Where is the Sex in Harm Reduction

Tim Rhodes and Alan Quirk, The Centre for Research on Drugs and Health Behaviour, London, UK

As the concept of harm reduction comes of age, it is becoming increasingly important to face up to the challenges of putting the ideas into practice. Every year, at the International Conference on the Reduction of Drug Related Harm, researchers and practitioners meet to discuss how this is best achieved. This March the conference was held in Florence. The conference was a mixed bag of presentations which told of harm-reduction successes as well as failures.


Harm reduction successes and failures are inextricably linked with the social, cultural and political context in which such activities are attempted. In many US states, for example, ‘harm reduction’ is fast becoming a dirty word. For many North American delegates in attendance at the conference in Florence, harm reduction has come to symbolise part of a wider political struggle where efforts are directed towards the protection of individual human rights against health and drug policies which often favour harm maximisation over harm minimisation.

One symposium session, in which George Clar; (NY), Joyce Rivera (NY) and Lisa Moore (SF) specifically discussed the harms and benefits associated with US drug and penal policy, concluded that one of the biggest problems for harm reduction in the USA was that it operated in the context of one of the most underdeveloped of overdeveloped nations (Clark, 1995; Moore, 1995; Rivera, 1995). Sound harm-reduction ideas may come to little when faced with the practical need to dismantle a political machinery intent on erasing the words ‘harm reduction’ from its policy vocabulary. This is a story of where the odds are stacked against ‘success’ yet where groups and communities of drug users themselves have proved that harm reduction is possible.

However, the challenge to remove policy impediments to individual and community attempts to minimise harm demands the efforts of drug users and researchers alike. Far less of the research presented at the Florence conference was attuned to tackling problems of social or political context than was attuned to tackling problems of individual attitude or behaviour change. The lesson to be learnt is that scientific endeavours will only have practical harm reduction outcomes when they understand the interplay between the individual and his or her context or environment.

This demands a social epidemiology of health behaviour which aims to make sense of individual ; beliefs and behaviours in the context of legal and policy influences as well as the actions of others. Here there are two challenges for future harm-reduction intervention and research. First, to understand that the ways in which individuals think and behave relates to how other individuals think and behave.


What you and I do when we take drugs or have sex depends on what we see as normal and appropriate behaviour. What we see as normal and appropriate depends not simply on some sort of endogenous ideal within our own heads, but on what is socially and culturally organised as acceptable behaviour. We act in direct response to others’ actions and they do likewise with us. The challenge for future harm reduction is therefore to support group ‘norms’ which are supportive of reducing harm and, where necessary, to encourage changes in unsafe ‘norms’ which encourage harmful or risky behaviour.

Second, we need to understand more about how group norms connect with the legal and policy context. Even peer- or group-oriented interventions and norm-changing strategies run the risk of failure if the costs of group change outweigh their benefits. In the same way as individual ‘choices’ and attempts at behaviour change can be seen to operate within the constraints of group and community norms, collective action and change can be seen to operate within the constraints of the political and legal environment. The challenge is for all harm-reduction intervention and research to recognise that there is often as much a need for political change as there is for individual or community change.


There is one incisive example of how interventions oriented towards individuals and individual change have had limited success because of the difficulties associated with encouraging concomitant changes in social ‘norms’. This is the example of safer sex. Just as there are harms as well as benefits associated with the use and injection of drugs, there are harms as well as benefits associated with having sex. It seems that many interventions continue to side-step the reduction of sex-related harm. Yet the potential harms associated with ‘unsafe’ or unprotected sex are enormous. As drug users continue to reduce their individual harms directly associated with drug use, and particularly injecting drug use, sexual transmission is becoming increasingly important in determining the future dynamics of HIV epidemic spread. The next stage of the HIV epidemic among drug injectors is likely to be significantly associated with whether or not, and with whom, sex is safe. This is the case worldwide in developed and developing countries.

HIV infection is primarily a sexually transmitted disease. Given that most drug users and injectors have sex (Rhodes et al.,1994), it is timely that we once again ask: ‘Where is the sex in harm reduction?’ This is not to suggest that a ‘sexual re-awakening’ has not occurred among drug workers and researchers. Over the last 5 years there has been increasing effort directed towards maximising sexual safety and minimising sexual risk in the lives of drug users and their sexual partners. These efforts are, however, often luke-warm and have had little effect in actually encouraging sexual behaviour change (Rhodes and Green, 1995). Almost all surveys among intravenous drug users (IDUs) worldwide indicate continued reductions in injecting risk behaviour yet little or no evidence of sexual behaviour change.

Research on injecting drug use presented at conferences, such as the harm-reduction conference in Florence, has repeatedly informed delegates that most IDUs are sexually active, that rates of partner change are relatively high, that there is a high degree of sexual mixing between IDUs and non-lDUs, that a significant minority of IDUs are involved in commercial sex work, that most IDUs report never having safe or protected sex with the* primary partners, and that a significant minority report never having safe or protected sex with their casual partners. There is an accompanying research literature which adds weight to these findings (Stimson,1991; Des Jarlais, 1992; Donoghoe, 1992; Rhodes, 1994; Rhodes et al.,1994). We have got as far as exchanging knowledge about the fact that there is a considerable amount of unsafe sex going on but we have made few effective attempts to understand systematically why sexual behaviour changes are so difficult or how such changes might be encouraged (Rhodes and Quirk, 1995a).


The first step towards answering these questions is to move beyond the strictures of survey designs towards research designs better suited to answering such questions as why and how. We need to begin by describing, from the perspectives of drug users them selves, what is meant by ‘safe’ or ‘safer’ sex. Ongoing qualitative research on sexual safety among opiate and stimulant users in London has begun by asking the obvious: what is sexual safety ( Rhodes and Quirk, 1995b;Rhodesetal.,1995a)

This research points to limitations in current epidemiological indicators of ‘safe’ and ‘unsafe’ sex. Contemporary research and intervention projects invariably equate ‘sexual safety’ with condom use. In most surveys, for example, unprotected penetrative sex is viewed as ‘unsafe sex’ whereas the use of condoms is seen as ‘safe’ or ‘safer’ sex. These indicators appear to be inadequate measures of how drug users perceive sexual safety.

First, our findings show that many drug users who report using condoms also have unprotected penetration prior to condom use. We have termed this ‘unsafe protected sex’ (Rhodes and Quirk, 1995b). Having ‘unsafe protected sex’ may prevent unwanted pregnancy but it does not necessarily minimise the risk of transmitting HIV or other sexually transmitted diseases (STDs). This means that ‘protected’ or ‘safer’ sex is often a matter of degree. Surveys which rely on drug users answering ‘yes’ or ‘no’ to whether or not, and how often, they use condoms run the risk of underestimating the proportions of drug users who may actually be having ‘unsafe sex’. This may seem an obvious research finding but it nevertheless has important implications for future harm reduction research and intervention.

Taken together with findings that suggest that the use of certain drugs may increase the likelihood of condom breakage or failure (e.g. as a result of pro longed penetration), it can probably be concluded that we currently know very little about how ‘safe the use of condoms actually is.

Second, future research needs to study not simply whether drug users say they use condoms or not, but the processes which influence why and how sex is deemed safe or unsafe. In the first instance this requires a description of how ‘risk’ and ‘danger’ are perceived. There are many potential harms associated with drug using and sexual lifestyles. Our research shows that perceptions of HIV risk are often viewed as relative concerns. HIV risk may be viewed as less important, for example, than the risk of overdose or of having no money or accommodation for the night.

Of crucial importance are findings that highlight that drug users do not consider unprotected sex to be a high risk priority. This is in spite of making considerable efforts to minimise their exposure to drugrelated harms. For some this is because they consider themselves to be having what we have termed ‘safe unprotected sex’ (Rhodes and Quirk,1995b). This is where partners’ knowledge of HIV or STD risk is based on trust. However, our findings also point to the relatively high degree of non-disclosure and dishonesty in relationships about drug use, injecting, sexual history and sometimes, also, HlV positive antibody status. For others, unprotected sex may be seen to carry less risk or danger than initiating condom use with their partners. This may be because of the threat of sexual violence or because of the loss of other ‘benefits’ associated with unprotected sex, such as displays of love, trust, commitment or permanency in relationships. But for most drug users we interviewed, unprotected sex was seen to be normal within heterosexual relationships.

Our data show that everyday ‘norms’ encourage safer drug use but unsafe sexual practices. The everyday normality and legitimacy of unprotected 19 sex in heterosexual relationships (particularly ‘long-term’ relationships) help explain some of the anomalies in drug users’ reported risk behaviour. One such anomaly is that injectors may never share needles, even with sexual partners, and yet continue to report unprotected sex, even with casual and sometimes also HIV positive partners. One of the most striking findings of survey research is that reported levels of condom use among drug injectors is almost identical to that among the heterosexual population in general. Perceptions that unprotected sex is ‘normal’ in heterosexual relationships are likely to be similar among drug users and non drug users alike. This points to the depth and parameters of social change necessary before future studies can report that most drug users always use condoms.


We have argued that future research needs to have a systematic understanding of how and why unsafe sex occurs and not simply an understanding of whether or not sex is safe or unsafe. This in itself would make a considerable advance on most current research findings. We have also suggested that, to do this, future research needs to employ a variety of research methods (particularly qualitative) so as to describe drug users’ sexual behaviours in the specific social contexts in which they occur. In short, the aim is to understand what sexual safety (and unprotected sex) means to drug users themselves. This, at the end of the day, is going to yield results of a far more practical value than measuring sexual safety as it is defined by epidemiologists for the purposes of scientific enquiry. Finally, we have emphasised that everyday ‘norms’ regulating drug users’ sexual behaviours tend to be more supportive of unprotected sex than they are of protected sex.

This leads to the question of whether and how social norms about sexual behaviour can be changed. We have argued that at the outset harm-reduction interventions often need to encourage changes in social norms and social context before individual behaviour changes become possible. Studies have observed how norms regulating everyday injecting behaviour have changed and how norms supportive of safer drug use can reinforce and sustain risk reduction behaviour (Burt and Stimson, 1993). We are equally aware that risk minimisation is not risk elimination and that unsafe drug use continues under certain social circumstances (Rhodes, 1994). But the aim and target of changing social norms about heterosexual sexual behaviour seem like a tall order. It would appear to require a major piece of social engineering which cannot (and perhaps should not) be the function of harm-reduction interventions.

It is our contention that future research and debate in harm reduction should not neglect the question of why sexual behaviour changes are so difficult or of how such change can be encouraged. We are aware, however, that this is a tempting option. It is tempting to view the changing of sexual norms as an intractable problem, but this is an adequate state of affairs. Harm-reduction research and intervention have got as far as recognising the importance of sexual norms in influencing whether or not unsafe sex occurs, and as far as advocating their change. It is necessary to debate whether and how this is possible if we are to prevent such research going round in circles. It is of limited practical value when attending international harm-reduction conferences to hear, yet again, the latest annual update on the fact that most drug injectors report never using condoms. We also need to hear of the practical successes and failures of interventions targeting sexual behaviour change and of research that questions whether, why and how sexual norms and behaviour can change.


There is an ongoing debate about whether it is possible to encourage group-mediated change within networks of drug users. Research on this topic has shown that group-mediated changes are possible. Studies show that changing peer-based norms about drug use can be a more effective way of creating and sustaining behaviour change than targeting and changing individual beliefs and behaviours alone (Friedman et al., 1992; Kelly et al.,1992; Trotter et al.,1993). It is now possible to say that in spite of the ‘predatory’ nature of some drug users’ social relationships, there is also the possibility for endorsing or encouraging an interconnected system of peer support. These methods of intervention, which borrow from social network analysis and social diffusion theory, hold more hope for the changing of sexual norms than most other methods of intervention.

One key problem remains. This is that sexual norms cut across drug-using and non-using populations. Indiv iduals interconnected within a drug-using network do so largely as a function of their drug use. The use, dealing, distribution and buying of drugs are facilitated by access to drug-using networks. It is perhaps fair to say that the ‘master status’ of these networks is that they are characterised by drug use. In contrast, drug-using networks are not often socially organised on aspects of sexual behaviour. This makes the introduct ion of group changes difficult. There may be no existing channels of communication or influence along which ideas of sexual responsibility and sexual behaviour change can diffuse. It is precisely the fact that sexual norms are part and parcel of everyday social life for us all, drug users or not, that makes the changing of sexual norms so difficult within the restricted parameters of subgroups or subcultures.

The social or sexual network is, however, perhaps the most viable agent and target of social change among drug-using populations. The job of harm reduction is to attempt experiments in social diffusion interventions designed to encourage mutual responsibilities about safer drug use and safer sex. We have yet to try such experiments systematically and are only just beginning to appreciate their value in encouraging group-mediated changes about injecting drug use. Not only is it necessary to encourage a closer understanding of what it is that is meant by sexual safety in the context of drug users’ lives but it is necessary to understand the social processes that influence why and how safe and unsafe sex occurs. It is on this understanding that it becomes possible to design and experiment with social diffusion interventions and other norm-changing strategies. We believe that it is at least possible to learn from within the parameters and lessons of previous harm-reduction research. We need look no further than ethnographic studies of the social function of syringe sharing to realise the practical advances of investigating the social processes which influence why individuals behave as they do and how such behaviours can change.


I am particularly grateful for the comments and help of Gerry Stimson and to the Department of Health for funding our ongoing qualitative work on sexual safety and drug use.

Tim Rhodes and Alan Quirk, The Centre for Research on Drugs and Health Behaviour (CRDHB), Charing Cross and Westminster Medical School, 200 Seagrave Road, University of London, SW6 1RQ, UK.CRDHB is core funded by North Thames Region.


Burt ), Stimson GV (1993) . Drug Ir4ectors and HIV Risk Reduction: Strategies for Protection. London: HEA.

Des Jarlais DC (1992). The first and second decade of AIDS among injecting drug users. British Journal of Addiction 87: 347-53.

Donoghoe MC (1992). Sex, HIV and the injecting drug user. British Journal of Addiction 87: 405-16.

Clark G (1995) . Harm reduction coalition: fostering change at a national level. Paper presented at the Sixth International Conference on the Reduction of Drug Related Harm, Florence.

Friedman SR, Neaigus A, Des Jarlais DC et al. (1992). Social intervention against AIDS among injecting drug users. British Journal of Addiction 87: 393-404.

KellyJ, StLawrence J, BradheldT (1992). Community AIDS/HIV risk reduction: the effects of endorsements by popular people in three cities. American Journal of Public Health 80: 1483-9.

Moore L (1995). The US war on drugs: hegemony versus harm reduction. Paper presented at the Sixth International Conference on the Reduction of Drug Related Harm, Florence.

Rhodes T, Quirk A (1995a). Heroin, risk and sexual safety: some problems for interventions encouraging community change. In T Rhodes, R Hartnoll (Eds), AIDS, Drugs and Prevention. London Routledge.

Rhodes T, Quirk A (1995b) . Sexual safety and drug use. Druglink in press.

Rhodes T, Donoghoe MC, Hunter GM et al. (1994). Sexual behaviour of drug injectors in London; implications for HIV transmission and HIV prevention. Addiction 89: 1085-96.

Rhodes T, Quirk A, Stimson GV (1995) . Sexual Safety in the Context of Drug Taking and Sexual Lifestyles. London: CRDHB.

Rivera J (1995) . The ‘Heart and Soul’ of harm reduction. Paper presented at the Sixth International Conference on thc Reduction of Drug Related Harm, Florence.

StimsonGV (1991) . Risk reduction by drug users with regard to HIV infection. International Journal of Psychiatry 3: 401-15.

Trotter RT, Potter JM, Bowen AM (1993) . Networks and ethnography: group-based HIV prevention for small town and rural drug users. Unpublished paper, University of Arizona.