai also had higher rates of untreatable herpes, another ulcerative STD. Furthermore, the trials tested different interventions and used different means of assessing outcome:

   [C]omputer simulations using data modeled on the Mwanza population suggest
   that mass treatment may have an substantial impact in areas with high
   prevalences of treatment STDs, and low or moderate but rising HIV-1
   prevalence. For example, a single round of mass treatment with improved STD
   case-management services was predicted to lead to a rapid, steep, and
   sustained reduction in HIV-1 incidence, reaching 63% after 5 years.

(Grosskurth et al., 2000, p.WA13)

Targeted interventions are specifically aimed at groups at high risk of STD/HIV infection. Groups most commonly targeted include commercial sex workers (CSWs), core groups at high risk of STDs/HIV, MSMs, and migrant workers among others. A particularly successful approach to such groups is the inclusion of peer leaders who work as opinion leaders, peer educators, and role models, following Bandura's (1986) social cognitive theory of health behavior. The approach used by Kelly et al. (1991, 1992) typifies this theoretical avenue. Such an approach has the advantage of making it possible to access many more individuals within the target group and providing educators who are familiar with the culture, language, and conditions of the target group. A good example of such an intervention was the Thai intervention in CSWs (Hanenberg, Rojanapithagakom, Kinasol, & Sokal, 1994). In this study, a government requirement that condoms be made available in all CSW establishments was associated with mass advertising and monitoring of CSWs. Over 2 years, reported condom use in brothels increased from 14% to 94%, and over 5 years there was a 79% decrease in the five major STDs in men. In Kenya and Zimbabwe, Ngugi, Wilson, Sebstad, Plummer, and Moses (1996) used peer educators to distribute condoms and provide safe sex information to CSW women at risk of STDs/HIV. After 1 year in Kenya, condom use had increased from 5% to 37%, and in Zimbabwe from 18% to 66% within 2 years. These data suggest that there is a place for government action in conjunction with conventional mass media and peer education interventions to modify sexual behavior (in this case, condom use).

CONCLUSIONS.

These data identify several important factors for interventions to modify sexual behavior. These factors include: first, the importance of using a theoretical framework to test the effectiveness of theoretical interventions; second, the importance of using rigorous scientific methodologies to compare intervention to control groups; third, the importance of changing social norms by using community opinion leaders, role models, and peer educators as intervention specialists; fourth, engaging the target community in the programs; fifth, the importance of diffusing the changes attributable to the intervention over as wide a range of participants as possible; sixth, the importance of providing booster interventions over time to strengthen the impact of the original intervention, like booster vaccinations; and seventh, that very behaviorally and proportionally significant changes are possible.

REFERENCES

Arnold, C. B., & Cogswell, B. E. (1971). A condom distribution program for adolescents: The findings of a feasibility study. American Journal of Public Health, 6, 1739-1750.

Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, NJ: Prentice Hall.

CDC AIDS Community Demonstration Projects Research Group. (1999). Community-level HIV intervention in 5 cities: Final outcome data from the CDC AIDS community demonstration projects. American Journal of Public Health, 89, 336-345.

Corby, N. H., Enguidanos, S. M., & Kay, L. S. (1996). Development and use of role model stories in a community-level HIV risk reduction intervention. Public Health Reports, 111(Suppl. 1), 54-58.

Grosskurth, H., Gray, R., Hayesm, R., Mabey, D., & Wawer, M. (2000). Control of sexually transmitted disease on the HIV epidemic in rural Tanzania: Baseline survey results. Lancet, 355, WA8-WA14.

Grosskurth, H., Mosher, F., Todd, J., Mwijarubi, E., Klokke, K., Senkoro, P., Mayaud, J., Changaluch, A., Nicoll, J. Ka-Jing, J., Newell, K., Mugeye, K., Mobey, D., & Hayes, R. (1995). Impact of improved treatment of STDs on HIV infection in rural Tanzania: Randomized control trial. Lancet, 346, 530-536.

Hanenberg, R. S., Rojanapithagakorn, W., Kinasol, P., & Sokal, D. C. (1994). Impact of Thailand's HIV control program as indicated by the decline of STDs. Lancet, 344, 243-245.

Job-Spira, N., Meyer, L., Bouvet, E., Janaud, A., & Spira, A. (1988). The prevention of sexually transmitted diseases which affect fertility: Methodological problems and initial results. European Journal of Obstetrics, Gynecology and Reproductive Biology, 27, 157-164.

Kegeles, S. M., Hays, R. B., & Coates, T. J. (1996). The Mpowerment Project: A community-level HIV intervention for young gay men. American Journal of Public Health, 86, 1129-1136.

Kelly, J. A., Murphy, D. A., Sikkema, K. J., McAuliffe, T. L., Roffman, R. A., Solomon, L. J., Winett, R. A., Kalichamn, S. C., & Community HIV Prevention Research Collaborative. (1997). Randomized, controlled community-level HIV prevention intervention for sexual-risk behavior among homosexual men in U.S. cities. The Lancet, 350, 1500-1505.

Kelly, J. A., St. Lawrence, J. S., Diaz, Y. E., Stevenson, L. Y., Hauth, A. C., Brasfield, T. L., Kalichman, S. C., Smitt, J. E., & Andrew, M. E. (1991). HIV risk behavior reduction following intervention with key opinion leaders of a population: An experimental community-level analysis. American Journal of Public Health, 81, 168-171.

Kelly, J. A., St. Lawrence, J. S., Stevenson, L. Y., Hauth, A. C., Kalichman, S. C., Diaz, Y. E., Brasfield, T. L., Koob, J. J., & Morgan, M. G. (1992). Community AIDS/HIV risk reduction: The effects of endorsement by popular people in three cities. American Journal of Public Health, 82, 1483-1489.

Meyer, L., Job-Spira, N., Bouyer, J., Bouvet, E., & Spira, A. (1991). Prevention of sexually transmitted diseases: A randomised community trial. Journal of Epidemiology and Community Health, 45, 152-158.

Ngugi, E. N., Wilson, D., Sebstad, J., Plummer, F. A., & Moses, S. (1996). Focussed peer-mediated educational programs to reduce STD and HIV transmission in Kenya and Zimbabwe. Journal of Infectious Diseases, 1996, 174 (Suppl. 2), S240-247.

Prochaska, J. O., DiClemente, C. C., & Norcross, J. C. (1992). In search of how people change: Applications to addictive behaviors. American Psychologist, 47, 1102-1114.

Prochaska, J., & Velicer, W. (1997). The transtheoretical model of health behavior change. American Journal of Health Promotion, 12, 38-48.

Rogers, E. (1983). Diffusion of innovations. New York: Free Press.

Santelli, J. S., Celentano, D. D., Rozsenich, C., Crump, A.D., Davis, M.V., Polacsek, M., Augustyn, M., Rolf, J., McAlister, A. L., & Burwell, L. (1995). Interim outcomes for a community-based program to prevent perinatal HIV transmission. AIDS Education and Prevention, 7, 210-220.

Michael W. Ross and Mark L. Williams
University of Texas-Houston

Parts of this research were supported by a grant from the Centers for Disease Control to the first author, and a grant from the National Institute on Drug Abuse to the second author. The opinions expressed herein are solely those of the authors.

Address correspondence to Dr. Michael Ross, WHO Center for Health Promotion and Prevention Research, School of Public Health, University of Texas, PO Box 20036, Houston, TX 77225; e-mail: [e-mail].