Injection drug use is a major risk factor for HIV infection in the United States and numerous other countries. In the United States, injecting drug users (IDU) now account for more than one third of new AIDS cases (nearly double the proportion of 5 years ago), and for a majority of new HIV infections. The Centers for Disease Control and Prevention estimate that about half of the 40,000 annual new infections are among IDU.(1) In addition, 70 to 80% of recent heterosexual AIDS cases attributed to heterosexual transmission are among women, most of whom were infected during sex with a male drug user. Injecting drug users are also the source of a large proportion of pediatric infections.
Specific behaviors associated with drug use that are risk factors for HIV transmission include shared use of drug injection equipment and unprotected vaginal or anal sex with multiple sexual partners.(2,3) For this reason, interventions that can reduce the prevalence of these practices are critical components of a comprehensive AIDS prevention policy.
This chapter reviews evidence of the effectiveness of strategies to prevent the spread of HIV in IDU. These strategies include interventions such as drug treatment, needle exchange, street outreach, educational interventions, and HIV counseling and testing. Also considered are several prevention strategies that have been tried with other populations and diseases other than HIV, several of which show promise of preventing HIV in drug users.
| Drug Treatment|
Increasing access to drug treatment is a frequently recommended approach to slowing the spread of HIV in IDU. In the United States, the National Academy of Sciences has urged that "treatment on demand" be made available to drug users,(4) and the National Institute on Drug Abuse has adopted drug treatment as its principal strategy for preventing HIV and AIDS.(5)
There is now a sizeable and growing literature (6-14) demonstrating that methadone maintenance treatment (MMT), the most widely available treatment for opiate addition, reduces both injection drug use and the risk of infection with HIV. Metzger and colleagues, for example, found a HIV seroconversion rate of only 3.5% over 18 months among methadone maintenance patients, compared with 22% in a comparison group of out-of-treatment opiate addicts.(11) Even when preexisting differences between MMT clients and nonclients are taken into account,(8,10,12) MMT reduces high-risk practices and the likelihood of seroconversion. Methadone's impact, however, appears to be related principally to its role in reducing injection frequency.(7,9,13)
Several studies have shown that methadone treatment increases in effectiveness the longer a patient remains in the treatment program.(8,12,14) Ball and Ross found that injection drug use among methadone maintenance patients dropped from 81% of admission to 42% at 3.5 years of treatment to 29% for patients enrolled 5 or more years.(7) Effectiveness, however, appears to be widely variable, depending both on the characteristics of patient populations and the array of ancillary services available to patients, such as counseling and psychiatric help. In a randomized trial, McLellan and colleagues found the addition of basic counseling and especially professional services were associated with substantial decreases in the proportion of clients testing positive for opiates and other drugs.(11) As a negative example, van Ameijden and colleagues found that "low threshold" methadone had little impact on injection behavior.(14,15) "Low-threshold" refers to the practice of not requiring patients to attend a clinic regularly or remain abstinent from injection drugs. van Ameijden et al. also speculate that the low doses typical of "low threshold" treatment may also have been responsible for the poor outcomes.(15) Several studies (e.g., Ball and Ross(7)) have found relationships between dose and opiate use, and thereby between dose and risk of infection with HIV.
| Needle Exchange|
The first needle exchange was formed by the Amsterdam junkiebond in 1984 in an effort to slow the spread of hepatitis B among IDU.(16) As the number of European AIDS cases increased sharply, needle exchange programs were soon adopted in other European cities. By the early 1990s, programs had been implemented in more than 20 countries, including the United States, Canada, Australia, and most member nations of the European Union.(17) Needle exchange programs developed more slowly in the United States than in Europe, primarily because of fierce political opposition. Nevertheless, at the end of 1996, 104 needle exchange programs were operating in more than 20 U.S. states (David Purchase, North American Syringe Exchange Network, personal communication, November 1996). Only about one half of U.S. needle exchange programs are legal, but others operate with the tacit support of local officials. Nearly all needle exchange programs require one-for-one exchange of syringes.
The effectiveness of needle exchange programs in reducing HIV incidence is not definitely known because of the difficulty associated with conducting rigorous evaluations. Practical and ethical considerations prohibit experimental trials in which IDU are randomized to attend or not attend a needle exchange. Studies of the impact of exchange programs have been limited by the difficulty of assessing client self-selection and by dependence on IDU self-reports. The preponderance of evidence, however, suggests that needle exchange has a beneficial impact on IDU behavior. A 1993 review of needle exchange programs in the United States and abroad,(17) commissioned by the Centers for Disease Control and Prevention, found that of 14 acceptably executed studies, 10 found attendance at a needle exchange to be associated with reduced syringe sharing, and four found no such association. Four additional studies that examined the impact of exchange programs on high-risk sexual behaviors found no clear evidence of beneficial or adverse effect.
Needle exchange programs in the United States serve only a small fraction of the IDU population. Of an estimated 1.1 to 1.5 million IDU in the United States,(18,19) only 15% are in drug treatment at any given time.(19) Between 1988 (when the first comprehensive needle exchange opened in Tacoma, Washington) and 1993, 5.4 million syringes were exchanged nationwide,(20) amounting to only about one syringe per IDU per year. Even in cities with small communities of IDU, such as Tacoma and New Haven, Connecticut, aggressive needle exchange reaches only about half of the estimated number of IDU, and for the most part, on an intermittent basis.
Mathematical modeling of the impact of needle exchange(21-23) suggests that impact is a function of penetration (the proportion of IDU who exchange) and volume (the number of needles exchanged). The effectiveness of needle exchange can be maximized by increasing both its penetration and volume.(24) Penetration is important independent of volume because the greatest benefit in decreased sharing comes with the first needle each client receives, because this event has the greatest impact on syringe circulation time; additional exchanges have decreasing marginal value. Volume depends in part on penetration and is important because each additional exchange further decreases the circulation time of used syringes, decreasing the likelihood that an infected syringe will be used. Using a dynamic mathematical model of needle circulation, researchers in New Haven, CT estimated that as a result of the needle exchange program, the number of times each circulating needle was replaced per year increased from 23 to 44, decreasing the mean circulation time from 2.3 to 1.2 weeks.(23)
Detailed ethnographic mapping of the IDU community may be helpful in identifying sites and times that make it convenient for the majority of IDU to attend. It may also be helpful to staff the exchange with recovering IDU drawn from the communities they serve. Recovering IDU have ethnographic knowledge of the IDU community, and are more likely to be viewed as peers or role models of behavioral change.
Penetration and volume can be achieved in other ways. Des Jarlais and colleagues recommend a social change approach to needle exchange, in which drug users exchange for others in their friendship networks.(25) Because of arrest warrants, parole violations, or other reasons, some IDU may not be willing to attend needle exchange themselves, but may know of others who would exchange for them. Surrogate or "secondary" exchanging can greatly increase the number of IDU effectively served by a needle exchange. A pager-initiated exchange in Sacramento, CA makes home delivery of syringes to more than 50 IDU who regularly exchange an average of 65 syringes for themselves and others.(26) Home delivery is an attractive alternative to fixed-site exchange for communities in which needle exchanges face NIMBY (not-in-my-backyard) opposition. Even where needle exchange has local support, home delivery complements fixed site exchange by making syringes available to IDU unable or unwilling to attend needle exchange.
| Street Outreach|
Educational street outreach has been one of the most successful strategies for slowing the spread of HIV among U.S. drug injectors. Under the National AIDS Demonstration and Research (NADR) projects, the U.S. National Institute on Drug Abuse funded limited and more elaborate forms of outreach in 68 cities. The standard form of the NADR outreach intervention consisted of brief (typically 10 minutes) educational encounters between outreach workers and drug injectors during which basic information about transmission of HIV was conveyed, and condoms and bleach were distributed. Results from a published evaluation of outreach conducted at 20 of the sites found dramatic decreases in HIV risk behavior among outreach clients.(27) For example, the proportion of clients judged to be at high risk of infection with HIV fell from 62% prior to receiving outreach to 31% at a 6-month follow-up interview, and similar decreases (16 to 8%) were noted in the proportion of clients judged to be at high sexual risk.
Even more striking results were obtained by Wiebel and colleagues,(28) who evaluated outreach targeted to drug users' social networks rather than to individuals. Using this approach, outreach workers promote dialogue about safer injection and safer sex. Distribution of bleach, sterile water, alcohol swabs, and condoms provide occasions for repeated contact with social networks, reinforcing the need for safer behaviors. In a 4-year follow-up evaluation of 641 IDU, Wiebel found a reduction from 100 to 14% in the proportion of IDU who engaged in unsterile sharing of equipment. Coinciding with these changes, HIV seroincidence decreased from over 5% per semester in the first year of the study to less than 1% per semester in the last year. Outreach may owe much of its effectiveness to the fact that it is usually conducted by peer educators who are recovering drug users known to the drug user community.
A limitation of the NADR studies, including Wiebel's, is that while the behavioral changes reported occurred in the context of community-based interventions, alternative explanations could account for the findings, including maturation out of drug use and the data collection process itself, which involved repeated waves of interviews and HIV testing. The behavioral changes in Wiebel's study, however, were apparently not due to historical factors, because no changes were observed in a nonequivalent control group.
| Counseling and Testing|
Little conclusive evidence exists that counseling and testing is an effective approach to reducing high-risk behavior in IDU. Few studies that have examined outcomes of HIV testing with IDU have reported reductions in risk behavior (for a review, see Higgins and colleagues.(29)) In two studies, greater needle hygiene was noted among tested users than among users who had never tested or received their test results.(30,31) Results of three studies, however, showed no differences in injection or sexual practices between tested and untested drug users,(32-34) and results of four other studies(35-38) demonstrated a mixed pattern of relationships between testing and behavior. Only two studies assessed outcomes with a randomized design that followed subjects after they had been tested.(34,39) In both studies, IDU assigned to group/individual counseling reduced high-risk behaviors, but comparison group subjects reported similar reductions.
| Other Interventions|
The pattern of results of these two studies has been mirrored in other controlled evaluations of interventions to prevent HIV in drug users. A review of 17 such evaluations identified only three without design limitations in which the experimental intervention was more effective than a comparison condition in reducing the prevalence of high-risk practices.(40) In the largest of three evaluations (n = 567), which involved following subjects for 12 months, a six-session skills building group intervention proved more effective in reducing high-risk injection practices than a two-session informational intervention; this was true, however, only for subjects initially at lower risk.(41,42)
In a second, smaller (n = 93) study,(43) IDU assigned to 6 hours of counseling designed to prevent relapse to high-risk behaviors reported less high-risk injection behavior during the heaviest risk-taking month following a baseline assessment; however, they did not differ from no-treatment controls for the most recent month, and there were no differences between the experimental and comparison groups in self-reported sexual practice.
In a third study,(44,45) a small (n = 84) sample of female methadone patients assigned to five 2-hour skills building intervention sessions reported greater condom use 15 months following intervention than subjects assigned to an information-only control group.
Comparing these outcomes with those of the other evaluations, the authors of the review(40) concluded that the success of the experimental interventions appeared to be due to their greater length as well as to having been conducted with stable and well-motivated (drug treatment) populations. Their more surprising finding, however, concerned 10 of the 17 studies in which no differences in outcomes were noted between the experimental and comparison group(s).
In these 10 studies, there was evidence of marked behavioral changes in both the experimental and comparison conditions, with the changes in several cases being sustained for as long as to 12 months. A close examination of the evidence for completing hypotheses for this pattern of results suggested that participation in research may have had contributed to this patterns of findings. In one study, subjects assigned to a wait-list control group had, by the time of a 4-month follow-up interview, reduced high-risk practices as much as subjects assigned to two experimental interventions. It is worth noting that all of the studies involved lengthy behavioral assessments prior to randomizing subjects.
That relatively brief assessments can have a marked impact on behavior should not surprise practitioners in the fields of public health and preventive medicine. In the last 20 years, health risk assessments have become commonplace in the workplace and other settings. The aim of such assessments is to influence individual health behavior in a "healthward" direction.(46) A health risk assessment typically consists of the following elements(46):
An assessment of a person's health habits and risk factors based on responses to questionnaire items, sometimes supplemented by biomedical measurements such as blood pressure and laboratory tests
A quantitative estimation or qualitative assessment of the individual's future risk of death and/or other adverse health outcomes
The provision of educational messages and/or counseling about how changing one or more personal risk factors would alter a person's risk of disease or death
There is evidence that health risk assessments can be a powerful tool for changing behavior. For example, a large randomized trial of a health risk assessment involving administration of health habit questionnaires, individualized health recommendation letters, newsletters, and other program materials found a sizeable impact on a variety of health behaviors over a 12-month period, including smoking, dietary intake of salt and fat, alcohol consumption, and exercise.(47)
Both similarities and differences exist between health risk appraisals and assessments employed in evaluating interventions to prevent HIV in drug users. The differences include the fact that health risk appraisals cover a wide variety of health-related risks, whereas assessments of drug users in intravenous behavioral research involve reviewing a relatively limited range of behaviors having to do with drug use and sex. Another difference is that behavioral assessments of drug users do not involve an evaluation of subjects' future risk of death per se, although given the life-threatening nature of AIDS and widespread knowledge about links between certain practices and infection with HIV, such an evaluation may not have been needed.
Health risk appraisals and assessments of drug users high-risk behaviors nevertheless are similar in that they involve a microscopic examination of participants'/subjects' behavior as it relates to the probability of disease and death. Behavioral assessments with drug users, however, frequently elicit information not only about behavior but also about an array of knowledge, attitudes, and beliefs that may support such behavior. For example, McCusker and colleagues, in addition to inquiring about the high-risk sexual practices and injection behavior of subjects, administered four knowledge scales, a susceptibility scale, and items measuring subjects' confidence or self-efficacy in carrying out eight specific behaviors. The potential for repeated questioning about a narrow range of practices to change behavior would seem to be considerable.
Several recommendations have been made with regard to preventing new HIV infections in IDU. First, as recommended by the National Academy of Sciences and the National Institute on Drug Abuse, drug treatment needs to be made available to all drug users seeking help for their addiction. Nadelmann and McNeely observed that if methadone were readily available, many tens and perhaps hundreds of thousands more IDU would trade their illicit heroin habit for a legal methadone dependence.(48) As a result, many would avoid painful, deadly HIV infections costing an average of $119,000 per person to treat.(17)
Second, the U.S. Congress needs to repeal its ban on the use of federal funds for needle exchanges, and state and local governments should repeal prescription and paraphernalia laws as they apply to syringes. The unanimous conclusion of six U.S. Government-funded reports was that needle exchange reduces HIV transmission, and four of the six recommended revoking both the federal funding ban and state prescription and paraphernalia laws (the other two took no position).(49)
Finally, although street outreach by peer educators should continue to be a staple of prevention strategies with drug users because it has led to dramatically reduced levels of high-risk behavior, a significant proportion of drug users continue to practice unacceptably high levels of risk behaviors. New, more powerful interventions are needed to slow the spread of HIV in drug users. A report recently prepared for the U.S. Office of Technology Assessment suggests that a social change approach to preventing HIV in drug users has much to add to individually oriented programs.(50) Community-level interventions are an important example of such an approach. Community interventions differ from those we have described in that they are designed to change the norms of entire communities or populations of individuals. Such interventions have proven effective in dealing with other problem health behaviors,(51-53) and show potential for preventing the spread of HIV.
As an example, a community intervention evaluated by Kegeles and colleagues(54) succeeded in decreasing by 45% the proportion of gay men in Eugene, Oregon, aged 18 to 29 years, reporting unprotected anal intercourse (UAI) with nonprimary partners, and by 24% reporting UAI with boyfriends. The intervention involved a major publicity campaign and grassroots efforts to mobilize Eugene's young gay men. Interventions such as these may succeed by reaching beyond those individuals directly touched by them and by changing community norms supporting unsafe behaviors.
|| ||Institute of Medicine/National Academic of Sciences. Confronting AIDS: Directions for Public Health, Health Care, and Research. Washington, DC: National Academy Press, 1988.|
|| ||Schuster CR. Drug abuse treatment and AIDS prevention. Int Working Group AIDS IV Drug Use Newsletter 1989;4:12-13.|
|| ||Ball JC, Ross A. The Effectiveness of Methadone Maintenance Treatment. New York: Springer, 1990.|
|| ||Kahn JG. Do NEPS affect rates of HIV drug and/or sex risk behaviors? In: Lurie P, Reingold AL, et al. The Public Health Impact of Needle Exchange Programs in the United States and Abroad. Atlanta: Centers for Disease Control and Prevention, 1993.|
|| ||Dondero T. Human immunodeficiency virus infection in the United States: A review of current knowledge. MMWR 1987;36:125-174.|
|| ||Turner C, Miller H, Moses L. AIDS: Sexual Behavior and Intravenous Drug Use. Washington: National Academic Press, 1989.|
|| ||Lane S. How and why did NEPs develop? In: Lurie P, Reingold AL, et al. Public Health Impact of Needle Exchange Programs in the United States and Abroad. Atlanta: Centers for Disease Control and Prevention, 1993.|
|| ||Kahn JG. Are NEPS cost-effective in preventing HIV infection? In: Lurie P, Reingold AL, et al. Public Health Impact of Needle Exchange Programs in the United States and Abroad. Atlanta: Centers for Disease Control and Prevention, 1993.|
|| ||Gibson DR, Kahn JG. Maximizing the impact of needle exchange: The role of penetration and volume. Presented at the North American Syringe Exchange Network annual meetings, San Juan, Puerto Rico, April 1995.|
|| ||Flynn NM, Anderson RA, Britton J. Syringe exchange in a hostile environment. Presented at the Annual Meeting of the American Public Health Association, San Diego, November, 1995.|
|| ||Stephens RC, Simpson DD, Coyle SL, et al. Comparative effectiveness of NADR interventions. In: Brown BS, Beschner GM, eds. Handbook on Risk of AIDS: Injection Drug Users and Sexual Partners. Westport, CT: Greenwood Press, 1993.|
|| ||Marlink RG, et al. High rate of HTLVIII/HIV exposure in IVDUs from a small-size city and the failure of specialized methadone maintenance to prevent further drug use. Presentation at III International AIDS Conference, June 1987, Washington, DC.|
|| ||van den Hoek JA, van Haastrect HJA, Coutinho RA. Little change in sexual behavior in drug users in Amsterdam: Implications for the AIDS epidemic. Presentation at the VI International AIDS Conference, June 1990, San Francisco.|
|| ||Farley T, Carter M, Hadler J. HIV counseling and testing in methadone programs: Effect on treatment compliance. Presentation at VI International AIDS Conference, June 1990, San Francisco.|
|| ||Gibson DR, Lovelle-Drache J, Young M, et al. Effectiveness of brief counseling in reducing HIV risk behavior in injecting drug users: Final results of randomized trials of counseling with and without HIV testing. Submitted for publication, 1997.|
|| ||Gibson DR, McCusker J, Chesney MA. Psychosocial interventions: How effective in reducing HIV risk behavior in injecting drug users. Submitted for publication, 1997.|
|| ||Nadelmann E, McNeely J. Doing methadone right. Public Interest 1996;123:83-93.|
|| ||Lurie P, Drucker E. An opportunity lost: Estimating the number of infections due to the U.S. failure to adopt a national needle exchange policy. Submitted for publication, 1997.|
|| ||Des Jarlais DC. HIV/AIDS prevention for injecting drug users. Report submitted to the Office of Technology Assessment by the American Psychological Association. Washington: American Psychological Association, 1995.|
|| ||Coates TJ, Greenblatt R. Behavior Change Using Community Interventions at the Community Level. New York: Plenum, 1989.|