The impact of the needle and syringe-exchange programme in Amsterdam on injecting risk behaviour
Christina Hartgers, Ernst C. Buning*, Gerrit W. van Santen*, Annette D. Verster and Roel A. Coutinho

We interviewed a group of 145 injecting drug users (IDUs) in Amsterdam about their drug use, participation in a needle-exchange programme and needle sharing. Approximately 1 year later 60 IDUs were followed up. IDUs who exchange regularly ('exchangers') are older, inject longer and are more often in contact with methadone programmes. Exchanging is associated neither with an increase in injecting nor with lending needles to other IDUs. The risk level of injecting of the exchangers is much lower than that of the non-exchangers. From this it can be concluded that a needle exchange is an effective prevention programme against the spread of HIV infection. However, efforts have to be made to reach the group of younger short-term injectors and those IDUs who are not in contact with methadone-maintenance programmes. Since there are indications that regular injectors in particular exchange, and since young male injectors are more at risk of borrowing independent of exchanging, it is argued that an exchange programme should be complemented with other prevention approaches, i.e. intensive counselling and the spread of leaflets with information on cleaning used needles with bleach.

Introduction

Studies on injecting drug users (IDUs) and AIDS have mentioned the ready availability of new needles as a method of preventing needle sharing and the spread of HIV infection among this group [1,2]. The needle and syringe-exchange programme in Amsterdam (hereafter called exchange programme) started in 1984 through an initiative of the drug users organization, the 'Junkiebond'. In 1986, needle-exchange schemes started in Sweden and Australia and in 1987 in England and Scotland [3,4]. While the initial aim of the Amsterdam programme was to prevent hepatitis B infection, this was soon overshadowed by the more important goal of AIDS prevention. The number of needles and syringes distributed increased greatly, from 100 000 in 1985 to 720 000 in 1988 [ 5 ]. At present it is t possible to exchange at 11 locations, consisting of methadone posts and other drug help organizations. At these locations, IDUs can exchange any kind or number of used needles and syringes for the same number of 2 cc needles and syringes. Condoms can be bought at a low price and leaflets on safe sex and safe use are also available. The main aim of the programme is to promote the once-only use of needles and syringes and to prevent needle sharing This is part of the general Amsterdam drug policy of harm reduction as the primary goal where the drug user is not willing or able to stop injecting drugs. The modus operandi of the exchange programme is that it removes used needles from circulation ('off the streets'), which is important from a public health point of view. In 1987, the percentage of returned needles was only 70%; instructions for exchange workers to keep more strictly to the one-to-one exchange rules led to an exchange percentage of 86%

in 1988 [5].

To keep the threshold as low as possible, counselling is optional and there is no registration at the exchange locations (except of the number of needles handed out per month per location). Therefore, there are no data available on number and characteristics of clients.

It is estimated that in 1987 there were about 7000 drug users in Amsterdam, of whom about 40% injected drugs.

There was, therefore, an estimated total of 2800 IDUs in Amsterdam [6]. Assuming that an average IDU uses 1.8 needles per day (84 recent IDUs from the Amsterdam cohort study among drug users (10) interviewed with a new questionnaire since February 1989 injected an average of 3.2 times a day and used the same needle on average 1.8 times; this means that they used 1.8 needles a day), in 1987 38% of all necessary needles were obtained at an exchange programme, while the other 62% were bought either in medical shops, a few pharmacies, small shops in the redlight (prostitution) district or on the street In 1987, a study was started to evaluate the exchange programme. In this report the relationship between participation in the exchange programme and risk level of injecting is analysed.

Study population and methods

The study population consisted of 145 Dutch and non Dutch IDUs in Amsterdam who volunteered to participate. The criterion for inclusion in the study was having injected during the previous 6 months. Subjects were recruited from exchange locations such as methadone maintenance programmes, drug advice agencies and the drug users' organization (Junkiebond) and from non-exchange locations such as consulting hour for foreign addicts, hospitals and police stations.

The mean age of 29.6 years and the male/female ratio of the subjects (2.4:1) are similar to those of the 4700 (approximately) drug users in Amsterdam (30.1 and 3.1: 1 years, respectively) who were seen by the Municipal Health Service in 1987 [6]. Since participants were self-selected, their representativeness is unknown.

From March to September 1987, participants were interviewed by nine different interviewers with a standardized questionnaire about their drug use, injecting behaviour and participation in the exchange programme in Amsterdam. All the interviewers, except one medical student, worked or had worked in the drug department and were experienced in working with drug users. None of the interviewers worked in the methadone-maintenance or treatmerit programmes in which their interviewees were participating.

Ten people were interviewed anonymously. When the other 135 participants in the original sample were approached for a follow-up in 1988, two had died and two refused a second interview. Twenty-three of the 83 foreign addicts were known to have returned to their country of origin. Of the remaining 108 people that we tried to contact, 60 participated in the follow-up interview. Although a follow-up of 41% seems low, one has to remember that tracing and motivating IDUs is very difficult.

Using the data from the first interview, we found a few significant differences between the 60 people who were reached for follow-up and the 85 who were not reached. The follow-up group consisted largely of Dutch IDUs (53 versus 35%) and therefore more of the follow-up group had had contact with a methadone agency in the previous 5 years than the group lost to follow-up (79 versus 49%). The follow-up group contained fewer people who had injected daily in the previous month (63 versus 88%), fewer people who had used heroin daily (57 versus 83%)

and more who had used methadone daily (72 versus 46%). However, there were no significant differences with regard to age, sex and duration of injecting. Risk levels with regard to injecting (see below) were also not significantly different and the same was true for the percentage of exchangers in both groups.

The mean interval between first and second interview was 13.5 months (range 10-20 months, median 14 months). Follow-up participants were interviewed with a follow-up version of the first questionnaire with a few added questions.

Demographic characteristics, drug use and injecting behaviour were analysed, with degree of participation in the exchange programme as an independent variable. Statistics used were chi-square, Student's t- and McNemar tests. P values less than 0.05 were taken as significant.

To study the risk level of injecting we combined the injecting variables 'to use your needle only once' and 'borrowing used needles from others in the previous month' [7]. We defined risk level 0 as not injecting in the previous month, and risk level 1 as the way of injecting with least risk, i.e. not borrowing needles and only using sterile needles. Risk level 2 was defined as 'not borrowing needles but re-using your own needle'. Although technically there is no difference with regard to AIDS risk between risk level 1 and 2, there are behavioural and psychological differences which are important with regard to AIDS prevention. Re-using your own needle is a high-risk activity because it can cause infections and because others may have used the needle in the meantime or needles may have been mixed up. From a psychological viewpoint it can be argued that it is easier not to share when you have already developed a habit of using only new (clean and sharp) needles than when you habitually use a needle more times. Risk level 3 was defined as borrowing used needles.

To study variables associated with needle sharing, 13 variables were selected for entry as independent variables in a logistic regression model with borrowing during the last two years as the independent variable. The 13 variables were: age, sex, nationality, duration of injecting, frequency of injecting heroin the month before, frequency of injecting cocaine the month before, frequency of injecting the month before, amount of injecting now compared with 6 months before, frequency of injecting alone the month before, frequency of exchanging needles within the last 6 months, using methadone daily, contact with methadone programme during the previous 5 years, having volunteered for an HIV test.



Results

General and drug-use characteristics at the first Interview

The study group consisted of 102 men and 43 women. Forty-three per cent had Dutch nationality, 30% German nationality and 27% another nationality. The mean age was 29.6 years (range 18-47 years). Males were, on average, 30.2 years old; women, on average, 28.4 years old.

Most (88%) of the interviewed IDUs reported that injecting had been their principal method of using drugs in the previous 2 years and 78% reported having injected daily during the previous month. The mean duration of injecting was 8 years and there was a significant correlation between age and duration of injecting: 0.48 (P < 0.001). The male IDUs had been injecting for a significantly longer time than the females: 9 versus 7 years (P < 0.05).

In the study group, 72% used heroin daily and 50% cocaine daily. Methadone was used daily by 57%, hashish and marijuana by 23%, tranquillizers by 18% and alcohol by 14%. Amphetamines were used daily by only 3% of the study group.

The German subgroup, although not older than the other nationalities, had been injecting for a longer time than the others: 10 versus 8 years (P < 0.05). Dutch IDUs were more often in contact with methadone programmes than foreign IDUs (90 versus 41%, P < 0.0001). This can be ascribed to the Amsterdam drug policy of preventing non-residents from entering methadone-maintenance programmes. Dutch IDUs also used methadone more often (66 versus 49%) at the time of the interview, but this difference was not significant.

Differences between exchangers and non-exchangers at the first Interview

Seventy-two people (50%) reported almost always (90% of the time or more) obtaining their new needles in the previ ous 6 months at the exchange programme. Forty-four people (30%) reported using the exchange programme never or rarely (10% of the time or less). The remaining 29 (20%) reported using the exchange programme, on average, 51% of the time. For the present analysis, we compared the group of 72 people who used the exchange programme more than 90% of the time with the other 73 people who used the exchange programme never or irregularly. We shall call the first group the 'exchangers' and the second group, the 'non-exchangers'.

There was no significant difference between these groups with regard to sex or nationality. The exchangers were older than the non-exchangers (31.2 versus 28.2 years, Student's t-test P=0.001) and had been injecting for alonger period of time (9.4 versus 7.4 years, Student's t-test P < 0.05). Exchangers had been in contact with methadone programmes in the last 5 years more often than non-exchangers (76 versus 48%, x2 = 10.4, d.f. = 1, P < 0.01). This also shows that 24% of the exchangers had not been in contact with methadone programmes in the last 5 years. There were no significant differences between groups with regard to the kind of drugs they used daily. Injecting behaviour in both groups also appeared to be similar: there were no significant differences between the percentages reporting predominant injecting during the previous 2 years, the percentages reporting injecting daily during the previous month, and the per centages reporting injecting only when being alone dur ing the previous month. Easy availability of free needles was not associated with an increase in injecting: 72% of the exchangers reported that they injected the same or less than 6 months previously compared with 49% of the non-exchangers (x2 = 7.0, d.f. = 1, P < 0.01). During the previous 6 months 3% of exchangers and 27% of non exchangers found themselves daily in the high-risk situa tion of possessing drugs but not clean needles (x2 = 15.2, d.f. = 1, P= 0.0001). Eighty-two per cent of the exchang ers used their own needles only once, compared with 29% of the non-exchangers (x2 = 39.3, d.f. = 1, P < 0.0001). More non-exchangers than exchangers had borrowed nee dles in the previous month as well as in the last 2 years; there was no such difference with regard to lending nee dles to other people (see Table 1).

Variables associated with needle sharing at the first interview

To study variables associated with needle sharing, 13 variables (see Methods) were selected for entry as indepen dent variables in a logistic regression model, with borrow ing during the last 2 years as the dependent variable. Five variables in the final model contributed significantly to the fit of the model (Table 2). As can be seen from the odds ratios, a level of 2, for example, for the frequency of inject ing in the previous month is associated with a 1.77 (77%) greater risk of borrowing during the previous 2 years than a level of 1. This means that (1) not injecting daily during the previous month, (2) not exchanging needles during the previous 6 months, (3) volunteering for an HIV test, (4) being male, and (5) being of a younger age are inde pendently associated with borrowing of needles during the previous 2 years.





Exchangers and non-exchangers at the second Interview

Thirty-five (49%) of the 72 exchangers and 25 (34%) of the 73 non-exchangers participated in the follow-up. The pro portion exchanging in the follow-up group did not change significantly (57% at interview 1 versus 63% at interview 2, n = 54).

Of the 35 exchangers, three had not injected during the previous 6 months at follow-up. Twenty-six were still exchanging more than 90% of the time (group EE) and 5 had started to exchange less than 90% of the time (group EN). The follow-up exchange data are missing for one former exchanger; this case is therefore excluded from analysis.

Of the 25 non-exchangers, two had not injected during the last 6 months at follow-up. Fifteen were still not or were only irregularly exchanging (group NN) and eight had started to exchange more than 90% of the time (group NE). General characteristics of the groups EE, EN, NN and NE can be compared in Table 3. As can be seen,'the group of stable exchangers (EE) is older and injects longer than the group of stable non-exchangers (NN).

Risk levels of injecting at first and second Interview

Risk levels were compared between the group of excharigers and non-exchangers. In 1987, at the first interview, the exchangers used at a significantly lower risk level than the non-exchangers (Table 4). Most of the exchangers (74%) used drugs at risk level 1, which means using only sterile needles (see Methods), versus 26% of the non-exchangers

Fewer exchangers (17%) used drugs at risk level 2 (meaning no borrowing but re-use of your own needle) than non-exchangers (48%). Ten per cent of the exchang ers used drugs at risk level 3 (which means borrowing) versus 23% of the non-exchangers.

When the group who used a needle once only (n = 80) is compared with the group which re-used needles (n = 65) we find that borrowing needles (risk level 3) occurs less frequently among the once-only users than among the re-users (9 versus 26%, P < 0.01).

In 1988, at the second interview, again a significant difference in risk level was found between exchangers and non-exchangers, as can be seen in Table 5.



At the second interview a question was added concerning the frequency of injecting over a longer period of time, i.e. 6 months. A relationship was found between exchanging and frequency of injecting during the last 6 months: 32% of the exchangers said they had injected irregularly during the last 6 months compared with 70% of the nonexchangers (X2 = 5.7, d.f = 1, P < 0.05).



Discussion

Exchangers appear to be older, to inject for a longer time and to be more frequently in contact with methadone programmes than non-exchangers. Participation in a needle exchange programme is not associated with an increase in injecting. Exchangers less often find themselves in a high risk situation (i.e. having drugs but no clean needle) than non-exchangers. Predominantly, they use their own needle only once, while most non-exchangers use it more than once. Exchangers also borrow used needles from other IDUs less often than non-exchangers: i.e. the injecting risk level of exchangers was much lower than of non exchangers in both 1987 and 1988. There seems to be no grounds for the fear that exchangers - because of their reliable supply of needles - will lend their used needles to others more often: there was no difference between ex changers and non-exchangers with regard to lending nee dles to others. Another positive fording is that the exchange programme seems to be able to reach IDUs who are not reached by methadone programmes.

Not exchanging needles in the last 6 months, being male and being younger were all associated with borrowing and these factors can be interpreted as predictive of borrowing. It is important to note that being male and being of a younger age are associated with borrowing independent of exchanging. This means that even among exchangers young IDUs and male IDUs have a higher rate of borrowing.

Volunteering for an HIV test was also associated with borrowing during the last 2 years. In this case the relationship is probably inverse: people who had been borrowing seemed to be more inclined to have an HIV test than people who had not been borrowing. This supports the current idea that drug users who volunteer for an HIV test have been more at risk than drug users in general.

The study was designed as a follow-up study. Followup rates would have been better had only Dutch subjects been selected for the first interview, since 57% of the initial interview sample were non-Dutch IDU s there was clearly less likelihood that they would be available for interview at follow-up.

The differences between the follow-up group and the group lost to follow-up can partly be explained by the larger proportion of Dutch users in the follow-up group. Dutch users have much easier access to methadone programmes than non-Dutch users, and, consequently, use more methadone than non-Dutch users. Daily injecting and daily heroin use was less common in the follow-up group than in that lost to follow-up. One possible reason could be that a considerable number of German and Italian high frequency users had returned to their country of origin.

There is an interesting relationship between infrequent injecting and not exchanging, which was found m1988 (comparable data were not available over 1987). This can be explained by the fact that those injecting irregularly might be in either a stopping phase or a re-starting phase, and therefore do not want to organize a supply of clean needles any longer or are not yet ready to organise themselves around their injecting behaviour. The financial ad vantage of exchanging over buying can also be expected to be more relevant for regular users.

We do not have a clear explanation for the finding that injecting infrequently in the previous month is associated with borrowing needles over the previous 2 years. It seems unwarranted to assume that injecting frequency over the previous month is an indicator for injection frequency over a longer period of time. The most plausible explanation is that serostatus is an intervening variable and that among those who have borrowed needles in the previous 2 years there are marry seropositives who have decreased their rate of injection because of their serostatus. In arty case, the relationship between irregular injecting, borrowing needles and participation in an exchange programme seems significant enough to be examined more thoroughly.

Until now a needle-exchange programme has been seen mainly from the 'supply side': offering a means of prevention. The 'demand side' is that an exchange programme is probably especially relevant to the more regular users or to users while they are in a period of regular use. In earlier articles [8,9] it was stressed that because of the great heterogeneity of the drug-using population it is very unlikely that only one method of HIV prevention will be universally effective. Since using irregularly seems to be associated with non-exchanging it becomes important to use (or keep using) other methods of HIV prevention in addition to an exchange programme. Another argument for this is that being young and being male is associated with borrowing needles independent of exchanging. This means that young or male exchangers are at greater risk of borrowing than old or female exchangers. Intensive counselling on risk reduction was advocated in a recent study (10) as a necessary ingredient for behaviour change. However, counselling has a higher threshold than simply exchanging, which means it may not be accepted by some users. A third prevention approach - giving information (by leaflets) about disinfecting a used needle with bleach - has been rejected until now in the Netherlands in order to keep the main prevention messages of 'no sharing' and 'always exchanging' clear. However, this approach does offer a compromise for those who find themselves in high risk situations. Ideally, of course, it would be better if every IDU used either his/her own needle, or did not inject at all, but this is not a realisdc universal objective. Knowledge about disinfecting with bleach might mean one less se roconversion each time a young male exchanger acquires drugs after office hours but has neither his own needles nor money. Since most of the IDUs we interviewed already practised cleaning a used needle with water when re-using, or borrowing to prevent the needle from getting clogged, it would appear to be a small step to add one more phase to this process (i.e. first rinsing twice with bleach).

We conclude that the exchange programme seems to be effective in lowering the risk level of injecting while not leading to an increase in injecting. It is possible to reach IDUs who are not reached by methadone programmes with an exchange programme. However, the exchange programme seems to attract mainly older, long-term injectors and regular injectors. Therefore, efforts have to be made to reach younger, short-term, and irregular injectors. Furthermore, the group that is not in contact with methadone programmes, predominantly non-Dutch IDUs, should be encouraged to exchange. Finally, to optimize risk reduction, the exchange programme should be one part of a set of varied preventive measures aimed at a heterogeneous IDU population.

Acknowledgements

We thank H.J.A. van Haastrecht for statistical advice, G. Hodge-Manos for correcting the English, Dr G.J.P. van Griensven for his critical com ments, J.S.A. Fennema for his efforts in tracing IDUs for follow-up, M. Grove and M. ter Pilli for preparing the manuscript and, finally, workers in the drug aid agencies and the participating IDUs for their cooperation.

References

1. Des Jarlais DC, Friedman SR, Hopkins W: Risk reduction for AIDS among i.v. drug users. Ann Intern Med 1985, 103:755-759.

2. Buning EC, Coutinho RA, van Brussel GHA. van Santen GW, van Zadelhofi AW. Preventing AIDS in drug addicts in Amsterdam [letter]. Lancet 1986, 1:1435.

3. Stimson GV: Syringe-exchange programmes for injecting drug users. AIDS 1989, 3'.253-260.

4. Stimson GV, Alldritt t.J, Dolan K, Donoghoe M: Syringe-exchange schemes for drug users in England and Scotland. Br Med J 1988, 296:1717-1719.

5. Drion J: Jaarverslag Sputtomruilprogramma 1988. Amsterdam: GG & GD 1989.

6. Buning EC: De GG & GD an her drugprobleem in cijfers, I, II & III. Amsterdam: GG & Go 1986, 1987, 1988.

7. Stallard A, Heather N, Johnston B: AIDS and intravenous drug use: what clinical psychology can offer. Bull Br Psychol Soc 1987, 40:365368.



8. Hangers C, Buning EC, van Santen GW, Verster AD, Coutinho RA: Intravenous druggebruik an her spuitenomruilprogramma in Amsterdam. T Soc Gezondheidsz 1988, 66:207-210.



9. Des Jarlais DC, Friedman SR: Target groups for preventing AIDS among intravenous drug users. J Appl Soc Psychol 1987, 17:251-268.



10. Van den Hoek JAR, van Haastrecht HJA, Coutinho RA: Risk re duction among intravenous drug users in Amsterdam under the influence of AIDS. Am J Publ Health (in press).