Examining links between cocaine use and street-based sex work in new South Wales, Australia.

There has been increasing interest in recent years in the study of sex workers as a marginalized group at increased risk for a range of poorer mental and physical health outcomes (Vanwesenbeeck, 2001). Previous research has documented the risks of blood-borne virus (BBV) transmission and sexually transmitted infections (STI) among sex workers due to unprotected sex with clients (Gossop, Powis, Griffiths, & Strang, 1995) and the potential risks posed to the broader community via BBV transmission through clients to the general population (Tuan et al., 2004).

Illegal drug use has previously been documented as common among street-based sex workers in particular (e.g., Gossop et al., 1995; Inciardi & Surratt, 2001; Leggett, 2001; Minichiello, Marino, Khan, & Browne, 2003; Roxburgh, Degenhardt, & Breen, 2005; Tuan et al., 2004). Drug use has been related to increased risk behaviors, including riskier sexual encounters (Minichiello et al., 2003) and risk of BBV transmission due to injection drug use and needle-sharing (Gossop et al.). Cocaine use in particular has been associated with significant BBV risk and sex risk behaviors among injection drug users (Hudgins, McCusker, & Stoddard, 1995; Tyndall et al., 2003).

Research indicates that drug use is an important predictor for poorer outcomes for sex workers, which has generated an interest in the role of drug use and drug use patterns among this group (Alegria et al., 1994). An ethnographic study of women in New York who engaged in sex work found that drug use played a substantial role in the way these women conducted their sex work (Maher, 1997). Crack-cocaine had a particularly deleterious effect on sex workers, as it was thought to lead to lowering of the price of sex work exchanges, to engender a hostile environment among sex workers and violent exchanges with clients, and to lead to an increased potential for high-risk sexual encounters. Many of the women interviewed also used crack in order to facilitate their engagement in sex work (Maher).

The Sex Work Industry in New South Wales

New South Wales (NSW), Australia, has a good tradition of measures designed to reduce harms related to the public health risks posed by sex work. Legalization exists in the form of brothels that operate under permits, and sex workers may operate from their own homes after obtaining permits from local councils, and along public thoroughfares as long as they are not in close proximity to churches or schools (Perkins, 1991). Outreach efforts have targeted sex workers in NSW with a strong focus on the reduction of sexual and other risk behaviors among this group. Harm-reduction and treatment measures for street-based sex workers engaging in injection drug use have been developed and targeted toward heroin use and injection, particularly outside the inner-city area.

Despite links between crack-cocaine use and sex work in the United Kingdom and United States (Gossop et al., 1995; Green, Day, & Ward, 2000; Inciardi & Surratt, 2001), less work has investigated in detail the potential relationships between cocaine use and street-based sex work in Australia. Given the apparent importance of context for both drug use and sexual encounters for sex workers (Aalbers, 2005; Fullilove, Lown, & Fullilove, 1992; Fullilove & McGrath, 2005; Green, Fullilove, & Fullilove, 2005; Maher, 1997), this is an important issue.

Drug Markets in NSW

Kings Cross has long been considered Australia's premier illegal drug market. American soldiers visiting Sydney on "rest and recreation" leave during the Vietnam War (1962-1972) introduced heroin to prostitutes and bohemians (McCoy, 1980). By the 1980s, Kings Cross was home to a thriving open-air drug market dealing in heroin, cannabis, and amphetamine. The drug market offered opportunities for small- and large-scale drug dealing (Southgate et al., 2003). The district attracted local and out-of-town injection drug users (IDU), sex workers, young people, a nightclub crowd, backpackers, and an array of tourists and visitors (Southgate et al.). There is good evidence that the NSW heroin market increased substantially during the 1990s (Gibson et al., 2003). IDU reports suggested that heroin dominated illegal drug markets in NSW; purity was relatively high; and the price of heroin either remained stable or decreased every year (Darke, Hall, & Topp, 2000).

In early 2001, however, there were reports of a dramatic decline in the availability of heroin in Sydney (Day et al., 2003; Weatherburn, Jones, Freeman, & Makkai, 2003). These were confirmed by the Illicit Drug Reporting System (IDRS), Australia's strategic early warning system, which documented a reduction in the availability and street-level purity of heroin and an increase in price (Day, Degenhardt, & Hall, in press; Topp et al., 2002). This reduction in availability--the so-called "heroin shortage" (United Nations Office for Drug Control and Crime Prevention, 2002)--was greatest from January to April 2001, but the market does not appear to have returned to levels of supply seen prior to the onset of the shortage. Higher prices, lower street purity, and the shrinking of street-level heroin markets has continued (Black, Degenhardt, & Stafford, 2005; Day et al., in press; Degenhardt, Conroy, Gilmour, & Collins, 2005).

Historically, cocaine was rarely used by IDU in NSW; little or no crack use occurs in Australia (Topp et al., 2001). Following the reduction in heroin supply, however, the availability of cocaine increased at the street level in areas where heroin had previously been distributed (Degenhardt, Conroy, et al., 2005; Roxburgh, Degenhardt, & Breen, 2004). Evidence suggests that those involved in street-level and mid-level heroin distribution began distributing cocaine when heroin became less available (Degenhardt, Conroy et al.), probably to replace lost income when heroin was not available. A street sale of cocaine decreased in price relative to 2000, but patterns of use were more frequent than for heroin (Roxburgh, Degenhardt, & Breen, 2004).

The reduction in heroin supply was associated with a short-term spike in robberies and home burglaries (Degenhardt, Conroy, et al., 2005; Weatherburn et al., 2003). It was hypothesized that the increase in acquisitive crime was related to increased cocaine use, due to its higher cost, greater frequency of use compared to heroin, and the pharmacological effects of the drug (Degenhardt, Day, Hall, Conroy, & Gilmour, 2005).

This Study

Given the association between cocaine use and street-based sex work in other countries, we conducted a natural experiment studying the relationship between cocaine use and street-based sex work. We were particularly interested in examining the changes that occurred in an area where harm-reduction measures had been well-established within the current injection drug market, which had historically been dominated by heroin. We had three goals in this study: (a) to examine changes in injection cocaine use in NSW over the past decade, particularly when the availability of heroin reduced in 2001 (see also Degenhardt, Conroy, et al., 2005; Degenhardt, Day, Gilmour, & Hall, in press; Roxburgh et al., 2004); (b) to examine whether there was evidence to suggest that street-based sex workers increased their cocaine use when heroin became less available, and if so, to consider the impact of such use upon these workers; and (c) to examine potential changes in the number of incidents of illegal street-based sex work over this time.

METHOD

We adopted three broad approaches to address the research questions. The first was an analysis of police data on recorded incidents for use and possession of cocaine and prostitution. We assumed in this analysis that incidents for cocaine use/possession and prostitution are, over the short timeframe covered by this study, a reasonable proxy for involvement in these two classes of offense. We examined this assumption by assessing whether specific operations were targeted toward either drug users or street-based sex workers during the study period. No such operations were reported by any of the key informants interviewed. The trends in cocaine use/possession offenses are also completely consistent with trends in other indicators of cocaine use, cocaine treatment episodes, hospital and emergency room admissions, and calls to drug telephone helplines about cocaine use (Black et al., 2005). Figure 1 shows the similarity of trends in possession/use offenses, telephone help-line calls, and emergency room admissions.

[FIGURE 1 OMITTED]

The second approach involved a series of semi-structured interviews with IDU and other key informants in the Sydney heroin market, such as police officers and health workers. These interviews assessed issues related to the heroin shortage and included discussion of trends in drug use among injection drug users and illegal activities such as street-based sex work. Third, we examined data collected since 1996 from interviews with regular injection drug users from the IDRS to consider trends in cocaine use over time among this group.

Human Participant Protection

This project was approved by the University of New South Wales Human Research Ethics Committee (HREC), the South East Sydney Area Health Service HREC, the South West Sydney Area Health Service HREC, the Central Sydney Area Health Service HREC, and the Aboriginal Health and Medical Research Ethics Committee. All participants gave written consent for the interviews and use of their information in the generation of project reports, journal articles, and conference proceedings relating to the project, providing they were unable to be identified.

Data Used in the Study

Structured Interviews With Regular IDU

As part of the IDRS, approximately 150 IDU have been surveyed annually in Sydney, NSW, since 1996. To be eligible to participate in the survey, IDU needed to have been injecting at least monthly during the six months preceding the interview, and to have been a resident for at least 12 months in Sydney. Participants were recruited through multiple methods, including treatment agencies, needle and syringe programs, and peer referral. (For further details on the IDRS IDU interview methodology and results, see Black et al., 2005, and Stafford et al., 2005). Each year, IDU were asked about their usc of a variety of drugs, including cocaine. The sample was recruited in the same way and with the same criteria across years but did not necessarily include the same participants. A significant proportion of the sample each year reported engaging in sex for money or drugs (Black et al.; Roxburgh et al., 2005).

Semi-Structured Interviews With Heroin Users

Heroin users were recruited via advertisements placed in opioid pharmacotherapy clinics. They had to have recent experience in the drug market and to have commenced pharmacotherapy between August and December 2000 (pre-shortage) or between February and April 2001 (during shortage). Fifty-three users were interviewed in total, approximately half in each time group. Users were surveyed on a range of issues, including their involvement and experience of drug markets prior to and during the heroin shortage. The surveys included discussions about income-generating activities of both the participants and the IDU with whom they had contact (Day, Gibson, Collins, Degenhardt, & Dietze, 2004).

Semi-Structured Interviews With Key Informants

A strategic approach was taken to the recruitment of key informants (KI), where we attempted to obtain a comprehensive coverage of drug markets from the health and law enforcement sides in each drug market. The different numbers across different areas reflect the size of the personnel base involved in work related to drug activities and the extent of service provision in that area. Selection of KI from all three sectors was based on one or more of the following criteria: (a) the extent of their contact with the illegal drug market; (b) their level of knowledge of the illegal drug market and drug users; (c) the focus of their position (e.g., direct/indirect, operational/policy); and (d) the length of time the KI had held the position, particularly with regard to their ability to comment on changes over time, pre- to post-heroin shortage.

All interviews were conducted by trained researchers using a semi-structured interview schedule. The interviews explored respondents' tasks at the time of the shortage; effects of the shortage on IDU drug use; IDU treatment-seeking behavior; psychosocial and general health of drug users; effects on communities and drug users; and the effect of the shortage on community and health service provision.

All interviews were tape recorded, unless otherwise requested by the interviewee, and transcribed, and thematic analysis of the transcripts was performed. Verbatim quotes are presented for illustration where appropriate. This analysis was performed with the aid of Microsoft Excel. All KI gave their informed consent to participate.

Law enforcement. The NSW Police Service includes three levels of command of relevance to this study: State, Region, and Local Area Command (LAC). Key informants were selected across all three levels and across the four LAC responsible for policing the three Sydney open-air drug markets, two region commands in which these LAC were located, and a range of squads within the State Command, including squads focused on organized crime groups and drug crime. A total of 29 law enforcement KI were interviewed for this study, 20 of which were sworn officers. Table 1 provides further detail on the numbers of KI recruited for each drug market and police command level.

Health. Health sector KI were recruited from organizations involved in the support of drug users across the three Sydney illegal drug markets of Kings Cross, Cabramatta, and Redfern (Gibson et al., 2003). KI from State Government bureaucracies and organizations were also recruited to provide state-wide observations. KI were drawn from a range of positions within their respective organizations, from middle and upper management to front-line positions. A total of 49 health sector KI were recruited for this study, including five from NSW State organizations. Table 2 provides further detail on the number of KI recruited for each drug market and health discipline.

Policy/advocacy. An additional three KI were recruited for their experience and expertise in drug policy and advocacy.

NSW Police Incident Data

NSW Police record all police activity in a centralized database known as the Computerised Operational Policing System (COPS). This information can be analyzed at the level of either event or incident and is not reliant on a charge having been placed. An event is a record created in COPS whenever police attend a criminal or non-criminal activity. An event includes the incidents that comprise it: what happened, where, who was involved and the actions taken by police in response to the event.

Two incident types were used in this study: cocaine possession/use and prostitution. Incidents of possession/use were unlikely to be influenced by the outcome of proactive investigations and, in turn, unlikely to be influenced by changes in police activity and resources, but rather, a reflection of trends in use and detection of illegal drugs (in this case, cocaine).

It is not an offense in NSW to offer sex in return for money. The offense of prostitution occurs when a person solicits for sex in close proximity to certain designated locations, such as schools or hospitals. There are several key areas in Sydney where such illegal prostitution occurs, most of which are close to the major Sydney street drug markets.

The indicator data series were analyzed using an ARIMA model time series. Intervention models were fitted using SAS version 8.2. If the number of events per month in a subpopulation of the series was too small to analyze statistically, that subpopulation was described qualitatively.

RESULTS

Trends in Cocaine Use

We found clear increases in the use of cocaine among injection drug users in 2001 (see Figure 2), whether IDU were asked about their use in the previous day, the number of days used in the past six months, or whether it was the last drug they had injected. Figure 3 shows the number of incidents recorded for cocaine possession/use in NSW. This peaked at 64 in March 2001 and remained high throughout the year but declined in 2002. The modelled series (see Figure 3) showed that although police incidents for cocaine possession or use were at a steady level prior to the reduction in heroin supply, they increased significantly over the six months following the reduced heroin supply before later returning to lower levels than were seen prior to the heroin shortage. The maximum mean increase on pre-shortage levels was 269% (p < .0001), occurring two months after the shortage. This peak decayed to a long-term level that was 73% lower than the numbers seen prior to the onset of the heroin shortage (p < .0001).

[FIGURES 2-3 OMITTED]

Data on possession/use offenses by gender (not shown) indicated similar patterns for males and females, but the small number of cases in 2003-2004 precluded statistical analysis. The duration of the increase in recorded incidents was marked for males, with only an apparently short spike of one month for females. As noted in the methods section, the increase in cocaine possession/use offenses was consistent with increases observed in treatment episodes for cocaine and in hospital admissions for cocaine dependence (Black et al., 2005).

Trends in Illegal Sex Work

There was a rapid increase in the incidence of prostitution offenses (i.e., in illegal street-based sex work offenses), from a mean of 31.4 offences per month to a peak of 116.7, p < 0.0001, which occurred across NSW one month after the onset of the shortage. This represented a brief 272% increase in the number of offenses. The number of prostitution offenses returned slowly to pre-shortage levels over the next 15 months. Figure 4 shows both the observed and the modelled series for this offense category. The brief peak which occurred in January 2003 is probably representative of a brief police campaign occurring in connection with the New Year, but it had no impact on the magnitude and significance of parameter estimates for the model presented here, even when modelled in statistical analyses as a separate event.

[FIGURE 4 OMITTED]

Both KI and IDU reported that some women felt they had been forced into sex work as a way to meet the increasing financial demands associated with cocaine use. In Cabramatta, where prostitution had not previously been an issue, this increase was thought by one law enforcement KI to be due to an increase in the number of sex workers coming in from outside the Cabramatta area, because the sex workers believed that Cabramatta would be the first market to be supplied with heroin when the supply of heroin returned.

KI also noted that women were working longer hours as a consequence of the behavioral effects of cocaine, and in doing so, became more visible to police. Reduced condom use, and therefore increased risk of sexually transmitted infections (STI) and HIV to both the workers and their clients, was also reported. This was particularly evident in the Kings Cross drug market and with the younger sex workers, as evidenced by this KI's comment:

   Services up there were really worried about some very young
   workers who were into sex work and were using heroin and then
   went straight over to coke [cocaine], and they were very concerned
   at their level of use, and the fact that they were working so
   much and so therefore were at greater risk in terms of the numbers
   of clients they were seeing each day and access to condoms,
   given that they were working around the clock, virtually. (Health,
   Community Welfare, Upper Management)

Similar patterns were observed among male and transgender sex workers in the same areas. An outreach worker reported the increased use of cocaine among this group resulted in longer work hours (and therefore more visibility), increased violence between the workers, and greater risk to clients.

   Those who went from heroin to coke [cocaine] or speed
   [methamphetamine powder] have really shaken up their old
   lifestyle or lost it completely.

   [Q: What would they have lost?]
   Their house ... [Their] looks. Their ability to put on a face of
   makeup and not have it look absolutely shocking.

The increase in sex work was also related to greater levels of violence among sex workers because of a perceived desire to work longer hours.

   I think there's also been an increased amount of violence between
   sex workers, because they're more hungry for work. They need to
   be standing on that particular spot 24 hours a day.... Their focus
   became more desperate. They're focused on getting the money off
   the client. So that means they're more likely to roll him or give a
   poor or short service. So a level of professionalism dropped.
   (NSW, Health, Community Welfare, Middle Management)

KI working in outreach programs for sex workers reported feeling it necessary to respond to the changes in drug use among their client base. In particular, they reported needing to develop new educational material and adapting to the changing needs of their clients due to the disruption that cocaine use caused within this group. For example, there was a greater perception that services were more "crisis-driven" and that sex workers' lives were "falling apart" due to their escalating cocaine use.

   I think that we saw a lot more demand for welfare-related or
   drug-and alcohol-related services and referrals.... We had
   referral around emergency accommodation, emergency assistance,
   financial assistance, and detox.

DISCUSSION

This study found evidence of an increase in cocaine use among injection drug users following a significant reduction in heroin supply (Roxburgh et al., 2004; Topp, Day, & Degenhardt, 2003). Concurrent with these changes were increased incidents of prostitution offenses. Interviews with IDU (a significant proportion of whom had engaged in sex work themselves) and KI who had had frequent, direct contact with sex workers indicated that the increase in prostitution offenses stemmed from increased sex worker visibility, due to longer working hours, and increased discord among sex workers, due to increased competition for business. It was consistently reported that sex workers were working longer hours aided in part by their cocaine use, but also necessitated in part by the greater cost of cocaine use compared to heroin use.

These findings suggest high levels of drug use may play a more important role than sex work in decisions to engage in risk-taking behavior. Sex workers were more likely to have spent money and to have spent more money on drugs on the previous day than non-sex workers, most likely because sex workers have more disposable income than non-sex workers. However, it may also be an indicator of heavier drug use.

Although sex work is more commonly reported by female IDU than male IDU (Roxburgh et al., 2005), KI reported the observed changes in cocaine use and sex work practices for all sex workers--female, male, and transgender--suggesting risk was not confined to one group of sex workers nor to a group of clients with specific gender preferences.

Key informants reported clients in contact with outreach services became more chaotic and took more risks, which they attributed to cocaine use; this is consistent with previous research examining risk behaviors among injection cocaine users (Hudgins et al., 1995; Tyndall et al., 2003). The need for outreach services to adapt to the changing drug market was apparent.

This study is subject to the flaws that beset all natural experiments in that it is not possible to guarantee that the intervention being studied was the only event that affected cocaine use or sex work during the time period. However, similar research on rates of crime conducted in the same time period in both Victoria and South Australia provided a control series. These states were geographically isolated from NSW and both experienced a heroin shortage, but neither experienced any increase in cocaine use or street-based sex work among IDU (Dietze et al., 2004; Harrison, Christie, Longo, Pointer, & Ali, 2004). Although it is possible that another event interfered in NSW drug markets at the same time as the heroin shortage, such possibilities were examined in a process of extensive crosschecking through KI, consultation with stakeholders, and analysis of other data sources in the wider project from which this study is drawn (Degenhardt & Day, 2004; Degenhardt, Day, & Hall, 2004). No plausible alternative explanations remained (Degenhardt & Day; Degenhardt et al., in press; Degenhardt et al., 2004; Roxburgh et al., 2004).

The increases in cocaine use and sex work offenses were not sustained, probably because of apparent decreases in cocaine availability in 2002 and a concomitant increase in heroin use among regular IDU (Roxburgh et al., 2004). However, there is no reason to believe that such a reduction in cocaine supply to IDU will persist. When cocaine use increases among injection drug users, so, too, does the extent of street-based sex work among this marginalized group, in part due to the higher cost of the drug and shorter duration of its effects, and in part due to its psychostimulant nature. The surge in prostitution was almost entirely a consequence of the temporary increases in cocaine consumption among IDU.

We obtained clear reports of increased sexual risk behaviors among street-based sex workers when cocaine use increased, including longer working hours, unprotected sex, and reductions in condom use in order to increase income. These behaviors increase the potential for transmission of BBVI and STIs among this group (as has been found previously; Hudgins et al., 1995; Tyndall et al., 2003). This has significant public health consequences which needs to be weighed against the public health benefits of reduced availability of heroin (van Beek, Dwyer, & Malcolm, 2001). Health services need to be adaptive to respond to such potential changes in drug use in this at-risk group if cocaine use returns.

In this study, increased cocaine use among injection drug users was associated with increased street-based sex work, increased visibility of these sex workers, and an apparent decline in the health and social stability of these workers. Positive effects of a sudden reduction in heroin use need to be balanced against a sudden increase in other behaviors that may have poor health consequences for IDU. Rapid changes in drug markets lead to rapid changes in the public health burden of injection drug use, which should be considered by health and law enforcement agencies. Given the increased risks among IDU during this period and the lack of appropriate harm-reduction strategies to target such behaviors, it would be useful to develop such interventions in anticipation of future returns to risky behavior.

Note. This research was part of a larger project funded by the National Drug Law Enforcement Research Fund. We thank Linette Collins, Amy Gibson, Libby Topp, Wayne Hall, and Peter Reuter for their input to the project, and staff at the Bureau of Crime Statistics and Research and NSW Police for providing data and assisting with interpretation.

Manuscript accepted August 30, 2005

REFERENCES

Aalbers, M. (2005). Big sister is watching you! Gender interaction and the unwritten rules of the Amsterdam red-light district. The Journal of Sex Research, 42(1), 54-62.

Alegria, M., Vera, M., Freeman, D. H., Robles, R., del C. Santos, M., & Rivera, C. L. (1994). HIV infection, risk behaviors, and depressive symptoms among Puerto Rican sex-workers. American Journal of Public Health, 84, 2,000-2,002.

Black, E., Degenhardt, L., & Stafford, J. (2005). New South Wales Drug Trends 2004: Findings from the Illicit Drug Reporting System (IDRS). NDARC Technical Report Number 211. Sydney: National Drug and Alcohol Research Centre, University of New South Wales.

Darke, S., Hall, W., & Topp, L. (2000). The Illicit Drug Reporting System (1DRS) 1996-2000. NDARC Technical Report 101 (NDARC Technical Report Number 101). Sydney: National Drug and Alcohol Research Centre, University of NSW.

Day, C., Degenhardt, L., & Hall, W. (in press). NSW heroin markets: Documenting the heroin shortage. Drug and Alcohol Review.

Day, C., Gibson, A., Collins, L., Degenhardt, L., & Dietzc, P. (2004). Research methodologies used to obtain retrospective self-reports of the impact of changes in heroin supply on regular users in and out of treatment. NDARC Technical Report No. 176 (NI)ARC Technical Report No. 176). Sydney: National Drug and Alcohol Research Centre, University of New South Wales.

Day, C., Topp, L., Rouen, D., Darke, S., Hall, W., & Dolan, K. (2003). Decreased heroin availability in Sydney in early 2001. Addiction, 98, 93-95.

Degenhardt, L., Conroy, E., Gilmour, S., & Collins, L. (2005). The effect of a reduction in heroin supply in Australia upon drug distribution and acquisitive crime. British Journal of Criminology, 45, 2-24.

Degenhardt, L., & Day, C. (Eds.). (2004). The course and consequences of the heroin shortage in New South Wales. NDLERF Monograph No. 4. Adelaide: Australasian Centre for Policing Research.

Degenhardt, L., Day, C., Gilmour, S., & Hall, W. (in press). Changes in community level drug use associated with a reduction in heroin supply. International Journal of Drug Policy.

Degenhardt, L., Day, C., & Hall, W. (Eds.). (2004). The causes, course and consequences of the heroin shortage in Australia. NDLERF Monograph No. 3. Adelaide: Australasian Centre for Policing Research.

Degenhardt, L., Day, C., Hall, W., Conroy, E., & Gilmour, S. (2005). Was an increase in cocaine use in New South Wales, Australia, accompanied by an increase in violent crime? BMC Public Health, 5(40).

Dietze, P., Miller, P., Clemens, S., Matthews, S., Gilmour, S., & Collins, L. (2004). The course and consequences of the heroin shortage in Victoria. NDLERF Monograph No. 6. Adelaide: Australasian Centre for Policing Research.

Fullilove, M., Lown, E., & Fullilove, R. (1992). Crack hos and skeezers: Traumatic experiences of women crack users. The Journal of Sex Research, 29(2), 275-287.

Fullilove, M., & McGrath, M. (2005). Introduction to the special issue on sexuality and place. The Journal of Sex Research, 42(1), 1-2.

Gibson, A., Degenhardt, L., Topp, L., Day, C., Hall, W., Dietze, P., et al. (2003). Global and Australian heroin markets. NDARC Technical Report No. 167. Sydney: National Drug and Alcohol Research Centre, University of NSW.

Gossop, M., Powis, B., Griffiths, P., & Strang, J. (1995). Female prostitutes in South London: Use of heroin, cocaine and alcohol and their relationship to health risk behaviors. AIDS Care, 7(3), 253-260.

Green, A., Day, S., & Ward, H. (2000). Crack cocaine and prostitution in London in the 1990s. Sociology of Health and Illness, 22(1), 27-39.

Green, L., Fullilove, M., & Fullilove, R. (2005). Remembering the lizard: Reconstructing sexuality in the rooms of narcotics anonymous. The Journal of Sex Research, 42(1), 28-34.

Harrison, A., Christie, P., Longo, M., Pointer, S., & Ali, R. (2004). The course and consequences of the heroin shortage in South Australia. NDLERF Monograph No. 5. Adelaide: Australasian Centre for Policing Research.

Hudgins, R., McCusker, J., & Stoddard, A. (1995). Cocaine use and risky injection and sexual behaviors. Drug and Alcohol Dependence, 37, 7-14.

Inciardi, J. A., & Surratt, H. L. (2001). Drug use, street crime, and sex-trading among cocaine-dependent women: Implications for public health and criminal justice policy. Journal of Psychoactive Drugs, 33(4), 379-389.

Leggett, T. (2001). Drugs, sex work, and HIV in three South African cities. Society in Transition, 32(1), 101-109.

Maher, L. (1997). Sexed work: Gender, race, and resistance in a Brooklyn drug market. New York: Oxford University Press, Inc.

McCoy, A. W. (1980). Drug traffic: Narcotics and organised crime in Australia. Sydney: Harper and Row.

Minichiello, V., Marino, R., Khan, M., & Browne, J. (2003). Alcohol and drug use in Australian male sex workers: Its relationship to the safety outcome of the sex encounter. AIDS Care, 15(4), 549-561.

Perkins, R. (1991). Working girls: Prostitutes, their life and social control. Canberra: Australian Institute of Criminology.

Roxburgh, A., Degenhardt, L., & Breen, C. (2004). Changes in patterns of drug use among injecting drug users following a reduction in the availability of heroin in New South Wales, Australia. Drug and Alcohol Review, 23(3), 287-294.

Roxburgh, A., Degenhardt, L., & Breen, C. (2005). Drug use and risk behaviors among injecting drug users: A comparison between sex workers and non-sex workers. Harm Reduction, 5(2).

Southgate, E., Day, C., Kimber, J., Weatherall, A., MacDonald, M., Woolcock, G., et al. (2003). Dealing with risk: A multidisciplinary study of injecting drug use, hepatitis C and other blood borne viruses in Australia (ANCD Research paper 7). Canberra: Australian National Council on Drugs.

Stafford, J., Degenhardt, L., Black, E., Bruno, R., Buckingham, K., Fetherston, J., et al. (2005). Australian Drug Trends 2004: Findings from the Illicit Drug Reporting System (IDRS). NDARC Monograph No. 55. Sydney: National Drug and Alcohol Research Centre, University of New South Wales.

Topp, L., Darke, S., Bruno, R., Fry, C., Hargreaves, K., Humeniuk, R., et al. (2001). Australian Drug Trends 2000: Findings from the Illicit Drug Reporting System (NDARC Monograph No. 47). Sydney: National Drug and Alcohol Research Centre.

Topp, L., Day, C., & Degenhardt, L. (2003). Changes in patterns of drug injection concurrent with a sustained reduction in the availability of heroin in Australia. Drug and Alcohol Dependence, 70, 275-286.

Topp, L., Kaye, S., Bruno, R., Longo, M., Williams, E, O'Reilly, B., et al. (2002). Australian Drug Trends 2001: Findings from the Illicit Drug Reporting System (IDRS). NDARC Monograph No. 48. Sydney: National Drug and Alcohol Research Centre.

Tuan, N. A., Hien, N. T., Chi, P. K., Giang, L. T., Thang, B. D., Long, H. T., et al. (2004). Intravenous drug use among street-based sex workers: A high-risk behaviour for HIV transmission. Sexually Transmitted Diseases, 31(1), 15-19.

Tyndall, M. W., Currie, S., Spittal, P., Li, K., Wood, E., O'Shaughnessy, M. V., & Schechter, M. T. (2003). Intensive injection cocaine use as the primary risk factor in the Vancouver HIV-1 epidemic. AIDS, 17(6), 887-893.

United Nations Office for Drug Control and Crime Prevention. (2002). Global Illicit Drug Trends 2002. Vienna: United Nations.

van Beek, I., Dwyer, R., & Malcolm, A. (2001). Cocaine: The sharp end of drug related harm! Drug and Alcohol Review, 20, 333-342.

Vanwesenbeeck, I. (2001). Another decade of social scientific work on sex work: A review of research, 1990-2000. Annual Review of Sex Research, 12, 242-289.

Weatherburn, D., Jones, C., Freeman, K., & Makkai, T. (2003). Supply control and harm reduction: Lessons from the Australian heroin "drought." Addiction, 98, 83-91.

Louisa Degenhardt, Carolyn Day, Elizabeth Conroy, and Stuart Gilmour

University of New South Wales, Sydney

Address correspondence to Louisa Degenhardt, National Drug and Alcohol Research Centre, University of New South Wales, Sydney, New South Wales, 2052, Australia; e-mail: [e-mail].

Table 1. Law Enforcement Sector Key Informants

                                Drug Market

Focus              Total   RF   KX   CAB   NSW   AUS   INT

Crime Mngt.          5           2    2     1
Drug Crime           7      1    2    1     1     1     2
Drug Policy          2                                  2
General Policing     5      3         2
Intelligence         5           1
Organized Crime      3           1          2
Serious Crime        1                1
Total               29      4    6    6     6     5     2

                              Role

                    UPR    MID
Focus              MNGT    MNGT   OPS    RCH

Crime Mngt.          1      3      1
Drug Crime           3      2      2      1
Drug Policy                        2
General Policing     3             2
Intelligence
Organized Crime      3
Serious Crime               1
Total               10      6      12     1

Note. RF = Redfern; KX = Kings Cross; CAB = Cabramatta; NSW = New
South Wales; AUS = Australia; INT = International; UPR MNGT = Upper
Management (no client contact); MID MNGT = Middle Management (some
client contact); OPS = Operations; RCH = Research.

Table 2. Health Sector Key Informants

                               Drug Market

Health
Focus               Total   RF   KX   CAB   NSW

Aboriginal            2      2
Community             3                 2    1
Community Welfare     5      1    1          3
Drug                 25      7   10     7    1
Emergency             1           1
Mental                2      1    1
Pre-natal             4      2    1     1
Primary Care          2      1          1
Youth                 5      1    2     2
Total                49     15   16    13    5

                            Role

Health               UPR    MID
Focus               MNGT    MNGT   CLIN

Aboriginal                    1      1
Community             2              1
Community Welfare     1       3      1
Drug                  6      13      6
Emergency                     1
Mental                        1      1
Pre-natal                            4
Primary Care                         2
Youth                         2      3
Total                 9      21     19

Note. RF = Redfern; KX = Kings Cross; CAB = Cabramatta; NSW = New
South Wales; UPR MNGT = Upper Management (no client contact); MID
MNGT = Middle Management (some client contact); CLIN = Clinical.