Schaffer Online Library of Drug Policy Sign the Resolution
Contents | Feedback | Search
DRCNet Home
| Join DRCNet
DRCNet Library | Schaffer Library |Miscellaneous


THE LANCET ******** Vol 349 - June 21, 1997: 1797-1800


Effectiveness of needle-exchange programmes for prevention of HIV infection

Susan F Hurley, Damien J Jolley, John M Kaldor



Background Needle-exchange programmes (NEPs) are potentially a key strategy for containing the spread of HIV infection among injecting drug-users, but their implementation has been limited by uncertainty about their effectiveness. We used an ecological study design to compare changes over time in HIV seroprevalence in injecting drug users worldwide, for cities with and without NEPs.

Methods Published reports of HIV seroprevalence in injecting drug users were identified, and unpublished information on HIV seroprevalence for injecting drug users entering drug treatment in the USA between 1988 and 1993 was obtained from the Centers for Disease Control and Prevention. Details of the implementation of NEPs were obtained from published reports and experts. For each of the 81 cities with HIV seroprevalence data from more than 1 year and NEP implementation details, the rate of change of seroprevalence was estimated by regression analysis. The average difference in this rate for cities with and without NEPs was calculated.

Findings On average, seroprevalence increased by 59% per year in the 52 cities without NEPs, and decreased by 58% per year in the 29 cities with NEPs. The average annual change in seroprevalence was 11% lower in cities with NEPs (95% Cl -176 to -39, p=0004).

Interpretation A plausible explanation for this difference is that NEPs led to a reduction in HIV incidence among injecting drug users. Despite the possibility of confounding, our results, together with the clear theoretical mechanisms by which NEPs could reduce HIV incidence, strongly support the view that NEPs are effective.

Lancet 1997; 349: 1797-1800



The HIV epidemic in injecting drug users has reached crisis proportions, moving from its original focal points in Europe and North America to major outbreaks in Asian cities. 1 In the USA alone, use of injection-drugs was the direct or indirect cause of almost 26 000 cases of AIDS in 1995, comprising 35% of all new cases. 2 Needle-exchange programmes (NEPs) are potentially a key strategy for containing this epidemic, but they are not universally accepted. Government-funded NEPs are well established in several countries, including the UK, the Netherlands, and Australia. 3 However, there is still strong opposition to their implementation in the USA, where federal funding of NEPs has been prohibited until "the Surgeon General determines that such programs are effective in preventing the spread of HIV and do not encourage the use of illegal drugs". 4 ! State prescription and paraphernalia laws, which restrict the possession and distribution of needles and syringes, are a further impediment to the operation of NEPs in the USA. 5 Although NEPs had been established in 55 US cities by 1994, 4 many operate illegally, with tenuous funding, and on a small scale. 4,5

Several observational studies have attempted to assess the effectiveness of NEPs in preventing HIV transmission. Most of these studies relied on surrogate endpoints such as self-reported risk behaviours, 4,5 or the occurrence of infections with hepatitis B or C virus. 6 The few studies that used HIV endpoints either compared HIV seroincidence rates or HIV seroprevalence in small numbers of NEP participants and non-partcipants, or correlated time trends in HIV seroprevalence for a single city with the introduction of NEPs. 4,7 Three reviews of NEP effectiveness commissioned by the US federal government since 1993 4,5,8 found that the evidence in favour of the effectiveness of NEPs was weak by accepted scientific standards. The goal of our study was to examine the effectiveness of NEPs with an approach that overcame so! me of the methodological limitations of previous studies. We used an ecological study design to compare changes over time in HIV seroprevalence in injecting drug users worldwide, for cities with and without NEPs.


Our analysis compared cities that had NEPs with those that had no NEPs. Programmes that distributed needles and syringes free of charge were regarded as NEPs, irrespective of how they operated (ie, from a fixed or mobile site, whether return of a used needle was mandatory, or the range of additional preventive services offered). We identified published reports of HIV seroprevalence surveys in injecting drug users that included 50 or more drug users and details of the implementation of NEPs in the corresponding city were obtained from reports and through communication with public-health officials and NEP researchers. We searched entries on Medline from between 1984 and May, 1994, HIV seroprevalence studies of injecting drug users and implementation of NEPs. Unpublished information on HIV seroprevalence was obtained from the Centers for Disease


Department of Public Health and Community Medicine, University of Melbourne, Australia (S F Hurley PhD, D J Jolley MSc), and National Centre In HIV Epidemiology and Clinical Research, University of Now South Wales, Sydney (J M Kaldor PhD)

Correspondence to: Dr Susan F Hurley, Research and Development Division, CSL Limited, 45 Poplar Rd, Parkville, Victoria 3052, Australia (e-mail:



THE LANCET ******** Vol 349 - June 21, 1997 Page 1798

Characteristics of cities

Cities with NEPs

Cities without NEPs

Location of cities
North America



South America






Asia and the South Pacific






Surveys, by type of recruitment site
Drug treatment centres






Health services



Other sites






Characteristics of surveys
Average number of surveys per city



Total number of drug users surveyed



Average number of drug users surveyed per city



Average seroprevalence period, in years, per city



Average HIV seroprevalence at start of seroprevalence periods (%)



Average HIV seroprevalence at end of seroprevalence periods (%)



Table 1: Characteristics of cities and HIV seroprevalence surveys, by NEP status

Control and Prevention on injecting drug users entering drug treatment in 27 cities in the USA from 1988 to 1993. 9

Cities were included in our analysis if HIV seroprevalence had been measured in injecting drug users in 2 or more calender years, and basic information on NEP implementation was available. We defined a survey as a measurement of HIV prevalence among injecting drug users in a single city at a single point in time. If data were gathered during a specific time period, the midpoint was used as the survey date. For each survey in a given city and year, the number of drug users, percentage seropositive for HIV, and type of recruitment site or sites were recorded. Population size, estimated number of injecting drug users and indices of NEP scope (presence/absence and the number of needles distributed in each year) for the relevant city were recorded from published material and information provided by experts.

For each city, the seroprevalence period was defined as the time in years from the first to the last available survey. The rate of change of HIV seroprevalence over this period was estimated by regression analysis, with the equation: logit(pij)~aj+Bjtij , where pij is the ith measurement of prevalence in a city j. 10 The parameter Bj represents the rate of change in prevalence on a log odds scale. Calender years, tij , were centred to 1990, so that the intercept terms represented the predicted seroprevalence in this year.

Average slopes (rate of change in HIV seroprevalence) were then calculated for cities with NEPs established during the period spanned by the surveys, and those without NEPs. The averages were weighted by the inverse of the squared SE of Bj, to reflect the relative accuracy of each city's slope estimate. The difference in average slopes between the two groups was calculated and tested for significance with a linear model. 10 To assess the effect of introducing NEPs when HIV seroprevalence was still low, a subgroup analysis was done among cities in which the first measured seroprevalence was less than 10% and the seroprevalence period was more than 3 years. The average annual rate of change of seroprevalence was then compared between cities in which no NEP was introduced during the seroprevalence period and those in which an NEP had been introduced when seroprevalence among injecting drug users was below 10%.


About 3500 published papers were reviewed, and information that satisfied the inclusion criteria for the study was obtained from 214 of these (references available from the authors on request). A total of 81 cities with HIV seroprevalence measurements from more than 1 year and information on NEP implementation were identified. 54% of these cities were in North America, 32% in Europe, and 124% in Asia and the South Pacific (table 1). In these cities, 1046 surveys of HIV prevalence involving 332892 drug users had been done between 1980 and 1993, 75% in drug treatment centres (table 1). Some serum specimens had been collected and stored, and were analysed when HIV tests became available. There were NEPs during the seroprevalence period in 29 (36%) cities; the mean duration of NEP implementation was 38 years (range 1-8). Seroprevalence was slightly higher at! the start of the seroprevalence period in cities without NEPs than in cities with NEPs, and by the end of the seroprevalence periods average HIV seroprevalence had changed little in cities with NEPs, but had increased in cities without NEPs (table 1).

To illustrate the regression analyses, the results for two cities are shown in figure 1. In Bern, Switzerland, where





THE LANCET ******** Vol 349 - June 21, 1997 Page 1799


Cities with NEPs

Cities without NEPs

Number of cities with increase in seroprevalence according to regression modelling*

14 (48%)

30 (58%)

Estimated annual rate of change (%)
in seroprevalence (95% CI)
All cities

1 to 14)

5 to 126)

Cities with intial seroprevalence <10% and seroprevalence periods longer than 3 years (95% CI)


(-212 to 242)


(30 to 312)

*Slope of plot of estimated odds of seroprevalence vs time.
Table 2: Effect of NEPs on HIV seroprevalence

NEPs were not operating during the seroprevalence period, the rate of change of the log odds of seroprevalence was estimated to be 0677 per year (SE 0123), corresponding to a 968% increase each year (95% CI 538-1500). In Glasgow, UK, NEPs were introduced in 1987. The estimated rate of change in the log odds of seroprevalence was -0286 (0042), corresponding to a 249% annual decrease in seroprevalence (-183 t! o -309).

Seroprevalence was estimated to have increased in 58% of cities without NEPs and 48% of cities with NEPs (table 2). The regression model showed that seroprevalence increased on average by 59% per year in the 52 cities without NEPs, and decreased by 58% per year in the 29 cities with NEPs--- a difference of 11% (-176 to -39, p=0004). The number of needles distributed over the seroprevalence period per 1000 injecting drug users was known for 21 cities, but was not a significant predictor of the rate of change of seroprevalence.

When analysis was restricted to cities with seroprevalence periods of more than 3 years with an initial seroprevalence below 10% we found that, on average, seroprevalence decreased by 1%, in the eight cities with NEPs compared with a 16% increase in the 18 cities without NEPs (figure 2), but this difference (-148 [95% CI -34 to 10]) was not significant. However, the regression modelling gave estimates of more than 10% for seroprevalence at the start of the period for four cities without NEPs, and one city with NEPs, apparently because the earliest surveys included fewer drug users and were consequently given less weight in estimation of slopes. Any bias would have caused the effectiveness of NEPs to be underestimated.


We found that in cities with NEPs HIV seroprevalence among injecting drug users decreased on average, whereas in cities without NEPs HIV seroprevalence increased. A plausible explanation for this difference is that the NEPs led to a reduction in HIV incidence among injecting drug users.

NEPs have the potential to decrease directly HIV transmission by lowering the rate of needle sharing 11 and the prevalence of HIV in needles available for reuse, 12 as well as indirectly through activities such as bleach distribution, referrals to drug treatment centres, provision of condoms., and education about risk behaviour. Although these mechanisms have strong theoretical support, the published evidence for NEP effectiveness is limited. Previous studies of the effect of NEPs on HIV incidence used observational designs or statistical models.

Observational designs included case studies; crosssectional, serial cross-sectional, and cohort studies (often without comparison groups); and case-control studies. 4,5 Only one study assessed the impact of NEPs on HIV incidence. Des Jarlais and colleagues 7 estimated that the hazard for incident HIV infection was 33 for injecting drug users in four high-seroprevalence cities without NEPs, compared with continuous users of NEPs in New York City. One case study investigated HIV prevention activities for five cities with low seroprevalence, but did not formally compare these with other cities that had high seroprevalence. 13 The most frequently cited statistical model for assessment of NEP effectiveness was developed by the New Haven NEP evaluators, and is based on the theory that NEPs decrease HIV transmission rates by lowering the time that needles are in circulation. 14

The conclusion of a 1993 review by a University of California team' was that NEPs are associated with decreased HIV drug risk behaviour and are not associated with negative outcomes, but that there is no clear evidence that they decrease HIV infection rates. 5 Few new data were available for the most recent US review by the Panel on Needle Exchange and Bleach Distribution Programs, 4 which concluded that NEPs are effective, but acknowledged that the evidence was weak.

Our study is distinguished from previous work by its worldwide scope and its design, which compares changes in HIV seroprevalence in cities with and without. NEPs, rather than changes within a single city. Nevertheless, our analysis clearly has limitations. First, the seroprevalence data were collected according to different protocols in diverse populations. However, any inaccuracies would be unlikely to differ systematically between cities with NEPs and those without, so the consequent bias would be in the direction of underestimating the effectiveness of NEPs. Second, cities were selected for analysis by the existence of HIV seroprevalence surveys, and bias may have been introduced by the decision to do a survey in a particular

THE LANCET ******** Vol 349 - June 21, 1997 Page 1800

city. Furthermore, HIV seroprevalence may have remained low in some of the cities with NEPs, irrespective of their introduction. Finally, implementation of NEPs was almost certainly confounded by the implementation of other HIV prevention strategies, and the sale of injection equipment through pharmacies is one such strategy that could work similarly to NEPs. Therefore, the difference in rate of change of seroprevalence between cities with and without NEPs may not be due solely to NEPs.

Despite these limitations, our study provides evidence that NEPs reduce the spread of HIV infection. With the theoretical mechanisms by which NEPs could reduce HIV incidence, and the interpretation of previous studies by the Panel on Needle Exchange and Bleach Distribution Programs, 4 the view that NEPs are not effective no longer seems tenable.


This work was supported by grants from Commonwealth AIDS Research Grants Committee, AIDS/Communicable Diseases Branch of the Commonwealth Department of Human Services and Health, and National Health and Medical Research Council of Australia (Neil Hamilton Fairley Fellowship Number 927215 to SH). The National Centre in HIV Epidemiology and Clinical Research is supported by the Australian National Council on AIDS through the Commonwealth AIDS Research Grants Committee.


We thank D Prevots for providing CDC seroprevalence data; for literature abstraction, data entry, and research assistance, R Billi, W Blake, M Jenkins, M Kermode, S Quirk, S Robson, and M Stebbing; for information regarding needle exchange programmes, A Abdul-Quader, P Aeberhard, M Rodriguez-Arenas, M Back, 0 Baker, R Braithwaite, M Batchelor, M Breda, E Buning, F Byers, J Byrne, N Crofts, K Dolan, P Dunlevy, J Gervasoni, S Friedman, J Golz, M Hare, Hartford Needle Exchange Program, Housing Works Inc, H Hagan, R Kemp, K Kerry, M Kinzly, L Lax, P Lurie, B Mackenzie, G Mathews, G McBride, M McDonald, R Moodie, A Morrison, K Nairn, New Haven Needle Exchange Programme, Olympia Needle Exchange, S Nimmo, P O'Hare, K O'Reilly, V Otte, C Paul, C Price, D Purchase, G Rezza, L Ribeiro, M Schemat, S Stafford, G Stimson, W Stronach, 'Me Exchange (Fairbanks), J Tranchina, T Trethowan, P Veugelers, K Viravaidya, 0 Westcott, J Wignell, J Witton, A Wodak, and M Zaccarelli; for discussions ! regarding analysis of NEP effectiveness, T Fleming and S Self.




1 Des Jarlais DC, Friedman SR, Choopanya K, Vanichseni S, Ward TP. International epidemiology of HIV and AIDS among injecting drug users. AIDS 1992; 6: 1053-69.

2 Centers for Disease Control and Prevention. AIDS associated with injecting-drug use-United States, 1995. MMWR Morb Mortal Wkly Rep 1996; 45: 392-97.

3 Stimson GV. Syringe-exchange programmes for injecting drug users. AIDS 1989; 3: 253-60.

4 Panel on Needle Exchange and Bleach Distribution Programs. Preventing HIV transmission: the role of sterile needles and bleach. Washington: National Academy Press, 1995.

5 Institute for Health Policy Studies. The public health impact of needle exchange programs in the United States and abroad. Berkeley: School of Public Health, University of California, 1993.

6 Hagan H, Des Jarlais DC, Friedman SR, Purchase D, Alter MJ. Reduced risk of hepatitis B and hepatitis C among injection drug users in the Tacoma syringe exchange program. Am J Public Health 1995; 85: 1531-37.

7 Des Jarlais DC, Marmor M, Paone D, et al. HIV incidence among injecting drug users in New York City syringe-exchange programmes. Lancet 1996; 348: 987-91.

8 US General Accounting Office. Needle exchange programs: research suggests promise as an AIDS prevention strategy. Washington, DC: US Government Printing Office, 1993.

9 Prevots, DR, Allen DM, Lehman S, Green TA, Peterson, LR, Gwinn M. Trends in human immunodeficiency virus seroprevalence among injection drug users entering drug treatment centers, United States, 1988-1993. Am J Epidemiol 1996; 143: 733-42.

10 Clayton D, Hills M. Statistical models in epidemiology. Oxford: Oxford University Press, 1993.

11 Nicolosi A, Leite ML, Musicco M, Molinari S. Parenteral and sexual transmission of human immunodeficiency virus in intravenous drug users: a study of seroconversion. Northern Italian Seronegative Drug Addicts (NISDA) Study. Am J Epidemiol 1992; 135: 225-33.

12 Kaplan EH, Heimer R. A circulation theory of needle exchange. AIDS 1994; 8: 567-74.

13 Des Jarlais DC, Hagan H, Friedman S, et al. Maintaining low HIV seroprevalence in populations of injecting drug users. JAMA 1995; 274:1226-31.

14 Kaplan EH. A method for evaluating needle exchange programmes. Stat Med 1994; 13: 2179-87.




Library Highlights

Drug Information Articles

Drug Rehab