Coffee shops and clinics: the give and take of doing HIV/AIDS research with injecting drug users Alison Marsh and Wendy Loxley National Centre for Research into t b Prevention of Drug Abuse, Curtin University of Technology Abstract: We discuss recruiting and interviewing injecting drug users and using research as health promotion in the context of collecting information related to human immunodeficiency virus (HIV) or acquired immune deficiency syndrome (AIDS) from a convenience sample of 200 injecting drug users, half in treatment and half not, in 1989 and 1990 in Perth, Western Australia. A variety of recruiting methods were used including advertising, referral by agency staff, 'snowballing' and approaches to personal contacts and others known to inject by the interviewer. Snowballing and personal contacts were the most successful means of recruiting those not in treatment, while advertising was comparatively unsuccessful with this group because of the importance of establishing the credibility of the study and the interviewer among injecting drug users before they will volunteer to be involved. The promotion of behavioural risk reduction among respondents during the interview is detailed. We argue that the traditionally rigid separation between research and intervention is inappropriate in the HIV/AIDS context. When lives are potentially at stake, any contact with injecting drug users, especially those not in treatment (where many receive HIV/AIDS education), must be used as an HIV/AIDS prevention opportunity, and the interview is an ideal opportunity. The employment of research as community intervention is also discussed. (AustJ Public Health 1992; 16: 182-7)
T"
e experiences of many countries around the world attest to the fact that the human immunodeficiency virus (HIV) can be spread rapidly among injecting drug users by the sharing of injection equipment and unprotected sexual and from them into the general population via unprotected sexual ~ o n t a c t . ~ Evidence from around the world and from within Australia indicates that the response of many injecting drug users to the threat of HIV and the acquired immune deficiency syndrome (AIDS) has been a reduction, but not necessarily a cessation, of risky injecting and sexual b e h a v i o ~ r and , ~ ~that those in drug treatment tend to reduce their risky injecting practices to a greater extent than those not in treatment.g*10 Consequently, although HIV infection among Australian injecting drug users is consistently estimated to be relatively low, between 1 and 5 per cent,!' there is still a very real potential for an exponential increase in seroprevalence among this populati~n.~ Most of our knowledge about seroprevalence and HIV/AIDS risk behaviour among injecting drug users is based on those who identify as injecting drug users and present for HIV antibody testing, the majority of whom have been in contact with treatment services that provide HIV education." The number of injecting drug users not in treatment may be three or four times higher than the number in contact with official agencies, and we know very little about this group. It becomes apparent, therefore, that any comprehensiveapproach to HIV/AIDS prevention among injecting drug users must entail conCorrespondence to Ms Alison Marsh, National Centre for Research into the Prevention of Drug Abuse, Curtin University of Technology, GPO Box U1987, Perth, WA 6001. 182
certed efforts to establish seroprevalence, risk behaviours and HIV/AIDS prevention knowledge and awareness, not only among those 'known' injecting drug users (those identified because of contact with treatment and other official agencies), but also the more 'hidden' portion of the population we know little about. This task is beset with difficulties. The illegality and stigma associated with injecting drug use mean that injecting drug users are generally unwilling to identify as such unless they are in treatment, or with people they trust not to be judgmental or punishing. Furthermore, many injecting drug users may have nothing more in common with each other than having injected illicit drugs. The absence of a single identifiable injecting drug user subgroup and the tenuousness of their identification with each other makes knowing where to look for them difficult." Even when located, injecting drug users are often reluctant to be recruited into research studies for fear of the information falling into the wrong hands. Nevertheless, injecting drug users are successfully recruited into studies in numerous ways, many of which have been developed specifically to collect data from hidden population^.^*-'^ The factors outlined above also make it difficult to target injecting drug users, especially those not in treatment, with HIV/AIDS risk reduction initiatives."J5 However, the seriousness of the HIV/ AIDS threat to life makes it imperative that any means of reaching injecting drug users with harm reduction information must be used to advantage. Researchers who contact injecting drug users, especially those not in treatment, for study purposes, must therefore face the issue of whether to maintain the traditionally objective researcher role, or
AUSTRALIAN JOURNAL OF PUBLIC HEALTH 1992
VOL.16 NO.
2
RESEARCH WITH INJECTING DRUG USERS
whether to extend their role with the people they study to encompass HIV/AIDS health promotion. This paper will address the issues of recruiting and interviewing injecting drug users, and using research as health promotion, in the context of collecting HIV/AIDS-related information from a Perth sample of injecting drug users.
The Australian National AIDS and Injecting Drug Use Study (ANAIDUS) The ANAIDUS is a survey of current or recent injecting drug users and HIV/AIDS-related issues being conducted in a number of centres around Australia. The aims of the study are to assess HIV seropositivity among this sample by the testing of blood samples (blots) from finger pricks and to investigate the risk behaviour of injecting drug users in relation to HIV infection. To collect data from the hidden portion of the injecting drug user population in Australia, all collaborating centres agreed to aim to obtain half of their respondents from injecting drug users not currently in drug treatment and the other half from those currently in treatment. In the first year of the study (May 1989 to May 1990), face-to-face interviews were conducted with 200 injecting drug users in the Perth metropolitan area by one trained interviewer (AM) with extensive experience in the injecting drug use field. The data were collected by means of a standard, structured interview schedule. Recruiting and interviewing subjects Recruiting injecting drug users The most commonly reported methods of recruiting known injecting drug users involve capitalising on their use of drug- or HIV/AIDS-linked services such as treatment agencies and needle and syringe outlets. Methods include recruitment and interviewing of subjects by service providers such as needle exchange scheme ~ t a f f ,referral ~ . ~ of subjects to researchers by agency staff,5researchers advertising in agencies and being approached by interested subjectsgand direct recruitment of subjects in the agency setting by the researcher.l6 Methods of recruiting and interviewing from the hidden injecting drug user population include establishing a recruiting and research field station in the general geographical area in which the population of interest exi~ts,’~J’ frequenting places or ‘arenas’ where drug users congregate and (preferably) waiting for an introduction from a previous contact in the ‘scene’;16and ‘snowballing’, whereby previous contacts pass word of the study on to members of their using networks who then contact, or are introduced to, the researcher.’8J9 Perth recruitment procedures The means of recruitment we adopted in Perth were determined by: the need for confidentiality; the length, structure and complexity of the interview, which necessitated a trained interviewer and up to two hours of relative privacy; the nonavailability of
funds to rent office or shopfront space; the necessity to have user-friendly venues; the very covert nature of the Perth drug scene, whereby most using occurs in people’s homes rather than on the street; and the extensive number of contacts the interviewer had in the Perth drug scene before conducting the interviews. Methods and sources of recruitment varied over the course of the study according to response rate and constraints in terms of treatment and gender ratios. Four main types of recruitment procedures were used advertising in various publications and places, referral of individuals by staff at various agencies, direct approach to individuals by interviewer, and snowballing.
Results and discussionof recruitmentprocedures From Table 1 it is clear that more injecting drug users were recruited from the methadone clinic than from any other individual source, and that personal contacts and referral by a previous respondent were the next most successful means of recruitment. Smaller but significant numbers of injecting drug users were recruited from agencies other than the methadone clinic. The number of replies from the SS5 kit fliers was very low. (SS5 kits are plastic disposal packs containing five needles and syringes, a condom, safe sex and safer injecting information. They are sold by many pharmacies in Perth.) The high proportion of injecting drug users recruited from the methadone clinic was part of our sampling strategy, and clients were recruited by a combination of all our recruitment approaches (advertising, referral by staff, approach to clients by the interviewer and snowballing). Obtaining this number of respondents from the clinic took much continual contact and ANAIDUS ‘profile raising’ on the part of the interviewer with both staff and clients. Table 1: Sources of referral of respondents Source of referral
Number interviewed
The known injecting drug user population William Street Methadone Clinic Central Drug Unit Detoxification Centre
72
8
Palmerston Drug Rehabilitation Agency
7
Cyrenian House Drug Rehabilitation Agency
5 8
Perth Inner City Youth Services
PSST Van Needle Exchange
2 102
Total The hidden injecting drug user population Personal contact of interviewer
39
Referred by a previous respondent
36
Xpress magazine advertisement SERA newsletter advertisement (sex industry workers’ collective)
13
SS5 kit flier advertisement Response to newspaper article Total
98
Total from known and hidden recruitments combined
AUSTRALIAN JOURNAL OF PUBLIC HEALTH
4 4 2
1992
VOL.16 NO.2
200
183
MARSH AND LOXLEY
Power makes the point that in setting up research projects the value of tactful and diplomatic negotiations with cooperating agencies cannot be overestimated, and may make, break, or maim a project.16 It is as important to gain the respect and cooperation of agency staff as it is to gain that of the individuals to be interviewed. The interviewer was well aware of these issues and made a point of recruiting and interviewing respondents in a manner which maintained a cooperative relationship with agency staff as well as clients. This involved making sure that the way she worked at the clinic did not hinder staff in carrying out their normal duties. She sought feedback from staff about this issue, kept staff informed about the progress of the interviewing by formal presentations at staff meetings, and ensured that she became a familiar face around the clinic, available for informal chats with both staff members and clients. When response rates were poor, she would approach clients as they left the dispensing counter in an attempt to recruit them. This technique was quite successful when the keen respondents had already been interviewed and a more aggressive recruitment procedure became necessary. Many of the injecting drug users who were approached in this manner agreed to be interviewed and commented that they had meant to book an interview earlier but had not got around to it. The number of respondents recruited from forms of treatment other than the methadone program was limited by selection considerations rather than difficulty in persuading them to be interviewed. Recruitment methods were primarily referral by staff and snowballing,and injecting drug users were keen to be interviewed because word spread very quickly that it was safe, confidential and lucrative. The majority of injecting drug users in treatment are on sickness benefits and $20 can mean the difference between buying their own cigarettes and ‘bumming’ other people’s. To quote Goldstein et al., who also reported success with paying respondents: ‘The problem is not how to attract subjects but how to reject those who, for one reason or another, do not meet the criteria for participation’ (p. S7).I4Suggestions have been made that paying respondents encourages them to present with fictitious information just to get the money.*O This was not our experience. The complex nature of the interview and the familiarity of the researcher with the injecting drug user culture ensured that unsuitable respondents who tried to fabricate a story of injecting drug use would have been quickly identified and rejected from the study. We believe that paying injecting drug users for their time is important not just for the incentive value, but also as an indication that their time and information is important and valued. The success of personal contacts and snowballing is a common finding in research on injecting drug users, since these methods help to establish the crediThe interviewer knew a bility of the intervie~er.’~*’~ number of injecting drug users who were keen to be interviewed and recruited others to the study. Snowballing proved effective at gaining access to friendship groups, particularly those of younger users who 184
tended to live and use in larger groups and have very little money. One respondent who was well known in the ‘younguser scene’ recruited another eight injecting drug users directly to the study and another eight or so were recruited via these respondents. As the study progressed the interviewer became so well known in the Northbridge area of Perth that she was approached by injecting drug users seeking to be interviewed at various times and places. The relative lack of success of advertising compared with personal contacts and snowballing highlights the trust and confidentiality aspects of dealing with hidden populations. The Xpress advertisement was the only one (other than that at the methadone clinic) which netted many respondents. Whether this was due to its being run when knowledge of the study was already widespread, or due to other factors, such as accessing a wider injecting drug user population, is unclear. Replies from fliers placed in the SS5 needle packs sold by pharmacists were very slow, but well into the second year of the study occasional responses were still received. It seems that the kits containing the fliers are distributed very slowly by the pharmacists and that a number of injecting drug users who would have been contacted through the SS5 kits had already been recruited to the study by other methods. Other reasons possibly relate to the population we hoped to contact in this manner: the elusive ‘recreational’ users who may have careers, families and no drugrelated problems. These injecting drug users may have no need for $20, have other things to do with their time than be interviewed and be reluctant to identify as injecting drug users for fear of repercussions in other areas of their lives. Recruitment procedures needed to be continually varied. Usually the initial response to a recruitment initiative was the strongest, and would dwindle rapidly. People would also assume that the study had finished and that the advertisements were old ones unless they were changed from time to time or unless the interviewer maintained a ‘high profile’ and kept informing people that it was continuing. We also discovered that the more obvious we made the ‘earn $20’ on our advertisements and fliers the better the response. Interview venues There was no fured venue for conducting the interviews due to the difficulties of finding somewhere easily accessible, free, user-friendly and anonymous. The interview venue was negotiated between respondent and interviewer in each instance and included drug treatment agencies, a Perth youth service, coffee shops, respondents’ homes, and occasionally, for personal contacts, the interviewer’s home. Coffee shops were a very popular venue for a number of reasons. They are anonymous, and respondents could smoke cigarettes, eat and drink during the interview. The noise level aided privacy, as it helped the interviewer to ensure that neither the questions nor the responses were audible to other customers. When interviewing in respondents’
AUSTRALIAN JOURNAL OF PUBLIC HEALTH 1992
VOL. 16 NO.
2
RESEARCH WITH INJECTING DRUG USERS
homes, she was careful to establish beforehand th.e degree to which the interview might be disturbed, whether by children, police or visitors, and to stress that it needed to be undisturbed or it would have to be postponed to a different time and venue. Interviewing injecting drug users Interviewing current injecting drug users means at times interviewing people in various states of intoxication. To refuse to interview anyone who is to some extent intoxicated would result in few interviews and an extremely biased sample. For example, all respondents on the methadone program would be excluded because they are to some extent ‘intoxicated’ on methadone 24 hours a day. Respondents behaved quite differently in the interview dependingupon the drug they had used. Some methadone clients on high doses seemed to have diminished time perception., and were often very slow to respond and very expansive in their answers. Some respondents on benzodiazepines had difficulty comprehending the syntax of many questions and had to be led very carefully through them. Some of those on amphetamine spoke fast and completed the interview very quickly as long as the interviewer kept them focused in their answers. At times, ensuring that intoxicated respondents understood the questions and answered accurately required much skill, concentration and ingenuity on the part of the interviewer. Occasionally the extreme measure of proposing to postpone the interview (and hence the $20 payment) proved an effective means of increasing effort and concentration with intoxicated respondents, but in two instances the interview had to be postponed. The interviewer tried to encourage maximum concentration and honesty of responses by emphasising the importance of respondents’ contributions. She explained that the aim of collecting the information was to aid in the formation of policy and interventions to reduce the spread of HIV among injecting drug users and that only the respondents themselves really knew what went on and what needed to be done. They were thus actively engaged as ‘experts’ with a stake in the outcome. The interviewer monitored the reliability of responses by comparing answers to different questions as the interview progressed. Inconsistencies were brought to the attention of the respondent and this often revealed a misunderstood question early on, or caused respondents to look more closely at what they thought they did and what they actually did. Impressingupon respondents the importance of the accuracy of their contribution also improved reliability. Upon completion of the questionnaire respon dents were provided with feedback and health care information (see discussion below), were paid $20, asked if they were prepared to volunteer a contact name and phone number for future follow-up and encouraged to provide finger-prick blood for seropositivity testing.
The interview process as health promotion The belief that the process of studying certain behaviours has a tendency to alter these behaviours and thus compromise the research has traditionally been seen as a problem to be overcome. In the context of studying HIV/AIDS risk behaviours, which have the potential to place lives at stake, we contend the opposite, that this is a benefit to be made explicit and maximised. This applies especially to doing research among a marginal population such as injecting drug users, who may not be in contact with any official sources of information about HIV/AIDS and may not receive any encouragement to reduce their risk behaviour. The process of being asked a series of questions relating to HIV/AIDS risk behaviours and knowledge has the potential to influence respondents towards behaviour change. We chose to maximise this process by providing information and the opportunity for discussion. The process of being interviewed and the order in which the questions were asked encouraged respondents to closely evaluate the riskiness of their own behaviour. Questions about the specificsof their own injecting behaviour in terms of frequency and recency of sharing, number of sharing partners and so forth, and similar very specific questions about types and frequencies of sexual behaviours, were asked early in the interview. Questions about how risky they believed various injecting and sexual practices to be, whether they could tell if anyone was HIV positive, how long it takes for a test to detect HIV antibodies and how likely it is that someone HIV positive will die were asked later. Because they were made to look closely at their own HIV/AIDS risk behaviour, then estimate how dangerous these same practices are for HIV/AIDS, and finally state the health and life dangers of HIV/AIDS, respondents were often placed in a state of cognitive dissonance. Frequently the behaviours they estimated as very dangerous were the very behaviours they described engaging in themselves. Thus the interview process could be seen as personalising the risk of contracting HIV/AIDS and heightening respondents’ awareness of any lack of correspondence between attitudes, self-perception, and behaviours; many respondents expressed discomfort at this. Increasing the individual’s personal, salient awareness of the riskiness of his or her behaviour is one of the essential features of motivational interviewing, one of the techniques outlined by Miller as designed to increase motivation to change behaviour.*l Thus the interview process itself can be seen as a potential motivator of behaviour change. At the conclusion of the interview we provided respondents first with factual information about HIV/AIDS and then with risk reduction information. Respondents were given feedback about the accuracy of their assessments of the relative dangers of certain behaviours and of their knowledge about the progress of the disease and the meaning of HIV test results. Many respondents expressed a keen interest in this feedback and in discussing and comparing it with their previous ideas and knowledge about HIV/AIDS.
AUSTRALIAN JOURNAL OF PUBLIC HEALTH 1992 vot. 16 NO. 2
185
MARSH AND LOXLEY
Wiebel, in a paper outlining the adaptation of what has become known as the ‘Chicago community outreach model’ to reducing the spreads of HIV/AIDS among injecting drug users, also discusses the issue of risk personalisation. He makes the point that the relevance of HIV/AIDS to the individual must be established before information about risk reduction is assimilated, and before risk reduction strategies are considered. Wiebel further contends that involving respondents as active participants in the process of risk assessment, and in identifyingspecific behaviour patterns which need to be changed, elevates them from the status of passive recipients of edicts about what they should and should not do to that of active participants in an evaluation process which they can then apply independently.21Thus respondents had the opportunity to identify particular areas of concern about which to further question the interviewer. In terms of Wiebel’s thesis, then, our research interviews personalised the risk of HIV/AIDS for respondents, thus preparing the ground for them to consider risk reduction information and strategies, and helped them to establish risk assessment techniques that they could then use independently. The next crucial component Wiebel identifies as encouraging HIV/AIDS risk behaviour change is the provision of a range of risk reduction alternatives so that the possibility that one will be found to be viable by the individual is maximised. We gave respondents information about where and how to have an HIV test, safe injecting, disposal and cleaning practices, safe sex practices, and the location of needle exchanges. If they were interested they were also told about treatment options and how to contact the users’ collective (WAIVE)or the sex workers’ collective (SIERA). Some respondents had little knowledge of needle exchanges and their locations, or of userfriendly chemists, or of safe cleaning methods, and questioned the interviewer closely about these issues. Providing this information removed some of the barriers to behaviour change for respondents, another technique considered by Miller to increase motivation to change behaviour.22 A number of respondents commented at the conclusion of the interview that not only did they feel that the study was necessary and worthwhile, but that they had found their participation in it a useful and worthwhile experience. Wiebel’s final point is that HIV/AIDS risk behaviour changes must be reinforced if they are to last.21 The interview itself was obviously a single intervention in respondents’ lives, and as such did not offer reinforcement. We hope that the provision of information about services for injecting drug users may have given some respondents the opportunity to engage in ongoing personal contact with sources of reinforcement for behaviour change. Respondents revealed a wealth of very personal and sensitive detail to the interviewer. For the interviewer to have then refused to provide them with information they requested about HIV/AIDSrelated issues could have been seen as disregarding the mutuality of the relationship that had been built up over the course of the interview. Respondents 186
could have been justifiably angered by this, which would have damaged the reputation of the study and the position the interviewer had forged among the injecting drug user community. Does providing information threaten the veracity of the research process? Although the information was provided after the data had been gathered, and thus would not have contaminated the data in this sense, it is conceivable that respondents may have passed on information to their peers who were subsequently interviewed, thus biasing the data. However, HIV/AIDS poses such a certain threat to life that we consider it unethical not to provide subjects with information which helps them to avoid contracting it. Power discusses the controversy which surrounds this issue.I6 He describes Feldman’s view that in situations when contributions to science are pitted against benefits to the research subjects, the latter should be preferred.23 Power, unlike Feldman, argues that unbiased research should be the primary focus in most instances. However, he holds that this general position may be called into question where observed activities may endanger life. He considers that intervention should occur only when there is serious threat to the individuals concerned, and he argues that HIV/AIDS presents such a threat so that ‘in the current climate research objectivity may at times have to take a back seat to health education and risk reduction’ (p. 50). He argues for a serious consideration of an expanded role of research to assist in risk reduction. It must be acknowledged that there is a further unresolved tension here, in that objective, unbiased research is most likely to result in accurate HIV/AIDS information, and without this, research subjects. may ultimately suffer by receiving risk reduction information which may be unknowingly biased. This distinction between research and intervention is also pursued by Merton in the context of community-based HIV/AIDS drug trials.24 More generally, Merton cites criticism of the randomised controlled trial on scientific grounds as enforcing rigid rules of statistical and other interpretation which ‘by-passjudgment’. In other words, decisions as to what constitutesa sufficiently rigorous study are not objective and purely scientific, but depend upon the values of the decision maker, so that even among clinical researchers there are ‘pragmatists’ who think trials should reflect the realities of practice so that results can be generalised immediately, and the ‘fastidious’ who believe that trials should be designed to e l i n a t e confounding factors and any possibility of bias [p. 5081.
Merton agrees with Sonnabend’s argument that in the context of HIV/AIDS ‘the traditionally emphatic separation between the goals of research and treatment just won’t wash’ (p. 503), and considers that community-based HIV/AIDS research ‘must balance the values of scientific credibility with the moral principles of respect for persons, including autonomy, beneficence and justice’ (p. 502).25 It is impossible, in a city like Perth, where there are ongoing HIV/AIDS prevention initiatives, to control confounding factors and eliminate bias in research
AUSTRALIAN JOURNAL OF PUBLIC HEALTH
1992
VOL. I6 NO.
2
RESEARCH WITH INJECTING DRUG USERS
among injecting drug users. HIV/AIDS risk behaviour change among injecting drug users is almost certain to be the result of a multitude of interventions including media campaigns, the provision of accessible sterile injecting equipment, peer group influences and possibly, for a few, the information we provided. Research that maps changes in behaviour, attitudes and knowledge over time, in relation to the occurrence of various HIV/AIDS interventions, and qualitative research that explores the factors to which injecting drug users attribute their behaviour changes, can contribute insights into the influence of interventions on risk behaviour.
The research process as community intervention The urgent nature of preventing the spread of HIV/ AIDS among injecting drug users has required that we communicate relevant findings and implications for HIV prevention in diverse ways as quickly and as widely as possible before publishing them in refereed journals. The ways in which we have communicated study findings rely strongly on maintaining close and regular liaison with those interested in and working in the HIV/AIDS and injecting drug user area. Many of these individuals and organisations have aided us in recruiting subjects and in providing them with up-to-date information. We hope our research has been reciprocally helpful in guiding them in the formulation of HIV/AIDS prevention initiatives. Examples of our community involvement are as follows: membership of a state interdisciplinary committee on HIV/AIDS and injecting drug users; membership of the local users’ collective, assistance with pharmacy workshops about the provision o:f needles to users; participation in drug use working parties; conference presentations; media input; the distribution of ANAIDUS findings to a number of places including the Health Department of Western Australia, the Western Australian AIDS Council, the Western Australian Alcohol and Drug Authority and other interested parties; and consultation about HIV/AIDS and injecting drug users with various people referred to us. We believe this involvement to have been mutually beneficial in that they received information from us as soon as it was available, and they were extremely helpful to us in both publicising the study and encouraging injecting drug users to participate. Acknowledgments Without the help and support of a number of agencies and individuals this research would have been very difficult, if not impossible to complete. In particular we would like to thank all those who attempted to help us in recruiting respondents to the: study, our co-investigators in Perth, Professor David Hawks (Director of the National Centre for Research into the Prevention of Drug Abuse) and Dr Allan Quigley (Director of Clinical Services for the West.. ern Australian Alcohol and Drug Authority), our colleagues involved in the ANAIDUS in other centres
around Australia and Bayer Glucolet for the provision of the glucolets and lancets with which we obtained blood samples. Most importantly of all, we would like to thank those injecting drug users who gave of their time, energy and trust to be interviewed. This research was funded by the Commonwealth AIDS Research Grants Committee. References Des Jarlais DC, Friedman SR, Novick DM, Sotheran JL et al. HIV-1 infection among intravenousdrug users in Manhattan, New York City, from 1977 through 1987.JAMA 1989;261: 1008-12. 2. Stimson GV, Donoghoe M,AUdritt L, Dolan K. HIV transmission risk behaviour of clients attending syrhge-exchange schemes in England and Scotland. BrJ Addict 1988;83: 1449-55. 3. Wodak A. Wffl the sky fall in? The prevention of HIV infection in intravenous drug w r s in Australia. Paper presented at the Fourth National Conference on AIDS, Canberra, 9-1 1 July 1990. Sydney: Alcohol and Drug %vice, St Vincent’s Hospital. 4. Donoghoe MC. Stimson GV, Dolan KA. Sexual behaviour of injecting drug users and associated risk of HIV infection for non-injecting sexual partners. AIDS Cure 1989;l(1):51-8. 5. McKeganey N, Barnard M, Watson H. HIV related risk behaviour among a non-clinic sample of injecting drug users. B r j Addid 1989;
1.
84;1481-90. 6. D e s Jarlais DC, Friedman SR. Target groups for preventing AIDS among intravenous drug users. J Appl Soc Psgchol 1987; 17(3): 251-68. 7. Robertson J, Skidmore C, Roberts J. HIV infection in intravenous drug users: a follow-up study indicating changes in risk-taking behaviour. BrJ Addid 1988;83: 387-91. 8. van den Hoek A, van Haastrecht HJ, Coutinho RA. Heterosexual behaviour of intravenous drug users in Amsterdam: implications for the AIDS epidemic. AIDS 1990;4:449-53. 9. Darke S, Hall W. Carless J. Drug use, injecting practices and sexual behaviour of opioid users in Sydney, Australia. BrJ Addict 1990;85:
1603-9. 10. Power K,Hartnoll R, Daiaud E.Drug injecting,AIDS, and risk behaviour: potential for change and intervention strategies. Br J Addict 1988;83: 649-54. 11. Watson J. Rcpoll of the selcd committee appointed to inquire into the National HW/AIDS Strategy White Paper. Perth: Australian Government Publishing Service, 1990. 12. Wiebel WW. Identifying and gainiing access to hidden populations. NIDA Rcs Mmgr 1990;98:4-11. 13. Adler P. Ethnopphic research on hidden populations: penetrating the drug world. NIDA Res Mmgr 1990;98:96112. 14. Goldstein PJ, Spunt BJ. Miller T, Bellucci P. Ethnographic field stations. NIDA Res Mmgr 1990; 98:80-95. 15. Dickie M.Can education influence the behaviour of injecting drug users? Paper presented at 3rd National Conference on AIDS: Living With AIDS in the Year 2000.Hobart, August, 1988.Sydney: Centre for Education and Information on Drugs and Alcohol. 16. Power R. Participant observation and its place in the study of illicit drug abuse. BrJ Addict 1989;84:43-52. 17. Dobinson I, Poletti P. Buying and selling heroin: a study of heroin user/dealers. Sydney: NSW Bureau of Crime Statistics and Research, 1988. 18. Biemacki P, Waldorf D. Snowball sampling. Biol Mcllbd Res 1981; lO(2): 141-63. 19. Morrison V, Plant M. Drug problems and pattems of service use among illicit drug users in Edinburgh. Br J Addid 1990; 85: 547-54. 20. Reilly C, Flaherty B, Home1 P. A study of 1621 year old illicit drug w r s in Sydney. Conference paper presented at W m t a School in the Sun. Brisbane, 3-6 July, 1990. Sydney: Dimtorate of the Drug Offensive, NSW Health Department. 21. Wiebel WW. Combining ethnographic and epidemiologic methods in targeted AIDS interventions: the Chicago model. NIDA Rrs M m g r 80; 1988:137-50. 22. Miller WR. Increasing motivation for change. In: Hester RK, Miller Wa, eds. Handbook of alcoholism treatment appmchcr. New Yorlr: P e r g a m ~1989. . 23. Feldman HW. Street status and the drug researcher: issues in participant-observation. Washington, D C Drug Abuse Council, 1974. 24. Merton V. Community-based AIDS research. Eual RN 1990;14(5): 502-97. 25. Sonnabend J. Do we need new ways to evaluate experimental AIDS treatment? AIDS Farm 1988; l(November): 4-8.
AUSTRALIAN JOURNAL OF PUBLIC HEALTH 1992 vot. 16 NO. 2
187
T"
e experiences of many countries around the world attest to the fact that the human immunodeficiency virus (HIV) can be spread rapidly among injecting drug users by the sharing of injection equipment and unprotected sexual and from them into the general population via unprotected sexual ~ o n t a c t . ~ Evidence from around the world and from within Australia indicates that the response of many injecting drug users to the threat of HIV and the acquired immune deficiency syndrome (AIDS) has been a reduction, but not necessarily a cessation, of risky injecting and sexual b e h a v i o ~ r and , ~ ~that those in drug treatment tend to reduce their risky injecting practices to a greater extent than those not in treatment.g*10 Consequently, although HIV infection among Australian injecting drug users is consistently estimated to be relatively low, between 1 and 5 per cent,!' there is still a very real potential for an exponential increase in seroprevalence among this populati~n.~ Most of our knowledge about seroprevalence and HIV/AIDS risk behaviour among injecting drug users is based on those who identify as injecting drug users and present for HIV antibody testing, the majority of whom have been in contact with treatment services that provide HIV education." The number of injecting drug users not in treatment may be three or four times higher than the number in contact with official agencies, and we know very little about this group. It becomes apparent, therefore, that any comprehensiveapproach to HIV/AIDS prevention among injecting drug users must entail conCorrespondence to Ms Alison Marsh, National Centre for Research into the Prevention of Drug Abuse, Curtin University of Technology, GPO Box U1987, Perth, WA 6001. 182
certed efforts to establish seroprevalence, risk behaviours and HIV/AIDS prevention knowledge and awareness, not only among those 'known' injecting drug users (those identified because of contact with treatment and other official agencies), but also the more 'hidden' portion of the population we know little about. This task is beset with difficulties. The illegality and stigma associated with injecting drug use mean that injecting drug users are generally unwilling to identify as such unless they are in treatment, or with people they trust not to be judgmental or punishing. Furthermore, many injecting drug users may have nothing more in common with each other than having injected illicit drugs. The absence of a single identifiable injecting drug user subgroup and the tenuousness of their identification with each other makes knowing where to look for them difficult." Even when located, injecting drug users are often reluctant to be recruited into research studies for fear of the information falling into the wrong hands. Nevertheless, injecting drug users are successfully recruited into studies in numerous ways, many of which have been developed specifically to collect data from hidden population^.^*-'^ The factors outlined above also make it difficult to target injecting drug users, especially those not in treatment, with HIV/AIDS risk reduction initiatives."J5 However, the seriousness of the HIV/ AIDS threat to life makes it imperative that any means of reaching injecting drug users with harm reduction information must be used to advantage. Researchers who contact injecting drug users, especially those not in treatment, for study purposes, must therefore face the issue of whether to maintain the traditionally objective researcher role, or
AUSTRALIAN JOURNAL OF PUBLIC HEALTH 1992
VOL.16 NO.
2
RESEARCH WITH INJECTING DRUG USERS
whether to extend their role with the people they study to encompass HIV/AIDS health promotion. This paper will address the issues of recruiting and interviewing injecting drug users, and using research as health promotion, in the context of collecting HIV/AIDS-related information from a Perth sample of injecting drug users.
The Australian National AIDS and Injecting Drug Use Study (ANAIDUS) The ANAIDUS is a survey of current or recent injecting drug users and HIV/AIDS-related issues being conducted in a number of centres around Australia. The aims of the study are to assess HIV seropositivity among this sample by the testing of blood samples (blots) from finger pricks and to investigate the risk behaviour of injecting drug users in relation to HIV infection. To collect data from the hidden portion of the injecting drug user population in Australia, all collaborating centres agreed to aim to obtain half of their respondents from injecting drug users not currently in drug treatment and the other half from those currently in treatment. In the first year of the study (May 1989 to May 1990), face-to-face interviews were conducted with 200 injecting drug users in the Perth metropolitan area by one trained interviewer (AM) with extensive experience in the injecting drug use field. The data were collected by means of a standard, structured interview schedule. Recruiting and interviewing subjects Recruiting injecting drug users The most commonly reported methods of recruiting known injecting drug users involve capitalising on their use of drug- or HIV/AIDS-linked services such as treatment agencies and needle and syringe outlets. Methods include recruitment and interviewing of subjects by service providers such as needle exchange scheme ~ t a f f ,referral ~ . ~ of subjects to researchers by agency staff,5researchers advertising in agencies and being approached by interested subjectsgand direct recruitment of subjects in the agency setting by the researcher.l6 Methods of recruiting and interviewing from the hidden injecting drug user population include establishing a recruiting and research field station in the general geographical area in which the population of interest exi~ts,’~J’ frequenting places or ‘arenas’ where drug users congregate and (preferably) waiting for an introduction from a previous contact in the ‘scene’;16and ‘snowballing’, whereby previous contacts pass word of the study on to members of their using networks who then contact, or are introduced to, the researcher.’8J9 Perth recruitment procedures The means of recruitment we adopted in Perth were determined by: the need for confidentiality; the length, structure and complexity of the interview, which necessitated a trained interviewer and up to two hours of relative privacy; the nonavailability of
funds to rent office or shopfront space; the necessity to have user-friendly venues; the very covert nature of the Perth drug scene, whereby most using occurs in people’s homes rather than on the street; and the extensive number of contacts the interviewer had in the Perth drug scene before conducting the interviews. Methods and sources of recruitment varied over the course of the study according to response rate and constraints in terms of treatment and gender ratios. Four main types of recruitment procedures were used advertising in various publications and places, referral of individuals by staff at various agencies, direct approach to individuals by interviewer, and snowballing.
Results and discussionof recruitmentprocedures From Table 1 it is clear that more injecting drug users were recruited from the methadone clinic than from any other individual source, and that personal contacts and referral by a previous respondent were the next most successful means of recruitment. Smaller but significant numbers of injecting drug users were recruited from agencies other than the methadone clinic. The number of replies from the SS5 kit fliers was very low. (SS5 kits are plastic disposal packs containing five needles and syringes, a condom, safe sex and safer injecting information. They are sold by many pharmacies in Perth.) The high proportion of injecting drug users recruited from the methadone clinic was part of our sampling strategy, and clients were recruited by a combination of all our recruitment approaches (advertising, referral by staff, approach to clients by the interviewer and snowballing). Obtaining this number of respondents from the clinic took much continual contact and ANAIDUS ‘profile raising’ on the part of the interviewer with both staff and clients. Table 1: Sources of referral of respondents Source of referral
Number interviewed
The known injecting drug user population William Street Methadone Clinic Central Drug Unit Detoxification Centre
72
8
Palmerston Drug Rehabilitation Agency
7
Cyrenian House Drug Rehabilitation Agency
5 8
Perth Inner City Youth Services
PSST Van Needle Exchange
2 102
Total The hidden injecting drug user population Personal contact of interviewer
39
Referred by a previous respondent
36
Xpress magazine advertisement SERA newsletter advertisement (sex industry workers’ collective)
13
SS5 kit flier advertisement Response to newspaper article Total
98
Total from known and hidden recruitments combined
AUSTRALIAN JOURNAL OF PUBLIC HEALTH
4 4 2
1992
VOL.16 NO.2
200
183
MARSH AND LOXLEY
Power makes the point that in setting up research projects the value of tactful and diplomatic negotiations with cooperating agencies cannot be overestimated, and may make, break, or maim a project.16 It is as important to gain the respect and cooperation of agency staff as it is to gain that of the individuals to be interviewed. The interviewer was well aware of these issues and made a point of recruiting and interviewing respondents in a manner which maintained a cooperative relationship with agency staff as well as clients. This involved making sure that the way she worked at the clinic did not hinder staff in carrying out their normal duties. She sought feedback from staff about this issue, kept staff informed about the progress of the interviewing by formal presentations at staff meetings, and ensured that she became a familiar face around the clinic, available for informal chats with both staff members and clients. When response rates were poor, she would approach clients as they left the dispensing counter in an attempt to recruit them. This technique was quite successful when the keen respondents had already been interviewed and a more aggressive recruitment procedure became necessary. Many of the injecting drug users who were approached in this manner agreed to be interviewed and commented that they had meant to book an interview earlier but had not got around to it. The number of respondents recruited from forms of treatment other than the methadone program was limited by selection considerations rather than difficulty in persuading them to be interviewed. Recruitment methods were primarily referral by staff and snowballing,and injecting drug users were keen to be interviewed because word spread very quickly that it was safe, confidential and lucrative. The majority of injecting drug users in treatment are on sickness benefits and $20 can mean the difference between buying their own cigarettes and ‘bumming’ other people’s. To quote Goldstein et al., who also reported success with paying respondents: ‘The problem is not how to attract subjects but how to reject those who, for one reason or another, do not meet the criteria for participation’ (p. S7).I4Suggestions have been made that paying respondents encourages them to present with fictitious information just to get the money.*O This was not our experience. The complex nature of the interview and the familiarity of the researcher with the injecting drug user culture ensured that unsuitable respondents who tried to fabricate a story of injecting drug use would have been quickly identified and rejected from the study. We believe that paying injecting drug users for their time is important not just for the incentive value, but also as an indication that their time and information is important and valued. The success of personal contacts and snowballing is a common finding in research on injecting drug users, since these methods help to establish the crediThe interviewer knew a bility of the intervie~er.’~*’~ number of injecting drug users who were keen to be interviewed and recruited others to the study. Snowballing proved effective at gaining access to friendship groups, particularly those of younger users who 184
tended to live and use in larger groups and have very little money. One respondent who was well known in the ‘younguser scene’ recruited another eight injecting drug users directly to the study and another eight or so were recruited via these respondents. As the study progressed the interviewer became so well known in the Northbridge area of Perth that she was approached by injecting drug users seeking to be interviewed at various times and places. The relative lack of success of advertising compared with personal contacts and snowballing highlights the trust and confidentiality aspects of dealing with hidden populations. The Xpress advertisement was the only one (other than that at the methadone clinic) which netted many respondents. Whether this was due to its being run when knowledge of the study was already widespread, or due to other factors, such as accessing a wider injecting drug user population, is unclear. Replies from fliers placed in the SS5 needle packs sold by pharmacists were very slow, but well into the second year of the study occasional responses were still received. It seems that the kits containing the fliers are distributed very slowly by the pharmacists and that a number of injecting drug users who would have been contacted through the SS5 kits had already been recruited to the study by other methods. Other reasons possibly relate to the population we hoped to contact in this manner: the elusive ‘recreational’ users who may have careers, families and no drugrelated problems. These injecting drug users may have no need for $20, have other things to do with their time than be interviewed and be reluctant to identify as injecting drug users for fear of repercussions in other areas of their lives. Recruitment procedures needed to be continually varied. Usually the initial response to a recruitment initiative was the strongest, and would dwindle rapidly. People would also assume that the study had finished and that the advertisements were old ones unless they were changed from time to time or unless the interviewer maintained a ‘high profile’ and kept informing people that it was continuing. We also discovered that the more obvious we made the ‘earn $20’ on our advertisements and fliers the better the response. Interview venues There was no fured venue for conducting the interviews due to the difficulties of finding somewhere easily accessible, free, user-friendly and anonymous. The interview venue was negotiated between respondent and interviewer in each instance and included drug treatment agencies, a Perth youth service, coffee shops, respondents’ homes, and occasionally, for personal contacts, the interviewer’s home. Coffee shops were a very popular venue for a number of reasons. They are anonymous, and respondents could smoke cigarettes, eat and drink during the interview. The noise level aided privacy, as it helped the interviewer to ensure that neither the questions nor the responses were audible to other customers. When interviewing in respondents’
AUSTRALIAN JOURNAL OF PUBLIC HEALTH 1992
VOL. 16 NO.
2
RESEARCH WITH INJECTING DRUG USERS
homes, she was careful to establish beforehand th.e degree to which the interview might be disturbed, whether by children, police or visitors, and to stress that it needed to be undisturbed or it would have to be postponed to a different time and venue. Interviewing injecting drug users Interviewing current injecting drug users means at times interviewing people in various states of intoxication. To refuse to interview anyone who is to some extent intoxicated would result in few interviews and an extremely biased sample. For example, all respondents on the methadone program would be excluded because they are to some extent ‘intoxicated’ on methadone 24 hours a day. Respondents behaved quite differently in the interview dependingupon the drug they had used. Some methadone clients on high doses seemed to have diminished time perception., and were often very slow to respond and very expansive in their answers. Some respondents on benzodiazepines had difficulty comprehending the syntax of many questions and had to be led very carefully through them. Some of those on amphetamine spoke fast and completed the interview very quickly as long as the interviewer kept them focused in their answers. At times, ensuring that intoxicated respondents understood the questions and answered accurately required much skill, concentration and ingenuity on the part of the interviewer. Occasionally the extreme measure of proposing to postpone the interview (and hence the $20 payment) proved an effective means of increasing effort and concentration with intoxicated respondents, but in two instances the interview had to be postponed. The interviewer tried to encourage maximum concentration and honesty of responses by emphasising the importance of respondents’ contributions. She explained that the aim of collecting the information was to aid in the formation of policy and interventions to reduce the spread of HIV among injecting drug users and that only the respondents themselves really knew what went on and what needed to be done. They were thus actively engaged as ‘experts’ with a stake in the outcome. The interviewer monitored the reliability of responses by comparing answers to different questions as the interview progressed. Inconsistencies were brought to the attention of the respondent and this often revealed a misunderstood question early on, or caused respondents to look more closely at what they thought they did and what they actually did. Impressingupon respondents the importance of the accuracy of their contribution also improved reliability. Upon completion of the questionnaire respon dents were provided with feedback and health care information (see discussion below), were paid $20, asked if they were prepared to volunteer a contact name and phone number for future follow-up and encouraged to provide finger-prick blood for seropositivity testing.
The interview process as health promotion The belief that the process of studying certain behaviours has a tendency to alter these behaviours and thus compromise the research has traditionally been seen as a problem to be overcome. In the context of studying HIV/AIDS risk behaviours, which have the potential to place lives at stake, we contend the opposite, that this is a benefit to be made explicit and maximised. This applies especially to doing research among a marginal population such as injecting drug users, who may not be in contact with any official sources of information about HIV/AIDS and may not receive any encouragement to reduce their risk behaviour. The process of being asked a series of questions relating to HIV/AIDS risk behaviours and knowledge has the potential to influence respondents towards behaviour change. We chose to maximise this process by providing information and the opportunity for discussion. The process of being interviewed and the order in which the questions were asked encouraged respondents to closely evaluate the riskiness of their own behaviour. Questions about the specificsof their own injecting behaviour in terms of frequency and recency of sharing, number of sharing partners and so forth, and similar very specific questions about types and frequencies of sexual behaviours, were asked early in the interview. Questions about how risky they believed various injecting and sexual practices to be, whether they could tell if anyone was HIV positive, how long it takes for a test to detect HIV antibodies and how likely it is that someone HIV positive will die were asked later. Because they were made to look closely at their own HIV/AIDS risk behaviour, then estimate how dangerous these same practices are for HIV/AIDS, and finally state the health and life dangers of HIV/AIDS, respondents were often placed in a state of cognitive dissonance. Frequently the behaviours they estimated as very dangerous were the very behaviours they described engaging in themselves. Thus the interview process could be seen as personalising the risk of contracting HIV/AIDS and heightening respondents’ awareness of any lack of correspondence between attitudes, self-perception, and behaviours; many respondents expressed discomfort at this. Increasing the individual’s personal, salient awareness of the riskiness of his or her behaviour is one of the essential features of motivational interviewing, one of the techniques outlined by Miller as designed to increase motivation to change behaviour.*l Thus the interview process itself can be seen as a potential motivator of behaviour change. At the conclusion of the interview we provided respondents first with factual information about HIV/AIDS and then with risk reduction information. Respondents were given feedback about the accuracy of their assessments of the relative dangers of certain behaviours and of their knowledge about the progress of the disease and the meaning of HIV test results. Many respondents expressed a keen interest in this feedback and in discussing and comparing it with their previous ideas and knowledge about HIV/AIDS.
AUSTRALIAN JOURNAL OF PUBLIC HEALTH 1992 vot. 16 NO. 2
185
MARSH AND LOXLEY
Wiebel, in a paper outlining the adaptation of what has become known as the ‘Chicago community outreach model’ to reducing the spreads of HIV/AIDS among injecting drug users, also discusses the issue of risk personalisation. He makes the point that the relevance of HIV/AIDS to the individual must be established before information about risk reduction is assimilated, and before risk reduction strategies are considered. Wiebel further contends that involving respondents as active participants in the process of risk assessment, and in identifyingspecific behaviour patterns which need to be changed, elevates them from the status of passive recipients of edicts about what they should and should not do to that of active participants in an evaluation process which they can then apply independently.21Thus respondents had the opportunity to identify particular areas of concern about which to further question the interviewer. In terms of Wiebel’s thesis, then, our research interviews personalised the risk of HIV/AIDS for respondents, thus preparing the ground for them to consider risk reduction information and strategies, and helped them to establish risk assessment techniques that they could then use independently. The next crucial component Wiebel identifies as encouraging HIV/AIDS risk behaviour change is the provision of a range of risk reduction alternatives so that the possibility that one will be found to be viable by the individual is maximised. We gave respondents information about where and how to have an HIV test, safe injecting, disposal and cleaning practices, safe sex practices, and the location of needle exchanges. If they were interested they were also told about treatment options and how to contact the users’ collective (WAIVE)or the sex workers’ collective (SIERA). Some respondents had little knowledge of needle exchanges and their locations, or of userfriendly chemists, or of safe cleaning methods, and questioned the interviewer closely about these issues. Providing this information removed some of the barriers to behaviour change for respondents, another technique considered by Miller to increase motivation to change behaviour.22 A number of respondents commented at the conclusion of the interview that not only did they feel that the study was necessary and worthwhile, but that they had found their participation in it a useful and worthwhile experience. Wiebel’s final point is that HIV/AIDS risk behaviour changes must be reinforced if they are to last.21 The interview itself was obviously a single intervention in respondents’ lives, and as such did not offer reinforcement. We hope that the provision of information about services for injecting drug users may have given some respondents the opportunity to engage in ongoing personal contact with sources of reinforcement for behaviour change. Respondents revealed a wealth of very personal and sensitive detail to the interviewer. For the interviewer to have then refused to provide them with information they requested about HIV/AIDSrelated issues could have been seen as disregarding the mutuality of the relationship that had been built up over the course of the interview. Respondents 186
could have been justifiably angered by this, which would have damaged the reputation of the study and the position the interviewer had forged among the injecting drug user community. Does providing information threaten the veracity of the research process? Although the information was provided after the data had been gathered, and thus would not have contaminated the data in this sense, it is conceivable that respondents may have passed on information to their peers who were subsequently interviewed, thus biasing the data. However, HIV/AIDS poses such a certain threat to life that we consider it unethical not to provide subjects with information which helps them to avoid contracting it. Power discusses the controversy which surrounds this issue.I6 He describes Feldman’s view that in situations when contributions to science are pitted against benefits to the research subjects, the latter should be preferred.23 Power, unlike Feldman, argues that unbiased research should be the primary focus in most instances. However, he holds that this general position may be called into question where observed activities may endanger life. He considers that intervention should occur only when there is serious threat to the individuals concerned, and he argues that HIV/AIDS presents such a threat so that ‘in the current climate research objectivity may at times have to take a back seat to health education and risk reduction’ (p. 50). He argues for a serious consideration of an expanded role of research to assist in risk reduction. It must be acknowledged that there is a further unresolved tension here, in that objective, unbiased research is most likely to result in accurate HIV/AIDS information, and without this, research subjects. may ultimately suffer by receiving risk reduction information which may be unknowingly biased. This distinction between research and intervention is also pursued by Merton in the context of community-based HIV/AIDS drug trials.24 More generally, Merton cites criticism of the randomised controlled trial on scientific grounds as enforcing rigid rules of statistical and other interpretation which ‘by-passjudgment’. In other words, decisions as to what constitutesa sufficiently rigorous study are not objective and purely scientific, but depend upon the values of the decision maker, so that even among clinical researchers there are ‘pragmatists’ who think trials should reflect the realities of practice so that results can be generalised immediately, and the ‘fastidious’ who believe that trials should be designed to e l i n a t e confounding factors and any possibility of bias [p. 5081.
Merton agrees with Sonnabend’s argument that in the context of HIV/AIDS ‘the traditionally emphatic separation between the goals of research and treatment just won’t wash’ (p. 503), and considers that community-based HIV/AIDS research ‘must balance the values of scientific credibility with the moral principles of respect for persons, including autonomy, beneficence and justice’ (p. 502).25 It is impossible, in a city like Perth, where there are ongoing HIV/AIDS prevention initiatives, to control confounding factors and eliminate bias in research
AUSTRALIAN JOURNAL OF PUBLIC HEALTH
1992
VOL. I6 NO.
2
RESEARCH WITH INJECTING DRUG USERS
among injecting drug users. HIV/AIDS risk behaviour change among injecting drug users is almost certain to be the result of a multitude of interventions including media campaigns, the provision of accessible sterile injecting equipment, peer group influences and possibly, for a few, the information we provided. Research that maps changes in behaviour, attitudes and knowledge over time, in relation to the occurrence of various HIV/AIDS interventions, and qualitative research that explores the factors to which injecting drug users attribute their behaviour changes, can contribute insights into the influence of interventions on risk behaviour.
The research process as community intervention The urgent nature of preventing the spread of HIV/ AIDS among injecting drug users has required that we communicate relevant findings and implications for HIV prevention in diverse ways as quickly and as widely as possible before publishing them in refereed journals. The ways in which we have communicated study findings rely strongly on maintaining close and regular liaison with those interested in and working in the HIV/AIDS and injecting drug user area. Many of these individuals and organisations have aided us in recruiting subjects and in providing them with up-to-date information. We hope our research has been reciprocally helpful in guiding them in the formulation of HIV/AIDS prevention initiatives. Examples of our community involvement are as follows: membership of a state interdisciplinary committee on HIV/AIDS and injecting drug users; membership of the local users’ collective, assistance with pharmacy workshops about the provision o:f needles to users; participation in drug use working parties; conference presentations; media input; the distribution of ANAIDUS findings to a number of places including the Health Department of Western Australia, the Western Australian AIDS Council, the Western Australian Alcohol and Drug Authority and other interested parties; and consultation about HIV/AIDS and injecting drug users with various people referred to us. We believe this involvement to have been mutually beneficial in that they received information from us as soon as it was available, and they were extremely helpful to us in both publicising the study and encouraging injecting drug users to participate. Acknowledgments Without the help and support of a number of agencies and individuals this research would have been very difficult, if not impossible to complete. In particular we would like to thank all those who attempted to help us in recruiting respondents to the: study, our co-investigators in Perth, Professor David Hawks (Director of the National Centre for Research into the Prevention of Drug Abuse) and Dr Allan Quigley (Director of Clinical Services for the West.. ern Australian Alcohol and Drug Authority), our colleagues involved in the ANAIDUS in other centres
around Australia and Bayer Glucolet for the provision of the glucolets and lancets with which we obtained blood samples. Most importantly of all, we would like to thank those injecting drug users who gave of their time, energy and trust to be interviewed. This research was funded by the Commonwealth AIDS Research Grants Committee. References Des Jarlais DC, Friedman SR, Novick DM, Sotheran JL et al. HIV-1 infection among intravenousdrug users in Manhattan, New York City, from 1977 through 1987.JAMA 1989;261: 1008-12. 2. Stimson GV, Donoghoe M,AUdritt L, Dolan K. HIV transmission risk behaviour of clients attending syrhge-exchange schemes in England and Scotland. BrJ Addict 1988;83: 1449-55. 3. Wodak A. Wffl the sky fall in? The prevention of HIV infection in intravenous drug w r s in Australia. Paper presented at the Fourth National Conference on AIDS, Canberra, 9-1 1 July 1990. Sydney: Alcohol and Drug %vice, St Vincent’s Hospital. 4. Donoghoe MC. Stimson GV, Dolan KA. Sexual behaviour of injecting drug users and associated risk of HIV infection for non-injecting sexual partners. AIDS Cure 1989;l(1):51-8. 5. McKeganey N, Barnard M, Watson H. HIV related risk behaviour among a non-clinic sample of injecting drug users. B r j Addid 1989;
1.
84;1481-90. 6. D e s Jarlais DC, Friedman SR. Target groups for preventing AIDS among intravenous drug users. J Appl Soc Psgchol 1987; 17(3): 251-68. 7. Robertson J, Skidmore C, Roberts J. HIV infection in intravenous drug users: a follow-up study indicating changes in risk-taking behaviour. BrJ Addid 1988;83: 387-91. 8. van den Hoek A, van Haastrecht HJ, Coutinho RA. Heterosexual behaviour of intravenous drug users in Amsterdam: implications for the AIDS epidemic. AIDS 1990;4:449-53. 9. Darke S, Hall W. Carless J. Drug use, injecting practices and sexual behaviour of opioid users in Sydney, Australia. BrJ Addict 1990;85:
1603-9. 10. Power K,Hartnoll R, Daiaud E.Drug injecting,AIDS, and risk behaviour: potential for change and intervention strategies. Br J Addict 1988;83: 649-54. 11. Watson J. Rcpoll of the selcd committee appointed to inquire into the National HW/AIDS Strategy White Paper. Perth: Australian Government Publishing Service, 1990. 12. Wiebel WW. Identifying and gainiing access to hidden populations. NIDA Rcs Mmgr 1990;98:4-11. 13. Adler P. Ethnopphic research on hidden populations: penetrating the drug world. NIDA Res Mmgr 1990;98:96112. 14. Goldstein PJ, Spunt BJ. Miller T, Bellucci P. Ethnographic field stations. NIDA Res Mmgr 1990; 98:80-95. 15. Dickie M.Can education influence the behaviour of injecting drug users? Paper presented at 3rd National Conference on AIDS: Living With AIDS in the Year 2000.Hobart, August, 1988.Sydney: Centre for Education and Information on Drugs and Alcohol. 16. Power R. Participant observation and its place in the study of illicit drug abuse. BrJ Addict 1989;84:43-52. 17. Dobinson I, Poletti P. Buying and selling heroin: a study of heroin user/dealers. Sydney: NSW Bureau of Crime Statistics and Research, 1988. 18. Biemacki P, Waldorf D. Snowball sampling. Biol Mcllbd Res 1981; lO(2): 141-63. 19. Morrison V, Plant M. Drug problems and pattems of service use among illicit drug users in Edinburgh. Br J Addid 1990; 85: 547-54. 20. Reilly C, Flaherty B, Home1 P. A study of 1621 year old illicit drug w r s in Sydney. Conference paper presented at W m t a School in the Sun. Brisbane, 3-6 July, 1990. Sydney: Dimtorate of the Drug Offensive, NSW Health Department. 21. Wiebel WW. Combining ethnographic and epidemiologic methods in targeted AIDS interventions: the Chicago model. NIDA Rrs M m g r 80; 1988:137-50. 22. Miller WR. Increasing motivation for change. In: Hester RK, Miller Wa, eds. Handbook of alcoholism treatment appmchcr. New Yorlr: P e r g a m ~1989. . 23. Feldman HW. Street status and the drug researcher: issues in participant-observation. Washington, D C Drug Abuse Council, 1974. 24. Merton V. Community-based AIDS research. Eual RN 1990;14(5): 502-97. 25. Sonnabend J. Do we need new ways to evaluate experimental AIDS treatment? AIDS Farm 1988; l(November): 4-8.
AUSTRALIAN JOURNAL OF PUBLIC HEALTH 1992 vot. 16 NO. 2
187
