What Individual-, Group-, or Community-Based Methods of Intervention Reduce Behavioral Risks? What Are the Benefits and Risks of These Procedures?

When we consider the available knowledge from the entire body of literature, we can reach a clear conclusion: Prevention programs significantly reduce HIV risk behaviors. This is true across a variety of risk behaviors and in a variety of populations at risk.

Do Prevention Programs Reduce Behavioral Risk?

Experts in the field have used different designs for evaluating prevention programs. The most rigorous design used in some areas of research, the randomized controlled trial, has been used in HIV prevention research but is more appropriate for testing some questions than others. For example, evaluating the effects of legislative changes would rarely be possible with randomized research. To draw its conclusions, the panel examined the body of literature in a given area by considering all existing approaches to research, the strength of a given design for addressing a specific question, the number and strength of existing studies, and the convergence of effects.

Men Who Have Sex With Men

Considerable research has focused on risk reduction in men who have sex with men. Descriptive studies and nonrandomized studies with control groups show positive behavioral effects, as do randomized studies. The studies with random assignment to groups are clustered in two areas: individual interventions delivered in small group settings and programs aimed at changing community norms (e.g., using peer leaders in community settings to deliver programs). These intervention programs focus on information, skills building, self-management, problemsolving, and psychological factors such as self-efficacy and intentions. Studies with clearly defined interventions, retention of samples to allow followup periods as long as 18 months, and reasonable sample sizes show substantial effects for intervention over minimal intervention or control conditions. More intensive interventions (e.g., more sessions) boost efficacy.

Heterosexual Transmission

Adult Women at Risk from Sexual Transmission

Data from a variety of settings demonstrate the ability to prevent HIV risk behaviors in women. A randomized trial involving a cognitive behavioral intervention aimed at inner-city women with high risk of acquiring HIV through heterosexual contact provides some of the strongest evidence of impact. Three months after intervention, women in the intervention reported a doubling of condom use from 26 percent to 56 percent for all intercourse occasions; no such change occurred for women in the comparison group. A second randomized trial, targeted at pregnant women, shows similar results at a 6-month followup. Results from a third randomized study yet to be published show reductions in unprotected sex and sexually transmitted diseases. A study in rural Tanzania involving treatment for sexually transmitted diseases, condom distribution, and health education found more than a 50 percent reduction in HIV seroconversion incidence over a 2-year period in women ages 15­24. Seroconversion also diminished in counseling programs for women attending a clinic in Kigali, Rwanda, and for sex workers in Bombay, India.

Couples

There is evidence that consistent and correct condom use reduces HIV seroconversion to nearly zero in both male and female heterosexual partners. Counseling of couples in a European study was associated with large increases in protected sexual behavior.

Adolescents

The strongest support for reductions in a broad array of risky sexual behaviors comes from rigorous studies. Five randomized controlled trials used cognitive and behavioral skills training and targeted male and female, African-American, Latino, and European-American adolescents in health clinics and inner-city schools. Studies varied in sample size, and followups were limited to 1 year or less, but results were consistently positive, with outcomes such as condom acquisition, condom use, and reduced number of partners.

Injecting Drug Users

Prevention for injecting drug users has involved drug abuse treatment in some cases, and outreach focused on both drug use and HIV risk behavior in others. Both approaches have been effective. Programs aimed specifically at treating drug abuse show positive effects on risk behavior and have the additional benefit of affecting drug use. These have shown minimal effects on high-risk sex. Community studies training outreach workers or using an educational media campaign to reduce the use of nonsterile needles show increased protected sexual behavior and slowing of seroconversion rates, along with impressive reductions in drug use.

Needle Exchange Programs

An impressive body of evidence suggests powerful effects from needle exchange programs. The number of studies showing beneficial effects on behaviors such as needle sharing greatly outnumber those showing no effects. There is no longer doubt that these programs work, yet there is a striking disjunction between what science dictates and what policy delivers. Data are available to address three central concerns:

1. Does needle exchange promote drug use? A preponderance of evidence shows either no change or decreased drug use. The scattered cases showing increased drug use should be investigated to discover the conditions under which negative effects might occur, but these can in no way detract from the importance of needle exchange programs. Additionally, individuals in areas with needle exchange programs have increased likelihood of entering drug treatment programs.

2. Do programs encourage non­drug users, particularly youth, to use drugs? On the basis of such measures as hospitalizations for drug overdoses, there is no evidence that community norms change in favor of drug use or that more people begin using drugs. In Amsterdam and New Haven, for example, no increases in new drug users were reported after introduction of a needle exchange program.

3. Do programs increase the number of discarded needles in the community? In the majority of studies, there was no increase in used needles discarded in public places.

There are just over 100 needle exchange programs in the United States, compared with more than 2,000 in Australia, a country with less than 10 percent of the U.S. population. Can the opposition to needle exchange programs in the United States be justified on scientific grounds? Our answer is simple and emphatic-no. Studies show reduction in risk behavior as high as 80 percent in injecting drug users, with estimates of a 30 percent or greater reduction of HIV. The cost of such programs is relatively low. Needle exchange programs should be implemented at once.

Policy and Large-Scale Interventions

As in other areas (e.g., smoking, injury control), policy interventions can remove barriers to protective behavior. In the United States and other countries, such interventions have resulted in dramatic reductions in risk behavior. In Connecticut, for example, a single legislative action legalizing over-the-counter purchase of sterile injection equipment led to an immediate and profound reduction in the sharing of nonsterile needles. A national campaign in Switzerland to promote the use of condoms dramatically reduced risky sexual behavior. Regulations on the use of condoms by sex workers in Thailand also led to fewer unprotected sex acts. The results thus far have been impressive. Given the potential benefit of policy changes, these should be implemented as local circumstances allow and, once implemented, should be evaluated as often and thoroughly as possible.

Issues in Need of Further Work

Populations and Settings

A promising start has been made to reduce risk in persons often marginalized. Homeless, chronically mentally ill, runaway, incarcerated, HIV-positive, and physically and developmentally challenged persons face obstacles that affect their ability to initiate and maintain behavior change. In addition, little is known about the risk behaviors of lesbians and bisexual women, heterosexual men, persons over 50 years old, and sexually active youth.

African-American and Latino communities experience disproportionate infection. The application of culturally appropriate strategies demands ethnographic research to understand values, attitudes, behaviors, and factors such as socioeconomic status in different communities. Cultural factors may affect the ability of individuals to change behavior. Researchers from different ethnic or cultural backgrounds may help address this issue. Language and cultural barriers to delivery of interventions must be addressed, with special consideration for individuals whose physical or other impairments limit access to most programs.

Prenatal care and sexually transmitted disease clinics are proven to be effective settings for delivery of HIV intervention. Further research is needed in these and other medical settings. In addition, individuals in institutions such as prisons and mental health facilities, and those in remote areas, require special attention.

Maintenance, Generalizability, and Theory

Understanding and evaluating the maintenance of behavior change requires multivariate, longitudinal studies. In this way, changes in patterns of behavior and causal associations can be estimated. Long-term followup of subjects is necessary. Similarly, more attention to generalizability is needed. An intervention proven effective in one city may not be applicable in another city with a similar population but with different community norms. Methodological issues in need of additional attention include research strategies that measure and enhance validity of self-report, standardization of risk behavior questions and questioning techniques, comparability of intervention conditions across different studies, examining participants and nonresponders to an intervention, and measuring changes in multiple risk profiles over time.

A developmental framework may be helpful for considering the origins of HIV risk behavior. Efforts are needed to incorporate knowledge of childhood antecedents of HIV risk behaviors in adolescents. Can early intervention that alters these antecedents reduce or delay HIV risk behaviors? The body of research now being done to reduce already existing risk behaviors such as unprotected sex and drug use needs to be linked with other research traditions that target antecedents of HIV risk behaviors.

Impact and Cost-Effectiveness

Reviews on HIV prevention conclude that programs produce significant effects, but a statistical advantage may not necessarily equate to meaningful change. An example comes from a study on condom use in more than 13,000 injecting drug users. An intervention nearly doubled consistent condom use, from a baseline level of 10 percent to 19 percent. Although the change was significant from a public health perspective, 81 percent of this high-risk population still engaged in high-risk sexual behavior. This highlights the importance of examining and improving impact as well as assessing statistical significance. Impact is assessed by understanding the efficacy of an intervention, the magnitude of behavior change, and the influence of this change on seroconversion.

A key issue is the degree to which the field has confronted the issue of efficacy (impact of interventions in controlled circumstances) versus effectiveness (effects in real-world setting). Little effectiveness research has been done. This limits the ability to estimate the impact likely to occur if the current generation of risk-reduction strategies, proven useful in efficacy trials, were applied on a large scale outside the research setting. The panel concluded that HIV prevention research is mature enough that some, but not all interventions, are ready for tests of effectiveness. This will require different research strategies and the involvement of professionals from additional disciplines beyond those used for efficacy trials.

The cost-effectiveness of interventions is an important issue in decisions about resource allocation. Research thus far has been positive, but more research is needed to examine the costs and benefits of HIV risk prevention programs.

Behavioral Issues Arising From Biomedical Advances

Important advances in medicine have created new and pressing behavioral issues. Pharmacologic treatment of HIV-positive individuals may increase longevity, but it is not known how such successfully treated individuals will alter their recreational drug use or sexual behavior. Complicated medical regimens raise issues of adherence, with the possibility that incomplete adherence will lead to resistant strains of the virus. Studies of biochemical preventive treatment after sexual exposure to HIV raise questions about risk-reduction counseling. For example, will individuals feel free to engage in risky sex as post-exposure treatment becomes more an option?

Pharmacologic treatment profiles now exist to reduce transmission of HIV from mother to newborn child. This demonstrated preventive intervention offers new opportunities to study behavioral issues and barriers to access in a new and important context.

Policy

Current evidence suggests that some of the most powerful positive effects on HIV risk behavior have been produced by legislative and regulatory changes. One need look no further than to the experience in Connecticut, where one legislative action permitting the purchase of sterile injection equipment had an immediate and pronounced effect on behavior. Here we see the potentially low cost and high effectiveness of intervention at the policy level. Policymaking can be conceptualized as behavior, and as such can and should be studied. Social policy, legal change, and community mobilization are powerful means of intervention and must be a legitimate area of inquiry at the National Institutes of Health and the Centers for Disease Control and Prevention.

Several examples beyond the Connecticut experience show the power of policy changes. Australia, for instance, has a low rate of HIV despite population profiles in some areas similar to profiles in areas in the United States that have high HIV seroconversion rates. Cities such as Tacoma, Toronto, Sydney, Glasgow, and Lund have kept the HIV infection rate low, coincident with policies making sterile needles available for injecting drug users, boosting education aimed at risk reduction, making condoms more available, and enhancing programs for the treatment of sexually transmitted diseases. Impressive results have been reported from around the world on government action to reduce risk and infection in many populations at risk.

Little qualitative and quantitative research has been done in HIV prevention policy, and no body of evidence exists to inform the field about the factors that influence policy, where policy intervention is most likely to be effective, and how best to encourage policy and legislative changes. We believe that funding should be devoted to the study of policy and legislative changes and that National, State, and local levels be considered.

Of utmost importance is that HIV prevention policy be based, whenever possible, on scientific information. This occurs too little-the behavior placing the public health at greatest risk may be occurring in legislative and other decisionmaking bodies. The Federal ban on funding for needle exchange programs as well as restrictions on selling injection equipment are absolutely contraindicated and erect formidable barriers to implementing what is known to be effective. Many thousands of unnecessary deaths will occur as a result.

The single greatest increase in HIV prevention funding occurred with 1996 Federal legislation in the United States providing $50 million within block grant entitlements for programs teaching adolescents abstinence from sexual behavior. Among the criteria for programs funded through the block grant program are the following two requirements: (1) "has as its exclusive purpose, teaching the social psychological, and health gains to be realized by abstaining from sexual activity" and (2) "teaches that a mutually faithful monogamous relationship in the context of marriage is the expected standard of human sexual activity" (Public Health Service Act, Public Law 104-193, Sec. 912). Some programs based on an abstinence model propose that approaches such as the use of condoms are ineffective. This model places policy in direct conflict with science because it ignores overwhelming evidence that other programs are effective. Abstinence-only programs cannot be justified in the face of effective programs and given the fact that we face an international emergency in the AIDS epidemic.

Another instance of policy conflicting with knowledge is in providing treatment for drug abuse. Research shows that treatment of drug abusers with methadone maintenance, outpatient drug-free treatments, residential treatment, or detoxification not only decreases drug use but has a substantial effect on risk behaviors (use of shared needles and protected sex). At the same time that this knowledge has reached a critical mass, funding of drug treatment programs has been reduced in many localities. This tragic trend must be reversed.

Policy and legislative change can have rapid, powerful, and positive results. This key area of the field has been given little attention, a problem that needs remedy. A coordinated effort is needed, and the Government must take strong and immediate steps to protect its citizens. Drawing together legal and policy changes and program implementation occurring at international, National, and local levels offers great promise. Strong political leadership is necessary to direct this effort. The United States has much to learn from other countries where political leaders have taken this issue seriously and, by supporting vigorous prevention strategies, have prevented even more tragedy from occurring from AIDS.

Back | Next
Table of Contents