You are on page 1of 19

Measuring harm reduction: the effects of needle and

syringe exchange programs and methadone
maintenance on the ecology of HIV
Ernest Drucker, Peter Lurie*, Alex Wodakt† and Philip Alcabes*
AIDS 1998, 12 (suppl A):S217-5230

Keywords: Harm reduction, injecting drug use, HIV transmission, methadone maintenance,
needle and syringe exchange

A public health approach
Minimizing the epidemic spread of an infectious disease requires that we carefully examine data on the effectiveness of
all interventions believed to reduce transmission. For AIDS, the absence of an effective vaccine and the continued high
cost and still undetermined long-term efficacy of antiviral medications dictate a focus on preventing HIV infection - our
only available tool to slow the global pandemic [1]. While the effect of combination therapies (using the new protease
drugs) on viral load suggests that competent and ongoing medical care for people with HIV/AIDS will reduce their
infectivity, the high cost of this approach means it has no public health consequence outside the developed world.

Interrupting transmission of HIV means reducing the susceptibility of the uninfected by controlling the quantity and
characteristics of contact with all infectious vectors. But the complex social and behavioral ecology associated with the
spread of HIV in human populations makes this very difficult, involving us in some of the most intimate (and often
proscribed) areas of personal life: sexuality and drug use. Nonetheless, in order to slow the spread of AIDS, we must
examine all available measures for reducing those risks that are associated with the transmission of HIV And, as is always
the case in public health, we must pay the closest attention and give the highest (priority to those factors that account for
the greatest proportion of incident infections and are most amenable to changes that reduce transmission. Illicit drug use,
particularly injecting drug use (IDU), qualifies on both counts it is a potent risk factor for transmission of HIV (as well as
for hepatitis and other infectious diseases) and is, we believe, accessible to risk reduction by a number of social and
behavioral interventions currently available to us.

Drug use and AIDS
The overall importance of nonmedical (illicit) drug use for the AIDS pandemic lies in its global prevalence, rapid growth,
and central role in igniting regional outbreaks of HIV infection. There are now an estimated 3.4-5.5 million injecting drug
users in 120 countries [1], one part of a larger population who constitute the illicit drug market for heroin (8-10 million
persons), cocaine (30--40 million), and amphetamines (at least 30 million) .While not all illicit drug use involves injection
(for example, cannabis), the practice of IDU has expanded rapidly in the 17 years since AIDS was first identified in 1981,
emerging in scores of nations with no prior history of IDU many of which are ill-equipped to restrict drug use in their
populations. The prevalence of IDU and the severity of its risks for HIV transmission are fueled by (and fuel) poverty,
social dislocation, and political instability. Finally, the rapid expansion of international drug markets creates a profitable
criminal `economy' that undermines and corrupts key institutions of civil societies, including those responsible for
addressing the public health problems associated with drug use and addiction [2]. IDU has proven itself capable of
igniting the regional spread of HIV infection in both developed and developing societies with astonishing speed. This
pattern was first evident in poor urban populations in the USA and Europe, but is now also seen in Asia, Africa, and Latin
America [3].
From the Montefiore Medical Center/Albert Einstein College of Medicine and the Lindesmith Center, New York, USA, the *University of
California, San Francisco and University of Michigan, Ann Arbor, Michigan, USA, the 'New York University School of Medicine, New
York, USA and 'St Vincent's Hospital, Sydney, Australia.
Sponsorship: This work was supported by NIDA under awards DA 09-712 (P.L. and E.D.), R01 DA 11324-01 (E.D.),the Lindesmith
Center/The Open Society Institute (E.D.) and the New York University CFAR under NIAID award P30 A127742 (PA.).
Requests for reprints to: Dr Ernest Drucker, Department of Epidemiology and Social Medicine, Montefiore Medical Center,
111 East 210th Street, Bronx, NY 10467, USA.
© Lippincott-Raven Publishers ISBN 1-41284-040-5 ISSN 1350-2840S217

Drug control policies and health risk
International drug control strategies and regimes based almost exclusively on prohibition (in other words, on the use of
criminal penalties to regulate the sale and use of certain drugs) appear to have been unsuccessful in containing IDU
geographically [3]. In the years of epidemic HIV spread (since the 1970s), despite massively increased interdiction and
drug enforcement activities, worldwide heroin and cocaine production have tripled [1], with comparable increases in the
prevalence of IDU and addiction. In addition, prohibitive drug policies also affect patterns of drug use in ways that
increase individual and public health risk. The necessarily clandestine life of the (criminal) addict fosters several specific
patterns of behavior known to increase HIV transmission, most important among these the sharing of scarce injection
equipment and increased sex work to get money for drugs. Stigmatizing and marginalizing the drug user also acts as a
barrier to access to medical and social services, isolating the drug user from education, prevention, and drug treatment that
might ameliorate AIDS and other health risks (for example, overdose).

In this context of rapidly expanding drug markets and the growing prevalence of drug use worldwide, the major test of the
public health efficacy of specific AIDS prevention measures is, therefore, not their impact on the prevalence of drug use
per se (which is a function of drug policies and markets) but, rather, their effect on the transmission risk associated with
continued and possibly increased drug use within the current environment, in other words in the context of drug policies
based on absolute prohibition.

Application of the communicable disease model
From a communicable disease standpoint, AIDS in the drug-using population is a vector-borne epidemic, with the needle
and syringe serving as the `vector' responsible for transmission of the biologic agent of infection, HIV AIDS control must
take into account a social ecology that includes the interaction of drug users with injection equipment (and with drug
policies) in ways that determine patterns of drug use and of risk [4]. Clearly, there also can be other modes of HIV
transmission in this population, notably sexual and (in countries that cannot protect the blood supply) by transfusion.
However, the salient role of the syringe in mediating transmission, the potentially higher transmission potential in
MISSING TEXT injection equipment (as opposed to most sexual contact), and the generally higher volume of contact
through needles and syringes for addicts (who inject at least once and often many times per day, every day) demand
attending to this `vector' in designing intervention strategies for restricting the spread of infection in this population.

Public-health practice for limiting communicable disease has long experience in control of arthropod vectors for malaria,
yellow fever, dengue, and other diseases. And, although some preventive interventions focus on wholesale ecological
change to halt disease spread (for example, draining swamps), the majority rely on one of four broad classes of
intervention: immunization, vector elimination, biologic vector control, and personal protection.

Immunization, which exists for other vector-borne diseases, is currently unavailable for HIV, and we are still years (at
least) away from the earliest possible large-scale vaccine intervention.

Vector elimination was embraced as a drug control policy long before AIDS, in the form of `drug paraphernalia' laws.
These were based on the (misguided) belief that criminal sanctions would eliminate or severely restrict needles and
syringes and thereby limit IDU This policy's disregard for the social ecology of the `vector' and the motivation of its host
caused it to backfire as a disease control strategy, forcing users to share injecting equipment more frequently [4]. In some
areas of the USA this policy drove IDU into the now-infamous `shooting galleries,' where the spread of HIV was
accelerated [5]. Under the circumstance outlined above, the total elimination of the critical vectors of IDU (injecting
equipment and addictive drugs) do not seem plausible goals for public-health actions

Biologic vector control generally recognizes the complex interactions among vector, pathogen, and host, and seeks to shift
the ecological balance by introducing variants of the vector (for example, new species of mosquitoes that will not carry
malaria) in hopes that they will outcompete the transmissive vector and take over its ecological niche. The initial use of
bleach to sterilize syringes, while raising awareness of risk and creating the vehicle for the first outreach and peer-Ied
AIDS prevention programs for IDU in the USA [6], was found to be only partly effective, in part because (like vector
elimination) it failed to adequately account for the social circumstance in which IDU occurred and the difficulties this
posed for interuse cleaning of injecting equipment [7]. These efforts. however, paved the way for the introduction of
sterilesterile injecting equipment in ways that displace contaminated equipment and significantly alter the risk
environment of IDU, acting as an effective form of biologic vector control [8].

Personal protection is a common strategy for prevention MISSING TEXT mosquito netting against malaria. yellow
fever, or dengue) and sexually transmitted diseases (STD; including condoms to prevent HIV transmission). This model
can be applied easily to understanding and modifying risks associated with the ecology of IDU, even within societies that
criminalize drug use and injection equipment.The medical provision to addicts of forms of opioid drugs that do not require
injection (for example, oral methadone in a maintenance regimen) or sterile injectable drugs and equipment (for example,
heroin maintenance) can eliminate exposure to HIV for the individual drug user in treatment. The more accessible these
substitute drugs are and the greater the proportion of addicts using them, the greater the protection they will confer on the
entire population of users and their secondary contacts.

Evaluating the evidence
In the next section of this paper we compile and review the evidence on the public-health impact of two such interventions
for interrupting HIV transmission among IDU: needle and syringe exchange programs (NSEP) and methadone
maintenance treatment. We summarize the findings and methods of this body of research and update earlier assessments
of their impact on AIDS risk factors.

A methodological note: in public health, it is often neither feasible nor ethical to conduct strictly randomized controlled
trials of interventions. In the midst of an epidemic, such as AIDS, circumstances may preclude the strict control of all
collateral risk factors and the illicit nature of drug use impedes the collection of complete and timely data to assess efforts
to reduce risks associated with an ever-changing social and behavioral environment in which most illicit drug use occurs.

But these limitations do not exempt us from making our best efforts to protect public health or from using existing
evidence to inform our actions. By employing a publichealth analvsis of the circumstances of IDU and transmission of
HIV (as above), we take advantage of ecological associations, observational studies, and `natural experiments' which
point the way towards the best available measures to blunt the epidemic's force. In addressing these same concerns, the
report of the USA National Academy of Sciences' Institute of Medicine on the efficacy of NSEP [8] (pages 200-202)
weighed both the `traditional approach' (to the preponderance of evidence) and an 'alternative approach' (to the patterns of
evidence), concluding that to reject NSEP, based on limitations of the design of single studies, ignores both the
preponderance and pattern of the evidence and `is both poor scientific judgment and bad public health policy'. The
Institute argued `that the improbability of being able to carry out the definitive study ... does not necessarily preclude the
possibility of making confident scientific judgments' and (citing the admonition of the biostatistician A. Bradford Hill in a
1965 address to the Royal Society of Medicine) that `incomplete' scientific evidence `does not confer upon us a freedom
to ignore the knowledge we already have, or to postpone the action that it appears to demand'. It is this approach that
guides our review and interpretation of the current evidence, and that has (fortunately) guided public-health officials and
community-health advocates throughout the world, who have acted based on assessing the relative impact of available
interventions and chosen to mount substantial national NSEP and methadone treatment to reduce HIV risk. Indeed, most
of the AIDS prevention measures in force today (for example, counseling, promotion of condom use) were begun and had
substantial effects long before any controlled trials had demonstrated their efficacy.

Needle and syringe exchange
NSEP are one of the hallmarks of public-health innovation for AIDS prevention among IDU. These programs
not only provide sterile injecting equipment to active IDU, they also offer sex education and prevention
materials, referrals to medical care, legal and social services, and drug treatment. And, by reducing the
marginalization of drug users, they increase the likelihood that IDU who continue to use drugs will do so more

The specific biologic action of NSEP (as a form of vector control) is to reduce the time that needles and
syringes spend in circulation; the less time that potentially contaminated injecting equipment circulates among
drug users, the fewer individuals it has a chance to `share' and the less chance it has of being contaminated and
of being used thereafter by an uninfected user [9]. Ideally, each syringe would be in circulation only long
enough to be used by a single person a single time (as the USA Centers for Disease Control recommends [10])
with the potential for eliminating syringe-mediated HIV transmission altogether. The NSEP accomplishes this
diminution in syringe lifetime in two ways. First, its volume-related circulation effect is to increase the total
number of syringes and needles available in a given time. Since NSEP do not recruit nonusers, the total number
of users does not change and the NSEP increases the number of syringes per capita, driving used syringes out of
circulation. Second, NSEP increase the rate at which syringes and needles are removed from circulation or,
equivalently, reduce the average circulation time of each syringe and needle by physically removing used
syringes from circulation, the essence of the `exchange.' This also helps eliminate used syringes in the
environment, reducing the risk of inadvertent needlesticks of nonusers.
In 1993, one of the present authors was part of a group of researchers who summarized all studies, published
and unpublished, then available on the efficacy of NSEP [11]. That review concluded that it was likely that
NSEP would reduce HIV incidence and that there was no evidence that NSEP were associated with increases in
IDU or community levels of drug use. Since that time, two dozen articles that assess the public health impact of
NSEP have been published; in this report we summarize their results.

We conducted a computerized literature review using the keyword `needle exchange' for English language
articles published from 1993 through to the end of 1997. We examined the references in articles identified by
our literature search and contacted NSEP experts to identify additional articles. Books with articles on NSEP
were also included, but abstracts from conferences were not. Articles that were only available in abstract form
at the time of our previous report were included if they presented at least some new data.

We included only articles that addressed the impact of NSEP upon drug use rates, injection-risk behaviors,
hepatitis and HIV (Table 1) [12-35]. Serial cross-sectional studies addressing the impact of NSEP upon the IDU
community were included as this a key question about the effect of NSEP Otherwise cross-sectional studies
were excluded, as they cannot address trends over time and serial cross-sectional studies of particular cities
were included only if the data for NSEP participants and nonparticipants were analyzed separately. All changes
reported are statistically significant at P < 0.05. If there was more than one statistically significant finding,
representative findings are reported. Some studies appear more than once in Table 1 because they addressed
more than one outcome or used more than one analytic approach; each is considered a separate study. Studies of
cost-effectiveness, changes in sexual risk behavior (there were few of these), and the impact upon referrals to
drug treatment were not included. We assumed that successful publication was a rough proxy for quality and
did not use a quality scale as in our previous report.

We identified a total of 24 articles that have been published since our previous report. The impact of NSEP on
drug use rates was addressed in 10 articles, on injection risk behaviors in 12 articles, on hepatitis B and C in one
article, and on HIV in 11 articles. The studies were generally of a higher quality than those available in 1993
with a wider range of evaluation designs. In particular, ecological and case-control designs were more common,
and prospective studies of HIV incidence had greater statistical power and were therefore analyzed using more
sophisticated methods than was possible with the small prospective studies that were available in 1993. Control
groups were more often used in the studies measuring NSEP impact upon injection risk behaviors and blood-
borne pathogen transmission. Half of the studies were conducted in the USA.

The results of the studies are summarized in Table 2 with the different study designs combined. Of the 10
studies MISSING TEXT rates, four serial cross-sectional studies examined the impact of NSEP on drug use
rates in the IDU community and the remaining six assessed NSEP impact on drug injection rates of IDU using
the NSEP Some studies used mean IDU age to investigate whether young people had begun injecting as a result
of the NSEP Seven studies reported reductions in measures of drug use, two reported no change, and one an
adverse effect associated with NSEP The latter finding, in a drug treatment cohort, was a greater decline in the
percentage of IDU who never used the NSEP and who continued to inject, compared to IDU who ever used the
NSEP This observation has been made once before [36], and likely represents self-selection: IDU who expected
to continue injecting drugs may have been more likely to select NSEP use. In any event, the number of studies
with beneficial or neutral effects of NSEP upon drug use measures greatly exceeds this single study.

Eleven out of 12 studies examining injection risk behaviors reported improvements, generally decreases in
syringe sharing rates. None showed an increase in injection risk behavior. By comparison, in our 1993 study
[37],16 out of 23 studies showed a beneficial NSEP effect, six showed a mixed or neutral effect, and one
showed an adverse effect (which was not apparent in subsequent studies [38]).
Of the 12 studies examining the impact of NSEP upon blood-borne infections, six reported beneficial effects for
NSEP, four reported neither a beneficial nor an adverse effect, and two reported an adverse effect. The adverse
effects were reported from a single study in Montreal that used two different methods to analyze the same data
set [33,39].

Although substantially more studies reported beneficial or neutral effects of NSEP upon blood-borne pathogen
transmission, the Montreal study has attracted significant attention from USA policy makers and merits special
mention [40].The study followed 974 HIV-negative drug users prospectively; IDU who attended the NSEP at
the intake interview were almost twice as likely to acquire HIV as those not attending the NSEP, even after
adjusting for potential confounders. Similar findings were apparent in the nested case-control analysis.
However, the Montreal authors demonstrated clearly that IDU who attend NSEP in Montreal are systematically
different than other local IDU: NSEP participants are much less likely to be enrolled in drug treatment and
much more likely to inject frequently, borrow injection equipment from other IDU, attend shooting galleries.
and share injection equipment with HIV-positive IDU. It is not surprising that a group of that description would
have higher seroincidence rates.

The on-the-ground realities of syringe availability in Montreal suggest why high-risk IDU may be more likely
to attend the NSEP. First, Montreal public-health officials MISSING TEXT provision of sterile syringes.
Second, for most of the study period, CACTUS, the major Montreal NSEP was only' open from 21:00 to 04:00
h and annually exchanged fewer than 200 ()00 syringes, not a large number for a city, of that size Finally,
CACTUS is located in an area noted for commercial sex work. The NSEP was therefore likely to serve IDU
who could not obtain syringes from pharmacies and who were in the streets in the early morning.

The Vancouver study [35] has also attracted the attention of NSEP opponents, even though we rate the study as
demonstrating neither an adverse nor a beneficial effect for NSEP That study demonstrated that a significant
HIV epidemic occurred despite the presence of an NSEP (this does not mean that the NSEP was ineffective,
only that it was not perfectly effective) and, like the Montreal study, that the highest-risk IDU attend the NSEP
The number of incident HIV infections was only 24 and almost all used the NSEP, so it was impossible to say
whether NSEP users were more likely to contract HIV infection than those not using the program or whether the
epidemic would have occurred sooner or been larger had there not been an NSEP

It is critical to recall that, by the authors' own descriptions, neither study was designed to evaluate NSEP But
both the Montreal and the Vancouver studies suggest that NSEP, because they attract high-risk IDU, would be
excellent places to implement behavioral risk-reduction programs. It is significant that public-health officials in
both Montreal and Vancouver, rather than curtailing these programs, have responded to these findings by
expanding the availability of NSEP services in their cities.

In summary, the data since our 1993 report confirm and extend the evidence of NSEP efficacy and the lack of
associated adverse effects. The combination of NSEP with efforts to increase syringe availability by modifying
restrictive laws and regulations [42] and outreach to increase pharmacist involvement in syringe sales [43,44]
holds promise for reducing the toll of HIV infection.

NSEP as part of a national risk-reduction strategy
As the Montreal and Vancouver studies of NSEP indicate, it is often difficult to isolate a single intervention approach
from the larger context in which it functions. And even an effective measure (such as NSEP) may not be able to overcome
all other epidemic forces, especially if applied piecemeal and late in the course of the epidemic. Conversely, if NSEP are
established early, in sufficient quantity, and are part of an overall drug policy which simultaneously promotes a range of
other prevention and treatment services, in other words, drug treatment, their effects on local and regional AIDS outbreaks
can be dramatic. The timely initiation of needle exchange has played a significant role in effectively restricting HIV
spread in IDU in Australia, the UK, Sweden, and Nepal (Table 1) [29]

Political obstacles to expanding NSEP
While these demonstrations of NSEP impact have been sufficient to lead more than 20 countries to adopt NSEP as part of
their national AIDS prevention strategy, USA Federal health authorities have been unyielding in their opposition to NSEP
[29]. At the time of writing, the USA still bans the use of Federal funds for NSEP despite the many USA studies
demonstrating their safety and efficacy, and the clear and repeated recommendations of two National AIDS Commissions,
the National Academy of Science's Institute of Medicine, and leading USA government experts calling for Federal support
of NSEP Further, the USA appears to have some influence in sustaining hostility to NSEP in other countries (Russia,
Sweden). This USA national policy is a landmark failure of public health in the USA: one study [31] found that expanded
provision of NSEP in the USA (with over one million injectors) would have prevented 10,000 to 20,000 HIV infections in
the past decade.

Substitution treatment or maintenance pharmacotherapy using methadone is today the most sought after and effective
form of treatment for opiate addiction [45]. Because methadone is a long-acting opiate whose dosage can be stabilized, it
is well suited to daily administration and has proven effective in the elimination of narcotic craving, a driving force
behind continued heroin use [46]. And, because it can be administered orally, methadone dramatically reduces heroin
injecting frequency and, with it, associated risks for HIV and other blood-borne pathogens (Table 3).

Methadone treatment effectiveness
The clinical effectiveness of methadone is most commonly measured by its retention of patients in care and by
reductions in heroin use as well as improvements in social outcomes, for example, employment, family integration, and
reduced arrests and incarceration for criminal offenses [47]. Both randomized trials and observational studies (Table 4)
[5,48-59] have determined that methadone maintenance retains patients at levels two to four times that of other treatment
modalities (in other words, >75%, 12-month retention) [60], and the longer patients remain in treatment, the better the
results, for example, for those in treatment more than 24 months, methadone reduces the use of heroin to levels below
15% of those in the period immediately before treatment [61]. Conversely, even among those who have greatly reduced
their heroin use while in methadone treatment, over 80% relapse to heroin use when they leave treatment [62]. The most
basic public-health benefit of methadone treatment can be seen in the reduction of mortality rates among IDU who
remain in treatment, observed in randomized clinical trials [63], and later follow-up [64], and, most recently, in
population data from Australia documenting that the (nonHIV-related) mortality rate of patients in methadone treatment
was 25% that of drug users not in treatment [65].
Table 1. Studies of needle and syringe exchange program effectiveness published since 1993
Reference Number Location Recruitment Year of Summary of Findings
Studies of changes in drug use rates
Serial cross-sectional studies
Walters et al. 5644 San Francisco Drug detox and 1986-1992 Percent using NSEP at least 25 times in previous
1994 [12] street recruited year increased from 14% in 1989 to 28% in
1992. Number of daily injections decreased from
1.9 in 1987 to 0.7 in 1992. Mean age increased
from 36 years in 1987 to 42 in 1992.

Hunter et al., "500/year London In- and out- of- 1990-1993 Percent using NSEP as main syringe source
1995 [13} treatment IDU increased from 37% to 50%. Mean age rose from
28 to 32 years. Percent injecting daily remained

Peak et al., 59-200/year Kathmandu IDU at NSEP 1991-1994 Number of injections per month decreased from
1995 {14] median of 21 to 15.

Des Jarlais et 1115 New York City Detox program 1984 Mean IDU age increased from 32 years to 36
al., 1990-1992 years. Number of cocaine injections per month
1994 [15] declined from 55 to 43 as injection frequency of
other drugs remained stable.

Fennema el al., "200/year Amsterdam Street recruited 1990, 1993 Duration of drug injection increased from 10
1997 [16] IDU 1996 years to 14 years. Mean age increased from 28
years to 37 years. Decrease in percent injecting
? twice a day from 74%, to 56%,.

Retrospective cohort studies
Hunter et al., 426 Tacoma IDU at NSEP 1988- Number of injections per month remained stable
1994 [17]

Paone et. al., 1752 New York City I 1 Numb
1994 [18] D9 inject
U9 mont
2 declin
a- from
t 1

Prospective cohort studies

Oliver et al., 77 frequent Portland NSEP and 1989-1991 Change in percentage of IDU reporting injecting
1994 [19] NSFP users outreach clients now similar in NSEP users and NADR clients.
and 355 NADR

Schoenbaum et 329 New York City Methadone 1989-1993 61 % relative decline in proportion of non-NSEP
al., 1996 [20] main-tenance users continuing to inject versus 14% relative
program decline in those ever using the NSEP.

Vlahov et al., 221 Baltimore IDU at NSEP 1994-1995 Percent injecting at least daily decreased from
1997 [21] 97% at baseline to 77% at 6 months. Number of
injections per clay decreased from 5.7 at
baseline to 4.2 at 6 months.

Studies of changes in injection risk behaviors

Serial cross-sectional studies

Walters et al., 5644 San Francisco Drug detox and 1986-1992 Percent using NSEP at least 25 times in previous
1994 [12] street recruited year increased from 14% in 1989 to 28% in
1992. Decrease in sharing from 66%, in 1987 to
36% in 1992 for full study population. In
multivariate analysis, greater use of NSEP
associated with less sharing.

Hankins et al., Montreal IDU at NSEP Lending of needles decreased from 31% at
1994 [22] opening of NSEP to 20%,. Bleach use increase
from 30%, at opening of NSEP to 62%.

van Ameijden 616 Amsterdam In- and out-of- t 1986-1992 In multivariate analysis, current NSEP use
et al., 1994 associated with reduction in reuse of syringes.

Peak et al., 59-200/year Kathmandu IDU at NSEP 1991-1994 Number of times syringes shared per month
1995 [14] declined from 13 to 8.

Retrospective cohort studies

Frischer and 503 Glasgow IDU using and 1990 In multivariate analysis, NSEP users had made
Elliot, not more harm reduction changes than nonusers.
1993 [24] using NSEP

Keene el al 328 Wales IDU using and 1990-1991 NSEP users and nonusers had similar rates of
1993 [25] not ever sharing syringes, but 18"6, of users shared
using NSEP syringes in past year compared to 52% of

Hagan et al., 426 Tacoma IDU at NSEP 1988- 64º/<, relative reduction in percent injecting
1993 [17] with at least one used syringe per month. 67%
relative reduction in percent passing on used
syringe al least once per month.
Paone et al., 1752 New York City IDU at NSEP 1992-1993 Injection with previously used syringes declined
1994 [18] from 12%, of the time to 4%. Percent injecting
with previously used syringes declined from 22%
to 6%.

Hahn 141 San Francisco IDU in drug 1989-1990 Number of sharing partners declined equally in
1997 [26] treatment IDU who ever and never used NSEP

Prospective cohort studies

Oliver et al., 77 frequent Portland NSEP and 1989-1991 Greater reductions in percentage reporting reuse
1994 [19] NSEP users outreach clients of syringes and throwing away syringes in NSEP
and 355 NADR clients than in NADR clients.

Schoenbaum et 329 New York City Methadone 1989-1993 Needle sharing declined more among IDU ever
al., maintenance using the NSEP than among nonNSEP users
1996 [20] program

Vlahov et al., 221 Baltimore IDU at NSEP 1994-1995 Per cent using someone elses’ syringe in previous
1997 [21] 6 months decreased from 22% at baseline to
8%, at 6 months follow-up. Per cent giving
syringe to friend in previous 6 months declined
from 27% at baseline to 12% at 6 months
follow-up. Number of injections per syringe
declined from 12.8% at baseline to 3.6% at 6
months follow-up.

Studies of Hepatitis B and C

Case-control studies

Hagan et al. 28 IDU with Tacoma IDU using health 1991-1993 For hepatitis B, ever use of NSEP associated with
1995 [27] hepatitisB and department adjusted odds ratio of 5.5. For hepatitis C, 'ever
20 with services use of NSEP associated with adjusted odds ratio
hepatitis C of 7.3

Studies of HIV

Ecological studies

Des larlais et >7300 Worldwide IDU in five cities 1984-1993 Cities with consistently low seroprevalence had
al., with stable all made efforts to provide sterile injection
1999 [34] seroprevalence equipment.

Hurley et a1., 332 892 IDU Worldwide IDU in 81 cities 1984-1994 HIV seroprevalence declined 5.8% per year in 29
1997 [29] (75% of IDU in cities with NSEP and increased 5.9% per year in
treatment) cities without NSEP.

Mathematical modeling studies

Kaplan et al. New Haven 1990-1992 HIV prevalence in NSEP syringes returned
1994 [30] decreased from over 69% to under 45% This
decrease not explained by changes in client

Lurie and USA 1987-2000 Up to 9666 HIV infections could have been
Drucker prevented by NSEP by 1995. Will cost $244
1997 [31] million to treat those infections.

Case-control studies

Patrick et al, 89 cases Vancouver Referrals, 1994 In multivariate analysis, no relationship between
1997 [32] 192 controls outreach, frequency of NSEP use and HIV seroconversion

Bruneau et al. 88 cases Montreal Treatment, 1988-1995 In nested case-control analysis, exclusive NSEP
1997 [33] 320 controls outreach users 6.5 times as likely to seroconvert.
and self-referrals

Prospective cohort studies
Hagan et al. Tacoma Users and 1988- No seroconversions among NSEP users (223
1994 [17] non-users of person-years of follow-up) or nonusers (67 person-
NSEP years of follow-up).

Des larlais et 1630 New York City In- and out- 1988-1995 NSEP attenders 3.5 times less likely to seroconvert
al., and Puerto of- treatment than nonattenders.
1999 [34] Rico

Schoenbaum et 329 New York City Methadone 1989-1993 HIV incidence similar between NSEP users and
al., maintenance nonusers.
1996 [20] program

Bruneau et al. 974 Montreal Treatment, 1988-1995 In multivariate analysis, NSEP users at baseline 2.2
1997 [33] outreach times more likely to seroconvert than nonusers.
and self-referrals

Strathdee et. 257 HIV-IDU Vancouver Street 1996-1997 Incidence was 18.6/100 person-years. 23/24
al., 1997 [35] outreach seroconverters reported NSEP as main source of
sample (92% syringes.
ever used

NSEP, needle and syringe exchange programs, IDU, injection drug users; NADR, National AIDS Demonstration Research project

Table 2 Summary of published studies of NSEP impace 1993-1997
Beneficial Neutral Adverse
Outcome NSEP effect NSEP effect NSEP effect

Drug use rates 7 studies 2 studies 1 study
injection risk 11 studies 1 study 0 studies
Blood-borne 6 studies 4 studies 2 studies

NSEP, needle and syringe exchange programs

Table 3 Methadone studies supporting HIV risk reduction (1969 - 1971)*
Outcomes RCT Obs Pre&Post
Reduced use of heroin 7 7 5
Reduced crime and incarceration 2 10 9
Improved social integration 2 1
(employment, family, housing)
Retained in treatment 3 3
(better access to medical care)
Reduced injection NA 4
Reduced sharing NA 11
Cohort studies showing length of NA 11
time in treatment associated with
low rates of HIV+
*The major published studies, most often cited in the literature, RCT, randomized control trials; Obs, observational
studies (prospective and retrospective); Pre&Post, pre- and postintervention

Methadone treatment as AIDS prevention
In addition to its well demonstrated clinical effectiveness in treating opiate dependency, there is strong evidence that
methadone treatment reduces AIDS risk among IDU (and their sexual partners and offspring) by reducing behavior
known to lead to transmission of HIV, in other words, the frequency of injection of heroin, the sharing of injection
equipment, and sex work needed to get money for drugs. In addition, methadone lowers the rate of imprisonment among
heroin users, itself a risk factor for HIV infection [66]. Further, programs which have integrated methadone treatment with
health services have improved access to both primary and specialized medical care, for example prenatal care, detection
and treatment of STD, and treatment of tuberculosis [67]. Involvement in methadone treatment also facilitates HIV/AIDS
screening programs and medical care, and helps adherence to the new antiviral combination therapies.

Table 3 shows a summary of these effects from the major studies on methadone published between 1969 and 1997,
documenting each specific area of methadone treatment's impact relevant to HIV risk reduction, by type of outcome.

Enrollment and retention in treatment and HIV
Retention in methadone treatment varies as a function of program characteristics [68] and dose [69], but length of time in
treatment is generally associated with lower rates of HIV, especially if treatment is initiated early in the course of the local
epidemic of HIV among IDU [5]. A striking example of this effect can be seen in the virtual absence of HIV infection
among IDU admitted for methadone treatment in New York City prior to 1978 [49] and lower HIV rates among those
with more time undergoing methadone treatment [70], a finding confirmed by subsequent studies in the USA, Europe, and
Austraha. Most studies of methadone are observational (random assignment to methadone being ethically inconsistent
with the evidence of its efficacy), and involve self selection into methadone treatment (or the choice to leave it). One early
study that did involve random assignment among applicants to a limited-capacity methadone program began in Sweden in
the early years of AIDS, and found that only 3% of those admitted before 1983 were HIV-positive. However, 16% of
those subsequently admitted (between 1984 and 1986) and 57% of those admitted after 1987 tested positive for HIV [64].
Further, among those HIV-negative individuals admitted since 1984, no seroconversions occurred in 3-4 years of follow-
up. Low-threshold methadone programs and those employing general practitioners have proven effective in attracting
more marginal IDU to longer term methadone treatment, but are more difficult to assess regarding reduced HIV rates

Reduced intravenous drug use
Early observational studies of methadone patients compared to active heroin users [73] found 80-90% reductions in daily
heroin use among those in treatment. The two most comprehensive prospective studies of continued IDU among USA
methadone patients, the Three Cities Study by Ball and Ross [62] (1991) and The Treatment Outcome Evaluation Study
(TOPS) [74], both found sharp reductions of injecting heroin within weeks of entering treatment and steady reductions in
the level of IDU as a function of time in treatment (in TOPS by 77% among those in treatment > l 8 months). Sufficient
methadone dose (6()-100 mg) is a critical factor in predicting reduced IDU and concurrent cocaine use often predicts
higher injecting; and sharing rates [55,75]. In prisons which provide methadone. similar positive results have been
observed, with reduced injecting both during incarceration and afterwards [76-78].

Reduced sharing of injecting equipment
Sharing of injecting equipment is generally related to the frequency of IDU, which is almost universally reduced by
methadone treatment. Concurrent use of cocaine and shooting galleries, however, increases the likelihood of sharing as it
does among drug users with the most marginalized social circumstances (for example, homelessness [79]). However,
cocaine use is lowered when higher doses of methadone are used and is inversely correlated with longer time in treatment
and withdrawal from the drug culture. In jails and prisons, the role of methadone is especially significant for HIV risk
because of the almost universal sharing of (contraband) injection equipment, and the high background prevalence rates of
HIV among incarcerated IDU [80], who commonly constitute 50% of prison populations in the USA and Europe.
Table 4. Studies on methadone treatment effectiveness published since 1987.

Author Year Place Study No. of Subjects Outcome
Moss et. al., 1994 San Francisco, USA Observational 2531 heterosexual HIV seroconversion rate 1.9'% ppy.
[48] intravenous drug Risk behaviour reduction over 5
users (681 HIV- years: intravenous cocaine 18%,
negative initially) shooting gallery 13% proportion
with more than 5 sex partners 15%
Methadone highly protective (less
than 1 year methadone hazard
ratio for seroconversion = 2.7: P =
0.02). Stable attendance at
methadone maintenance was
highly protective: the
seroconversion rate in subjects
with one year or more in
methadone was 1% ppy.
Novick et. al., 1990 New York, USA Retrospective 58 long-term, socially 0/58 (0%) HIV-positive: 53/58
[49] review rehabilitated (91%) at least one HBV marker.
maintained, pre-
10.3 +/- 1.7 years
heroin injecting

Vanichseni et. 1991 Bangkok, Thailand Controlled trial 240 male heroin Methadone maintenance clients
al., [50] (methadone addicts with at least more likely to complete 45 days
maintenance, 45 day Six prior detoxification treatment (P <0.0001); less likely
methadone treatment episodes to have used heroin during
detoxification) treatment (P < 0.0002); less likely
to have used heroin on the 45 day
of treatment (P <0.000007)

Ball et. al., 1988 Six programs in three Observational 633 males were Of the 388 patients still in
[51] cities in eastern USA interviewed initially treatment, 36% had not injected
(113 new admissions again after the first month of
and 520 already in treatment; an additional 22% had
treatment for at least not injected in the past year and a
6 months). further 13% had not injected for a
A year later 388 period of between 1 and 11
remained in treatment months. The remaining 29%, had
and 107 had left injected in the past month. Overall,
treatment during the 71% had not injected in the last
previous year month with similar results for
sharing (both numbers of sharers
and frequency of sharing).

Metzger et. 1993 Philadelphia, USA Observational 152 in-treatment, 103 At baseline 10% of methadone
al., [52] out-of-treatment from maintained group and 16% of out-
same neighborhood of-treatment group HIV-positive.
as methadone Over 18 months, 3.5% of
maintenance unit methadone maintained subjects
and 22% of those out-of-treatment
initially HIV-negative
seroconverted. Out-oftreatment
subjects injected drugs, shared
needles, visited shooting galleries,
and practised unsafe sex at
significantly higher rates than in-
treatment subjects.

Stark and 1993 Berlin, Germany Cross-sectional, 472 Current borrowing of syringes
Muller [53] multisite sampling was associated with younger
age, shorter history of
intravenous drug use, negative
HIV serostatus and
nonparticipation in methadone
maintenance program.

Schoenbaum 1990 New York, USA Observational 452 patients in HIV infection associated with more
et. al., [5] methadone treatment injections per month, higher
percentage of injections with used
needles, more injections with
cocaine per month, higher
percentage of injections with
needles shared with strangers or

Marmor et. 1987 New York, USA Observational 308 hospital-based 50.7% HIV-positive. HIV infection
al., [54] methadone associated with more frequent drug
maintenance or drug injection, higher proportion of
detoxification, injections in shooting galleries,
Manhattan, New York male to male sex.
Chaisson et. 1989 San Francisco, USA Observational 633 heterosexual HIV infection associated with
al., [55] intravenous drug intravenous cocaine use, drug use
users in shooting galleries, sharing of
drug injection equipment.
Multivariate analysis found black
race, daily cocaine injection by
blacks and Hispanics, all other
cocaine injection, heavy drug use
prior to entry into methadone
treatment by blacks and use of
drugs in shooting galleries were
independent predictors of HIV
infection. Methadone associated
with substantial reduction in heroin
use and some reduction in cocaine

Brown et. al., 1989 New York, USA Multisite,cross- 454 patients in Frequent use of intravenous drugs,
[56] sectional methadone duration of drug use, and duration
maintenance of drug treatment enrolment were
significantly associated with HIV
infection and complications.

Williams et. 1992 New Haven, CT, USA Observational 56 patients in Subjects in continuous treatment
al., [57] continuous reported less needle sharing, fewer
methadone needle sharing partners, fewer
maintenance sexual partners and were more
treatment, 42 with likely to be women.
methadone treatment

Selwyn et. 1987 New York, USA Cross-sectional 261 intravenous drug Continued needle sharing
al., [58] users either in behaviour associated with
methadone detention facility site and lower
maintenance scores on age knowledge
treatment or a large questionnaire. Reduced needle
detention facility in sharing among methadone
NYC maintained and subjects with
greater knowledge about AIDS.

Serpelloni et. 1994 Verona, Italy Nested case- 40 cases HIV seroconversion associated
al., [59] control study (seroconverters), 40 with number of cycles of treatment,
controls (HIV- daily dose and time out of
negative injecting treatment (univariate analysis).
drug users). Daily dose (linear effect) and time
out of treatment most important
risk factors (multivariate analysis).
Risk of HIV infection increased 1.5
times for every 3 months spent out
of treatment.

ppy, Per person-years; HBV, hepatitis B virus

Reduced sex work
Drug treatment is often found to be less effective in reducing sexual risk behavior than injecting or sharing [81], in part
because female IDU may use condoms with sexual customers, but not with their regular partners who are more likely also
to be IDU and HIV-positive [82]. However, retention in methadone treatment is associated with reduced arrest rates for
women for prostitution, a marker for higher HIV risk. And some studies [83] indicate that, for those who continue to
inject, participation in methadone treatment is associated with greater use of NSEP, which provide condoms and other
social supports favoring reduced sexual risks.

Reduced arrests/incarceration and the role of methadone in prisons
In the earliest demonstration of effects on criminal behavior, the growth of New York City's methadone treatment
capacity (in the 1970s) from 20 000 to 34 000 patients corresponded with 25 000 fewer drug arrests [84]. A 90%
reduction in drug imprisonment in Hong Kong also followed initiation of a low-threshold methadone program [85]. It is
the potential to expand methadone treatment in jails and prisons, however, that offers the greatest promise for HIV
reduction [55] because IDU continues in these settings, albeit at lower levels of use but with far fewer syringes and greater
likelihood of sharing. Furthermore, because drug users spend so much of their careers incarcerated and because relapse
frequently occurs shortly after release, they are particularly vulnerable populations for HIV infection and are prime
candidates for methadone to begin during incarceration and to be continued post release

New forms of methadone treatment
In the face of the HIV epidemic, many countries have developed low-threshold approaches which expand the availability
of methadone treatment, for example, the Netherlands' Methadone Bus and Beth Israel's interim program in New York
City. Most significantly, the rapid expansion of office-based methadone within general medical practice has become the
norm in many countries [86-88], and allowed some countries to expand rapidly the availability of methadone treatment
(for example, in Australia, tenfold increase in 10 years) and help avert the growth of HIV prevalence among IDU [89,90].
Methadone is also now being incorporated within some residential therapeutic communities [91]. In addition, a longer-
lasting version of methadone (LRAM) is now being used to reduce the frequency of dosing (and costs) and to alleviate
concerns about dispensing take-home doses of potentially dangerous narcotics.

International differences in access to methadone
The number of methadone treatment places is a critical determinant of its potential public health (or population) impact.
Worldwide, there are dramatic differences in the availability of methadone treatment: from <10 in France and Belgium to
80 in Denmark per 100 000 population [92]. Availability is 45 per 100 000 in the USA, but, with 115 000 patients in care,
10 states still have no methadone treatment [93]. While some countries (such as the Netherlands) have provided a wide
range of drug treatment options (including methadone) that reach over 75% of the IDU population, fewer than 15% of the
2 000 000 opiate addicts in the USA and Europe today have access to methadone. In eastern Europe and most of Asia,
these proportions are far lower [94].

Expansion of methadone is still limited by widespread medical misunderstanding of methadone treatment and (often) very
deep-seated hostility to the concept of ,substitution' treatment for opiate dependency. Thus, in Germany, until very
recently, doctors were barred by law from using methadone (some were even imprisoned for prescribing it) and
methadone is still in only limited use, mainly for patients with HIV infection. In Belgium, there were only a few hundred
patients on methadone treatment until 1992, when the general physicians' union successfully mounted a legal challenge to
psychiatric restraints on its use [84]. France (with >50 000 IDU) had only 50 patients in methadone treatment as late as
1993, when the medical human-rights communities and a powerful social and health insurance organization successfully
fought a resistant psychiatric profession to institute methadone treatment as a national public-health program within
general medicine [95].

In the USA, where it was first developed and adopted on a massive scale, however, methadone has lost ground.
Throughout the two decades of the AIDS epidemic among IDU in the USA, there has been no real expansion of the
methadone maintenance treatment program in the USA and, in some states, severe limits are placed on time in publicly
financed treatment, doses are subtherapeutic, and large fees are charged for private programs [96]. Strong prejudice
against methadone continues in the (generally) abstinence-oriented USA drug treatment world.

Beyond methadone
Methadone is not the only drug substitution option available to medical practitioners. Faced with severe restrictions on
methadone prescribing, German physicians developed a system of maintenance using oral codeine, with over 35 000
patients in treatment. Results have been positive. French general practitioners are prescribing buprenorphine in far greater
numbers (30 000 patients) than they do methadone. These new developments are pushing the boundaries of addiction
treatment within medical practice and offering significant therapeutic advantages over the current narrow spectrum of
available pharmacotherapies, along with the prospect of including more patients in care and effecting decisive reductions
in the spread of HIV

Most significantly, in 1995, the Swiss began a landmark program of injectable heroin maintenance for longer-term addicts
who had previously failed in treatment with methadone [97]. The 1200 patients included can visit any of 20 clinic sites
throughout Switzerland up to three times a day, to inject pharmaceutical heroin under medical supervision. The
preliminary evaluation by the Swiss Federal Health Ministry (with over 1800 person-years of data) indicates that 1-year
retention in treatment is >80%, sharp reductions have occurred in arrests and criminal activity, and there is greater
involvement with non-drug using family and friends, all while injecting a median of 400 mg of heroin each day.
Conclusion: harm reduction
These two public health strategies, NSEP and maintenance pharmacotherapy for opiate dependency (especially
methadone), are among our best available response to the AIDS epidemic in drug users. Both build on established public-
health principles of disease control, acting on one or more elements of the complex ecology of HIV among drug injectors
to reduce transmission. These elements include the vector-control approach mentioned above, separation of the potential
host (the user) from the vector (in other words, needles and syringes), and reducing the susceptibility of opiate dependent
drug users through making safer alternative forms of opiates more accessible through medical services.

The common goal of these efforts is `harm reduction'; not ending drug use, but reducing the population morbidity and
mortality rates associated with HIV [98-101]. This approach to IDU and AIDS prevention is clearly along the lines of the
steps taken to deal with other dangerous drugs, for example, alcohol. In countries that tolerate and regulate alcohol use
(rather than prohibiting it) harm-reduction strategies now include standard unit labeling, availability of low-alcohol
beverages, designated driver programs, and coordinated national media campaigns, backed up by road testing for blood
alcohol level with strict enforcement. National Harm Reduction Drug Strategies for all drugs (licit and illicit) can be seen
today in many countries [102]. These strategies have proven effective in stopping or slowing the spread of AIDS in quite
different societies (for example, the UK, Australia, Nepal, Thailand, and parts of the USA).

Today, however, instead of instituting more harm reduction services and working for social improvement that might
prevent drug abuse (for example, jobs, housing, healthcare, and schools), many nations continue building prisons, and
most addicts spend more time in prison than in treatment. In the USA, which leads the world in the rate of incarceration of
its citizens (most of it associated with the prosecution of drug users), over 3.5 million Americans are in the control of the
criminal justice system and 1.5 million behind bars, including 35% of all young African-American men [103]. These
national programs of massive prosecution and incarceration of addicts further destabilize exactly those communities most
vulnerable to drugs and AIDS, destroying social capital, and damaging many social and civic institutions vital both to
order and to public health.

In public health, we seek pragmatic solutions which produce demonstrable and reproducible results, measured in reduced
rates of morbidity and mortality. In the case of AIDS, because of the large populations and many societies at risk, the
stakes are very high. The uneven burden of HIV infection and AIDS faithfully reiterates all of our social, economic, and
political tensions and inequalities. How could it be otherwise' As the rapid globalization of the economy proceeds
unchecked in much of the world, by restraining social welfare policies and uprooting large populations it creates a lethal
ecology of poverty, social dislocation, and despair [104,105] from which, predictably, many people seek escape in drugs.
And, just as predictably, others will seek to profit from the physical dependency that these drugs engender. So it will take
more than sterile needles and methadone to stop the AIDS epidemic associated with drug use. But while we (literally)
await the millennium and AIDS continues to spread, we have a clear obligation to employ the best public-health tools
available to protect the millions of people affected.


The authors wish to thank Alicia Knowles. Ann Arbor, Michigan, for her assistance.
1.Mann J, Tarantola D: AIDS In The World 11. New 17. Hagan H, Des Jarlais DC, Friedman SR, Purchase
York: Oxford; 1996. D: Risk for human immunodeficiency virus
2. Drucker E: Harm reduction: a public health and hepatitis B virus in users of the Tacoma
strategy. Curr issues Public Health 1995, 1:64- syringe exchange program. In: Proceedings:
70. Workshop on Needle Exchange and Bleach
3. Ball A, Des Jarlais D, et al.: Multi-Center Study on Distribution Programs. Washington: National
Drug Injecting and Risk of HIV Infection. Geneva: Academy Press; 1994:24-34.
WHO; 1991. 18. Paone D, Des Jarlais DC, Caloir 5, Freedman P,
4. Grund JPC, Stern LS, Kaplan CD, Adriaans NFP, Ness 1, Friedman SR: New York City syringe
Drucker E: Drug use contexts and HIV- exchange: an overview. In: Proceedings:
consequences: the effect of drug policy on Workshop on Needle Exchange and Bleach
patterns of everyday drug use in Rotterdam Distribution Programs. National Academy Press,
and the Bronx. Br ] Addiction 1992, 87:41-52. Washington, 1994:47-63.
5. Schoenbaum EE, Hartel D, Selwyn PA, et al.: Risk 19. Oliver K, Maynard H, Friedman SR, Des Jarlais
factors for human immunodeficiency virus DC: Behavioral and community impact of
infection in intravenous drug users. N Engl J the Portland syringe exchange program. in:
Med 1989, 321:874-879. Proceedings:' Workshop on Needle Exchange and
6. Watters JK: Historical perspective on the use Bleach Distribution Programs. National Academy
of bleach in HIV/ AIDS prevention. J Acquir Press, Washington, 1994:35-46.
Immun Defic Syndr 1994, 7:743-747. 20. Schoenbaum EE, Hartel DM, Gourevitch MN:
7. Haverkos HW, Jones ST: HIV, drug-use Needle exchange use among a cohort of
paraphernalia, and bleach. J Acquir injecting drug users. AIDS
Immune Defic Syndr 1994, 7:741-742. 1996,10:17291734.
8. Normand J, Vlahov D, Moses L (eds): Preventing 21. Vlahov D, Junge B, Brookmeyer R: Reductions in
HIV Transmission. The Role of Sterile Needles and high-risk drug use behaviours among
Bleach. Washington: National Research Council, participants in the Baltimore needle
IOM National Academic Press; 1995. exchange program. J Acquir Immune Defic
9. Kaplan EH: Probability models of needle Syndr Hum Retrovir 1997, 16:400-406.
exchange. Operations Res 1995, 43:558-569. 22. Hankins C, Gendron S, Bruneau 1, Roy E:
10. US Public Health Service: HIV Prevention Bulletin: Evaluating Montreal's needle exchange CACTUS-
Medical Advice For Persons Who Inject Illicit Montreal. In: Proceedings:
Drugs. Atlanta: US Public Health Service; 1997. 23. van Ameijden EJC, van den Hoek JAR, Coutinho
11. Lurie P, Reingold AL, Bowser B, et al.: The Public RA: Injecting risk behavior among drug users in
Health Impact of Needle Exchange Programs in the Amsterdam., 1986 to 1992, and its
United States and Abroad. San Francisco: relationship to AIDS prevention programs.
University of California; 1993. Am J Public Health 1994, 84:275-281
12. Watters JK, Estilo MJ, Clark GL, Lorvick J: 24. Frischer M, Elliot L: Discriminating needle
Syringe and needle exchange as HIV/AIDS exchange attenders from non-attenders.
prevention for injection drug users. JAMA Addiction 1993, 88:681-687.
1994, 271:115-120. 25. Keene J, Stimson GV, Jones S, Parry-Langdon N:
13. Hunter GM, Donoghoe MC, Stimson GV, Rhodes T, Evaluation of syringe-exchange for HIV
Chalmers CP: Changes in the injecting risk prevention among injecting drug users in
behaviour of injecting drug users in rural and urban areas of Wales. Addiction 1993,
London, 1990-1993. AIDS 1995, 9:493-501. 88:10631070.
14. Peak A, Rana S, Maharjan SH, Jolley D, Crofts N: 26 Hahn JA, Vranizan KM, Moss AR: Who uses
Declining risk for HIV among injecting drug needle exchange? A study of injection drug
users in Kathmandu, Nepal: the impact of a users in treatment in San Francisco, 1989-
harm-reduction programme. AIDS 1995, 90. J Acquir Immune Defic Syndr Hum Retrovir
9:10671070. 1997, 15:157164.
15. Des Jarlais DC, Friedman SR, Sotheran JL, et al.: 27. Hagan H, Des Jarlais DC, Friedman SR, Purchase D,
Continuity and change within an HIV Alter MJ: Reduced risk of hepatitis B and
epidemic: injecting drug users in New York hepatitis C among injection drug users in the
City, 1984 through 1992. )AMA 1994, Tacoma syringe exchange program. Am J
271:121-127. Public Health 1995, 85:1531-1537.
16. Fennema JSA, Wiessing LG, Coutinho RA, van den 28. Des Jarlais DC, Hagan H, Friedman SR, et al.:
Hoek A, Houweling H, van Ameijden EJC: Maintaining low HIV seroprevalence in
Trends in injection drug use in a city with populations of injecting drug users. JAMA
harm reduction. In: HIV infection among 1995, 274:1226-1231.
drugusers and the potential for heterosexual 29. Hurley SF, Jolley DJ, Kaldor JM: Effectiveness of
spread. Edited by Fennema H. Wageningen: needle-exchange programmes for
Ponsen and Looijen BV; 1997:51-62. prevention of HIV infection. Lancet 1997,
349: 1797-1800.
30. Kaplan EH, Khoshnood D, Heimer R. A decline in
HIV infected needles returned to New
Haven's needle exchange program: client
shift or needle exchange? Am J Public Health
1994, 84:1991-1994.
31. Lurie P, Drucker E: An opportunity lost: HIV 45. Ward J, Mattick R, Hall W: Key Issues In
infections associated with lack of a national Methadone Treatment. Randwick, Australia:
needle-exchange programme in the USA. University of NSW Press; 1992.
Lancet 1997, 349:604-608. 46. Dole VP, Nyswander M: Methadone
32. Patrick DM, Strathdee SA, Archibald CP, et al.: maintenance treatment: a ten-year
Determinants of HIV seroconversion in perspective. ]AMA 1976, 235:80-84.
injection drug users during a period of 47. Drucker E: From morphine to methadone. In:
rising prevalence in Vancouver. Int J STD Harm Reduction: Policy and Practice. Edited by
AIDS 1997, 8:437445. Inciardi J, Harrison L. Thousand Oaks, CA: Sage;
33. Bruneau J, Lamothe F, Franco E, et al.: High 1998:in press.
rates of HIV infection among injection drug 48. Moss AR, Vranizan K, Gorter R, Bacchetti P,
users participating in needle exchange Watters J, Osmond D: HIV seroconversion in
programs in Montreal: results of a cohort intravenous drug users in San Francisco,
study. Am J Epidemiol 1997, 146:994-1002. 1985-1990. AIDS 1994, 8:223-231.
34. Des Jarlais DC, Marmor M, Paone D, et al.: HIV 49. Novick DM, Joseph H, Croxson TS, et al.:
incidence among injecting drug users in Absence of antibody to human
New York City syringe-exchange immunodeficiency virus in long-term,
programmes. Lancet 1996, 348:987-991. socially rehabilitated methadone
35. Strathdee SA, Patrick DM, Currie SL, et al.: maintenance patients. Arch Intern Med 1990,
Needle exchange is not enough: lessons 150:97-99.
from the Vancouver injecting drug use 50. Vanichseni S, Wongsuwan B, Choopanya K,
study. AIDS 1997, 11:1`59-1`65. Wongpanich K: A controlled trial of
36. Dolan K, Donoghoe M, Jones S, Stimson G: A methadone maintenance in a population of
cohort study of syringe exchange clients and intravenous drug users in Bangkok:
other drug injectors in England, 1989 to 1990. implications for prevention of HIV lnt J
London: Centre for Research on Drugs and Addictions 1991, 26:1313-1320.
Health Behaviour; 1991. 51. Ball JC, Lange WR, Myers CP, Friedman SR:
37. Kahn JG: Do NEPs affect rates of HIV drug Reducing the risk of AIDS through
and/or sex risk behaviors? In: The public methadone maintenance treatment. J Health
health impact of needle exchange programs in Soc Behav 1988, 29:214-226.
the United States and abroad. Edited by Lurie P, 52. Metzger DS, Woody GE, McLellan AT, et al.:
Reingold AL, Bowser B, et al. San Francisco: Human immunodeficiency virus
University of California; 1993:403-427. seroconversion among intravenous drug
38. Klee H, Morris J: The role of needle exchange users in- and out-of-treatment: an 18-
in modifying sharing behaviour: cross- month prospective follow-up. J Acquir
study comparisons 1989-1993. Addiction Immun Defic Syndr 1993, 6:1049-1056.
1995, 990:1635-1645. 53. Stark K, Muller R: HIV prevalence and risk
39. Bruneau J, Franco E, Lamothe F: Assessing behaviour in injecting drug users in Berlin.
harm reduction strategies: the dilemma of Foren Sci Int 1993, 62:73-81.
observational studies. Am J Epidemiol 1997, 54. Marmor M, Des Jarlais DC, Cohen H, et al.: Risk
146:1007-1010. factors for infection with human
40. Lurie P: Le mystere de Montreal. Am J Epidemiol immunodeficiency virus among intravenous
1997, 146: 1003-1006. drug abusers in New York City. AIDS 1987,
41. Lurie P, Jones TS, Foley J: A sterile syringe for 1:39-44.
every drug user injection: how many 55. Chaisson RE, Bacchetti P, Osmond D, Brodie B,
injections take place annually and how Sande MA, Moss AR: Cocaine use and HIV
might pharmacists contribute to syringe infection in intravenous drug users in San
distribution? J Acquir Immune Defic Syndr Francisco. ]AMA 1989, 261:561-565.
Hum Retrovir 1998, in press. 56. Brown LS Jr, Chu A, Nemoto T, Ajuluchukwu D,
42. Gostin LO, Lazzarini Z, Jones TS, Flaherty K: Primm BJ: Human immunodeficiency virus
Prevention of HIV/ AIDS and other blood- infection in a cohort of intravenous drug
borne diseases among injection drug users: users in New York City. Demographic,
a national survey on the regulation of behavioral, and clinical features. NY State J
syringes and needles. JAMA 1997, 277:53-62. Med 1989, 89:506-510.
43. Valleroy LA, Weinstein B, Jones TS, Groseclose 57. Williams AB, McNelly EA, Williams AE, D'Aquila
SL, Rolfs RT, Kassler WJ: Impact of increased RT: Methadone maintenance treatment and
legal access to needles and syringes on HIV type 1 seroconversion among injecting
community pharmacies' needle and syringe drug users. AIDS Care 1992, 4:35-41.
sales- Connecticut, 1992-1993. J Acquir 58. Selwyn PA, Feiner C, Cox CP, Lipshutz C, Cohen
Immune Defic Syndr Hum Retrovir 1995, 10:73- RL: Knowledge about AIDS and high-risk
81. behavior among intravenous drug users in
44. Groseclose SL, Weinstein B, Jones TS, Valleroy New York City. AIDS 1987, 1:247-254.
LA, Fehrs L), Kassler WJ: Impact of increased 59. Serpelloni G, Carried MP, Rezza G, Morganti S,
legal access to needles and syringes on Gomma M, Binkin N: Methadone treatment as
practices of injecting-drug users and police a determinant of HIV risk reduction among
officers Connecticut, 1992-1993. J Acquir injecting drug users: a nested case-control
Immune Defic Syndr Hum Retrovir 1995, 10:82-89. study. AIDS Care 1994, 6:215-220.
60. Hubbard RL, Marsden ME, Rachal JV, Harwood HJ, 76. Dolan K, Wodak A: An international review of
Cavanaugh ER, Ginzberg HM: Drug Abuse methadone in prisons. Addiction Res 1996,
Treatment: A National Study of Effectiveness. Chapel 4:85-97.
Hill, NC: University of North Carolina Press; 77. Wale S, Gorta A: Pre-Release Methadone Program.
1989. Sydney: NSW Department of Corrective
61. Gerstein DR, Hendrick HJ: Treating Drug Services; 1987 [Research Publication No. 11.
Problems. Vol. 1. Washington: National Academy 78. Gorta A: Monitoring The NSW Methadone Program
Press; 1990. 1986-1997. Sydney: NSW Department of
62. Ball JC, Ross A: The Effectiveness of Methadone Corrective Services; 1987 [Research Publication
Maintenance Treatment: Patients, Programs, No. 2].
Services and Outcome. New York: Springer- 79. Klee H, Faugier J, Hayes C, Boulton T, Morris J:
Verlag; 1991. AIDS-related risk behaviour, polydrug use
63. Gunne L-M, Gronbladh L: The Swedish and temazepam. Br J Addiction 1990,
methadone maintenance program: a 85:1125-1132.
controlled study. Drug Alcohol Dependence 80. Hall W, Ward J, Mattick RE: Methadone
1981, 24:31-37. maintenance in prisons: the NSW
64. Blix O, Gronbladh L: AIDS and IV heroin addicts: experience. Drug Alcohol Rev 1993, 12:193-
the preventive effect of methadone in Sweden. 203. 81. Donoghoe MC: Sex,
4th international Conference on AIDS. Stockholm, HIV and injecting drug users. Br J Addict
1988. 1992, 87:405-416.
65. Hall W: Australia's Methadone Program. Randwick, 82. Van den HoekA, van Haastrecht HJ, Coutinho RA:
Australia: National Drug and Alcohol Research Heterosexual behavior of intravenous drug
Center ARC; 1997. users in Amsterdam: implications for the
66. Dolan K, Wodak A, Hall W, Gaughwin M, Rae F: AIDS epidemic. AIDS 1990, 4:449-453.
HIV risk behaviour before during and after 83. Hartgers C, van den HoekA, Krijnen P, Coutinho
imprisonment in New South Wales Addiction RA: HIV prevalence and risk behaviors
Res 1997, 4:151-160. among injecting drug users who participate
67. Herman M, Gourevitch MN: Integrating primary in' low-threshold' methadone programs in
care and methadone maintenance Amsterdam. Am J Public Health 1992, 82:547-
treatment: Implementation issues. ] Addict 551.
Dis 1997, 16:91-102. 84. Joseph H: The criminal justice system and
68 General Accounting Office: Methadone opiate addiction: a historical perspective.
Maintenance: Some Treatment Programs Are In: Compulsory Treatment of Drug Abuse:
Not Effective; Greater Federal Oversight Needed. Research and Clinical Practice. Edited by
Washington: US General Accounting Office; Leukefeld CG, Tims FM. Maryland: National
1990 (GAO/HRD-90-1041. Institute on Drug Abuse; 1984:106-125 INIDA
69. D'Aunno T, Vaughn TE: Variations in research monograph 861.
methadone practice. /AMA 1992, 267:253- 85. Newman RG, Whitehill WB:Double-blind
258. comparison of methadone and placebo
70. Abdul-Quader AS, Friedman SR, Des Jarlais D, maintenance treatments of narcotic addicts
Marmor MM, Maslansky R, Bartelme S: in Hong Kong. Lancet 1979,ii:485-488.
Methadone maintenance and behaviour bv 86. Strang J, Farrell M (eds): Heroin Addiction and
intravenous drug users that can transmit Drug Policy. Oxford; 1994.
HIV. Contemporary Drug Problems 1987, 87. Reisinger M, Picard E: Methadone in Belgium:
14:425-434. physicians gaining therapeutic freedom.
71. Yancovitz SR, Des Jarlais DC, Peyser NP, et al.: A Addiction Res 1996, 3:369-371.
randomized trial of an interim methadone 88. Carpentier J: La Toxicomanie a L'Heroine en
maintenance clinic. Am J Public Health 1991, Medecine Generale. Paris: Ellipses; 1994.
81:1185-1191. 89. Wodak A: HIV infection and IDU in Australia:
72. Buning E, van Brussel GHA, van Santen GW: The responding to a crisis. J Drug Issues 1992,
'Methadone by Bus' Project in Amsterdam. 22:549-562.
Br J Addictions 1990, 85:12471250. 90. Stimson G: AIDS and IDU in the UK. 1988-
73. Anglin MD, McGlothin WH: Outcome of narcotic 1993. The policy response and prevention of the
addict treatment in California. In: Drug epidemic. 1995.
Abuse Treatment Evaluation: Strategies, 91 Introduction of Methadone into a therapeutic
Progress and Prospects. Edited by Timms FM, Community in Leicester UK, 1997.
Ludford JP. Maryland: National Institute on Drug 92. Farrell M: Methadone in the European Community.
Abuse; 1984 INIDA Research Monograph, 511. Brussels: EU; 1996.
74. Hubbard RL, Rachal JV, Craddock SG, Cavanaugh 93. Parrino MW (ed): State Methadone Treatment
ER, Treatment Outcome Prospective Study Guidelines. Center for Substance Abuse Treatment,
(TOPS): Client characteristics and Substance Abuse and Mental Health Administration.
behaviours before, during and after Rockville: US Public Health Service; 1993
treatment. In: Drug Abuse Treatment (publication SMA 93-1991].
Evaluation: Strategies, Progress and Prospects. 94. Grund JP: Harm reduction and Methadone Treatment
Edited by Timms FM, Ludford JP. Maryland: in Eastern Europe. Open Society Institute; 1997
National Institute on Drug Abuse; 1984 INIDA 95. Exclusion: Report on Tri Ville Paris
Research Monograph, 51]. Conference, 1992. In: Les Temps Modernes, No.
96. Rosenbaum M, Murphy S, Beck ): Money for 101. Erickson PG, Riley D, Cheung E (eds): Harm
methadone. 1 Psychoactive Drugs 1987, Reduction: A New Direction For Drug; Policy and
19:13-19. Programs. Toronto: University of Toronto:
97. Uchtenhagen A, Gutzwiller F, Dohler-Mikola A 1997.
(eds): Program For Medical Prescription of 102. Nadelmann E: Common Sense Drug Policies.
Narcotics. Interim Report. Zurich: University of Foreign Affairs; 1998.
Zurich; 1996. 103. Mauer M: Report on Drug Incarcerations in the US.
98. Nadelmann E, McNeely I. Drucker E: Harm Washington: The Sentencing Project; 1996.
reduction: an international perspective. In: 104. Lurie P, Hintzen P, Lowe RA: Socioeconomic
Substance Abuse: A comprehensive Textbook. obstacles to HIV prevention and treatment
Edited by Lowinson IH, Ruiz P. Millman RB. in developing countries: the roles of the
Langrod ). New York: Wiley; 1997. International Monetary Fund and the
99. O'Hare P, Matthews A, Buning E, Drucker E (eds): World Bank. AIDS 1995, 9:539-546.
The Reduction of Drug Related Harm. London: 105.. Drucker E: Epidemic in the war zone: AIDS
Routledge; 1994. and community survival in New York City.
100. Heather N, Wodak A, Nadelmann E. O'Hare I' Int I Health Services 1990. 20:601615.
(eds): Harm Reduction: From Faith to Science.
London: Whurr; 1996.