This article was downloaded by: [Memorial University of Newfoundland] On: 20 January 2015, At: 20:55 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
AIDS Care: Psychological and Sociomedical Aspects of AIDS/HIV Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/caic20
Differences across sexual orientation on HIV risk behaviours in injecting drug users a
b
c
M. W. Ross , A. Wodak , J. Gold & M. E. Miller
d
a
National Centre in HIV Social Research , University of New South Wales b
Alcohol and Drug Service , St Vincent's Hospital
c
Albion Street AIDS Centre, Sydney Hospital
d
Directorate of the Drug Offensive, New South Wales Department of Health , Australia Published online: 25 Sep 2007.
To cite this article: M. W. Ross , A. Wodak , J. Gold & M. E. Miller (1992) Differences across sexual orientation on HIV risk behaviours in injecting drug users, AIDS Care: Psychological and Socio-medical Aspects of AIDS/HIV, 4:2, 139-148, DOI: 10.1080/09540129208253085 To link to this article: http://dx.doi.org/10.1080/09540129208253085
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &
Downloaded by [Memorial University of Newfoundland] at 20:55 20 January 2015
Conditions of access and use can be found at http://www.tandfonline.com/page/termsand-conditions
AIDS CARE, VOL. 4, NO. 2, 1992
139
Differences across sexual orientation on HIV risk behaviours in injecting drug users
Downloaded by [Memorial University of Newfoundland] at 20:55 20 January 2015
M. W. ROSS,A. WODAK',J. GOLD^ &I M. E. MILLER3 National Centre in HIV Social Research, University of New South Wales, 'Alcohol and Drug Service, St Vincent's Hospital, *Albion Street AIDS Centre, Sydney Hospital, 3Directorate of the Drug Ofensive, New South Wales Department of Health, Australia
Abstract Injecting drug users (IDUs) play a disproportionate role in the spread of HIVgiven
their injecting and sexual contacts, and thereby act as conduits between these risk groups. We investigated diferences in risk behaviour and HIV seroprevalence in a Sydney sample of 1,245 IDUs. Significant diferences were observed across sexual orientation in HIV serostatus for males, with homosexual men having the highest HIV seroprevalence rate (35%), bisexual men intermediate (12%) and heterosexual men lowest (3%). Sexual HIV risk behaviours were lowest for homosexual men, intermediate for bisexual men, and highest for heterosexual men in the case of condom use: however, for numbers of partners, seroprevalence, and anal sex the trends were reversed. There were no diferences across sexual orientation fm either sex for injecting drug risk behaviours. Both male and female respondents reported having more than 50% of sexual contacts while under the influence of drugs. This study suggests that risk reduction in the sexual domain has not generalized to the injecting risk domain regardless of sexual orientation, and demonstrates that sexual risk behaviours in IDUs are lowest in homosexual, intermediate in bisexual, and highest in heterosexual IDU men. Introduction
In many western countries, the bisexual injecting drug user is positioned at the conjunction of a number of important bridges for HIV transmission. First, they are a potential conduit of HIV transmission from homosexual men into the wider drug using population; second, they are a potential conduit from injecting drug users (IDUs) to non drug-using heterosexual people (Marmor et al., 1990); third, they are independently of drug use a potential conduit between homosexual men and heterosexual men and women; and fourth, they are indirectly together with heterosexual IDUs, a conduit for vertical transmission. For these reasons, the bisexual IDU may play a disproportionately prominent rBle in transmission of HIV infection between and across groups of individuals whose behaviour places them at risk of HIV infection. These links are all the more important because a substantial proportion of male and female prostitutes also use drugs. Boles et al. (1989) found that 13% of a sample of male
Address for correspondence: Dr Michael Ross, National Centre in HIV Social Research, University of New South Wales, 345 Crown Street, Surry Hills NSW 2010, Australia.
Downloaded by [Memorial University of Newfoundland] at 20:55 20 January 2015
140 M. W.ROSS ET AL. prostitutes in the USA engaged in recreational sex with both males and females, and that 62% of the bisexuals (as compared with 37% of heterosexuals and 43% of homosexual hustlers) had injected drugs, particularly cocaine and heroin. Moreover, while male IDUs may sometimes provide homosexual sex commercially, they will more often have recreational heterosexual partners; conversely, a large proportion of female IDUs provide sexual services to men but some may also have homosexual contacts. Drug use and prostitution are commonly linked because prostitution is often used as a means of generating income to support a drug habit, or alternatively drugs are sometimes consumed to make the commercial provision of sexual services bearable emotionally. Williams (1990) provides clear evidence of the HIV transmission prospects from bisexual IDUs. He found that IDUs under 25 generally have partners who are not themselves IDUs, and that this appeared to be related to prostitution. In his sample of 131 IDUs recruited in the USA, Williams found that 15% reported sexual contact with both sexes in the past 6 months. Of those young adults who had had sex in that period, 38% had exchanged sex for money. The majority of male prostitutes reported that their sexual identity was strictly heterosexual and that they usually had relationships with young women who were frequently also IDUs and prostitutes. Williams suggests that contact between IDU and non-IDU populations is simple and direct: young female IDU prostitutes sharing injection equipment and having sex with clients at no other risk for HIV, as well as through bisexual clients of male prostitutes who are IDUs. Chu et al. (1989) found from an assessment of CDC data that bisexual men with AIDS in the USA were more than twice as likely to have injected drugs than homosexual men. This finding remained when the data were adjusted for race or ethnicity. Further, Battjes et al. (1989) noted that equipment sharing by male homosexual and bisexual IDUs with heterosexual IDUs was a common and efficient means of introducing HIV to low-prevalence areas. These studies suggest that the combination of high risk equipment sharing and sexual behaviours among IDUs is a potent means of HIV transmission. Although sexual transmission of HIV appears to be less important than transmission by sharing of injecting equipment in IDUs, sexual behaviour has been found to be more resistant to change than needle sharing (Donoghoe et al., 1989b; Hart et al., 1989). Further, a high prevalence of sexual HIV risk behaviours have been found in this population. Donoghoe et al. (1989b) reported that IDUs at a needle exchange did not use condoms for sexual activity up to 79% of the time: Jones & Vlahov (1989) reported similar figures. These studies raise the possibility that bisexual IDUs may have a special rBle in the dissemination of HIV transmission through multiple risk behaviours and multiple contacts. However, no research has been carried out on differences in HIV risk behaviour across sexual orientation in IDUs. The aim of this study was therefore to investigate differences between heterosexual, bisexual and homosexual IDUs in both sexual and injecting HIV risk behaviours. Method
This study formed part of a national HIV-IDU research project. Respondents were obtained by three forms of advertising. First, by interviewers distributing cards with study details and the telephone and address of the interview site, and the indication that IDUs would be paid $A 20 for an anonymous interview; second, by putting advertisements with the same message in employment and social security offices, needle exchanges and pharmacies which sold needles and syringes; and third by placing the same advertisement in a popular free central city magazine. Interviews took place in an unmarked building several blocks from the
Downloaded by [Memorial University of Newfoundland] at 20:55 20 January 2015
RISK BEHAVIOURS IN INJECTING DRUG USERS
141
centre of the drug-using sub-culture in the Kings Cross-Darlinghurst area of Sydney, with direct access off the street into a waiting room. These interviews were conducted in an individual private cubicle by interviewers who had extensive personal or professional experience in the area of injecting drug use. A single receptionist recorded initials of first and surnames and date of birth and where respondents were suspected or recognized as having attended previously, these data were checked to ensure that there was no double interviewing. In addition, interviews were conducted by one interviewer in the western suburbs of Sydney to obtain a broader geographical distribution of injecting drug users: these interviews were conducted at a community health centre under the same conditions. All data were entered on the interview schedule by the interviewer. At the conclusion of the interview, respondents were invited to contribute a drop of blood obtained by MicroletR from the tip of the central digit on to a strip of blotting paper, which was then coded, dried and sealed in a plastic bag for laboratory analysis. Because in an earler study in the same city (Wolk et al., 1990) self-reports of serostatus by IDUs were 100% concordant with their subsequent serological testing, those respondents who had already reported a positive HIV test were not retested, following the approach of McCusker et al. (1990). Payment took place at the completion of the interview preceding the blood collection. Interviews were conducted between May and December 1989. The study employed an interview schedule which had been piloted on over 100 injecting drug users and subsequently modified. Sections covered demographics, drug use behaviour, use of new equipment/reuse of own equipment, sharing of injection equipment, cleaning of injection equipment, disposal of used injection equipment, social context of injecting drug use, sexual history, knowledge and attitudes about HIVIAIDS, HIVIAIDS prevention behaviours, sources of HIVIAIDS information, HIVIAIDS antibody testing, and modules on treatment and prison use if appropriate. Response possibilities ranged from closed options to open-ended questions. All response possibilities were provided on showcards where appropriate (a copy of the full 36 page interview schedule is available from the first author on request). The interview took on average 1.25 hours to complete. Interview data were coded and punched for analysis by the SPSS-X package. Categorical data were analysed by xztest with Yates correction for discontinuity where appropriate, and by one way analysis of variance with post hoc Scheffk tests. Significance was set at the 5% level. HIV-1 estimations The dried bloods were sent to the National HIV Reference Laboratory (NRL) and stored according to the method of George et al. (1 989). The ELISA used throughout this study was the Genetic Systems LAV EIA (product number 0218). The kit insert was strictly adhered to for DBS elution and eluate test procedures. Eluates initially reactive by ELISA were repeated in duplicate: repeatedly reactive eluates were tested by Western Blot (WB).The WB method routinely used by the NRL was adapted to the Immunetics miniblotter system (MN45) (Maskill et al., 1989). The required modification to the published protocol was as follows. The uncut casein blocked nitrocellulose sheet was loaded into the miniblotter according to the manufacturer’s protocol. Individual lanes were cleared of residual buffer by vacuum prior to loading 100 1 of a 2:3 dilution of eluate with 1% casein. The manifold was then vigorously agitated on a platform shaker for 1.5 hours at room temperature. After this incubation a 300 ml volume of wash buffer was passed through the lanes of the miniblotter. The nitrocellulose sheet was removed from the miniblotter for subsequent incubations and washes. Quality control DBS samples used in this study were provided by the Centers for
142 M. W. ROSS ETAL.
Downloaded by [Memorial University of Newfoundland] at 20:55 20 January 2015
Disease Control, Atlanta, GA 30333, USA. These DBS represented negative, low positive and high positive serostatus respectively (identification numbers 01,02,03) and were eluted in duplicate with each elution batch. When members of an elution batch were tested by ELISA or WB, corresponding quality control samples were tested also. WB band patterns were categorized according to NRL HIV-1 WB interpretation guidelines (Maskill et al., 1989). HIV prevalence data were based on laboratory analysis of dried bloods obtained at interview for 14 positive results from 696 respondents (dried blood spot seroprevalence =2.0%), and for the remainder data were based on self-report of previous laboratory analysis (combined prevalence =6.8%; combined n =939). Results Sample characteristics
A total of 1,245 respondents who had injected drugs in the last 2 years were interviewed: these included 908 males, 331 females, and six male to female transsexuals. Characteristics of the sample are described in Table 1. Because some questions were not answered by all respondents, the ns will not always reflect the total numbers of respondents in each category. Table 1. Sample characteristics (Mean fSD or %) ~
Variable
n Demographics Age Born in Australia Some high school education Pension or benefit Pension or benefit for more than a year Employed (full- or part-time) Been in prison Years since release Number of children Number of dependent children Others financially dependent Moved to Sydney Years since moved to Sydney Lhug behaviour Been in previous drug treatment Currently in treatment If drug treatment, years since treatment ended Age first drug injection Age first injecting drug monthly or more often Frequency of injecting per month
~
~
Males
Females
908
331
27.9 2 6.7 79.2% 57.2% 59.2% 50.8% 13.5% 45.2% 3.6 f3.8 0.6f1.3 0.1f0.5 0.2 f2.6 55.4% 4.8 +. 7.7
26.3 f7.6 84.6% 53.5% 58.5% 57.6% 16.7% 19.6% 4.9 f 8.4 0.9 f2.7 0.6f2.1 0.1 20.4 54.1% 4.9 2 8.4
63.8% 30.4%
55.9% 15.7%
2.3? 4.6 18.6 f4.4
2.3? 5.5 18.1 k3.8
20.0 f4.6 49.0k 66.0
19.0f4.4 53.0 f66.0
Sexual orientation was defined by sexual partners in the past 5 years, with bisexuals reporting at least one sexual contact with males and females in this period. Unsafe sex was defined as reporting anal or vaginal sex without a condom in the past 6 months and having more than one partner in that period. This is a liberal definition and could exclude those
Downloaded by [Memorial University of Newfoundland] at 20:55 20 January 2015
RISK BEHAWOURS IN INJECTING DRUG USERS
143
cases where the HIV status of the partner is unknown or discordant and thus there is a risk of transmission. While it does accord with current advice to IDUs to practice safer sex or limit sex to one partner, it will underestimate risk by excluding oral sex and risks associated with the partner’s other (and possibly unknown) unsafe activities. Unsafe needle sharing practices were defined as having accepted a used needle or syringe from another person in the past 6 months. Analyses of variables by sexual orientation appear for males in Table 2 and for females in Table 3. The category ‘anal receptive sex, without a condom’ included insertion of objects such as sex toys or digits into the rectum, and thus heterosexual men report some anal receptive activity without a condom in Table 2. This inclusion of sex toys was noted during data analysis, and it was not possible to exclude such cases although the risk of HIV transmission using sex toys is probably significantly lower than that for unprotected anal intercourse. Use of injecting equipment by sexual orientation Variables not included in the tables were not significantly different across sexual orientation. There were no differences across sexual orientation by use, reuse or sharing of injecting equipment, or cleaning or disposal of equipment. The mean percentage of the time a used needle or syringe was used for males, heterosexual 18.1%, bisexual 16.1%, homosexual 21.1%, N.S.; for females, heterosexual 20.1%, bisexual 25%, homosexual 14%, N.S. The mean percentage of the time a needle or syringe was shared and the respondent used first was for males, heterosexual 52.7%, bisexual 45.1%, homosexual 41.5%, N.S.; for females, heterosexual 51.6%, bisexual 49.3%, homosexual 47.7%, N.S. The mean number of people respondents accepted a used needle and syringe from in the past 6 months was for males, heterosexual 1.9, bisexual 2.8, homosexual 0.9; and for females, heterosexual 1.1 ,bisexual 2.5, homosexual 0.9, both N.S. The number of different people (including the respondent) who would use the same needle and syringe before it was thrown away was for males, heterosexual 2, bisexual 1.8, homosexual 1.4, N.S.; for females, heterosexual 2, bisexual 2.5, homosexual 1.6, N.S. Mean percentage of the time respondents were ‘high, stoned or drunk’ during sex was 57.9 for males, 56.8 for females, with no significant differences across sexual orientation for either gender. The proportion of occasions that respondents reported being ‘high, stoned or drunk’ when having sex was for males, heterosexual 57.8, bisexual 59.1, homosexual 58.9 (F2=0.2, N.S.) and for females, lesbian 54.7, bisexual 66.0, and heterosexual 53.1 (F2=7.0, ptO.001).
Discussion The representativeness of samples of IDUs is problematic as the size and characteristics of the drug-using population in a community has never been defined confidently. However, this is a large sample, containing both in-treatment and out-of-treatment populations recruited systematically and is comparable on many indices with another large sample of IDUs recruited from multiple sources in a city 300 km from Sydney (Kieboom et al., 1989). Selfreport of sexual behaviours (McLaws et al., 1990) and drug use behaviours (Darke et al., 1991) is reported to be of acceptable reliability and validity, although this may be an additional source of error. However, such error is unlikely to be systematic. Because of the size of the data set and the number of variables, we report only on data relating to sexual orientation and have not included analyses of other subgroups. These data indicate that differences across sexual orientation in injecting drug users are almost exclusively limited to sexual behaviour, as there is little difference in type of drug
144 M. W. ROSS E T A . Table 2. Differences across sexual orientation in male injecting drug users (%) Heterosexual (n=719)
(n=117)
Homosexual (n=50)
18.7 26.7 47.5
23.1 42.1 35.7
22.0 38.0 35.4
Had HIV Test HIV positive Changed behaviour since test
76.2 3.2 45.9
88.9 12.1 60.4
94.0 35.4 59.6
Sexual behaviour Regular sex partners bisexual Pay for sex Paid for sex
9.1 13.1 4.0
46.4 1.8 33.9
18.8 6.1 32.7
1.5 1.5 6.2
5.1 6.8 8.6
14.0 16.0
63.3 8.8 18.5 7.2 5.4 56.1
53.8 16.2 24.8 14.5 44.4 70.1
42.0 10.0 36.0 14.0 22.0 64.0
90.4 27.4 17.7 39.6 16.3
84.8 44.7 57.1 64.5 64.5 42.9
0 0 44.0 71.9 72.4 46.7
0
62.5
85.0
**
0
53.5
78.4
**
76.0 0
80.7 66.7
0 85.0
** **
26.8
69.2
26.0
**
STDs Gonorrhoea Genital warts Non specific urethritis No STDs reported
11.0 9.5 11.3 35.0
19.7 17.1 19.7 33.3
30.0 24.0 26.0 18.0
Sex with HIV seropositive person After knew were HIV+ Took precautions Before knew
1.5 72.7 1.4
12.2 68.5 12.5
30.4 92.9 44.8
Variable
Demographics Tertiary educated Always lived in Sydney Been to prison
Bisexual
Downloaded by [Memorial University of Newfoundland] at 20:55 20 January 2015
HIV status
Oral drugs used Benzodiazapines Ecstacy LSD h g s used when have sex Heroin Other opiates Benzodiazapines LSD Ecstacy Alcohol Sexual practices past 6 months Vaginal Vaginal, withdrawal Anal receptive minus condom Anal insertive with condom Anal receptive with condom Anal insertive, withdrawal Oral receptive with male partner Oral receptive, male partner, withdrawal Oral Insertive, female partner Manual, male partner Practice unsafe sex, past 6 months
0
Significance
** ** ** ** **
2.0
** * ** ** ** ** **
** **
** ** ** ** ** **
Mean Number of HIV tests
3.0
3.8
3.7
Bi Ho>He**
RISKBEHAVIOURSIN INJECTING DRUG USERS 145 Table 2.-Cont. Heterosexual (n=719)
Downloaded by [Memorial University of Newfoundland] at 20:55 20 January 2015
Variable Number female partners, past year Number male partners, past year W time condoms used: Vaginal Anal insertive Anal receptive Oral receptive Oral insertive Knowledge of AIDS *pHo*
6.1
6.2
0
0
10.8
23.1
Ho>Bi>He**
30.9 22.0 0 0 10.2 12.1
41.7 45.6 58.4 31.5 21.5 12.5
0 70.7 75.6 22.4 19.2 13.5
Bi >He> Ho* Ho>Bi>He** Ho>Bi>He* Bi>Ho>He* Bi>He** Ho>He*
**pHo* Bi>He>Ho** Ho Bi>He** Bi>Ho He* Bi>Ho He* Bi>Ho He**
7.6
3.7
85.0
Ho>Bi
Sexual practices past 6 months Vaginal Anal insertive, without condom Fellating male partner Fellating male partner, withdraw Active oral sex with female partner Receiving cunnilingus Masturbating male Masturbating woman Being masturbated Practice unsafe sex, past 6 months STDs Genital warts Hepatitis B Pelvic Inflammatory Disease No STD Reported
** **
**
50.0
Mean Number HIV tests Male partners past year Female partners past year Number regular partners Number sexual partners IDUs % time intoxicated during sex % time condoms used, oral receptive sex *p
AIDS Care: Psychological and Sociomedical Aspects of AIDS/HIV Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/caic20
Differences across sexual orientation on HIV risk behaviours in injecting drug users a
b
c
M. W. Ross , A. Wodak , J. Gold & M. E. Miller
d
a
National Centre in HIV Social Research , University of New South Wales b
Alcohol and Drug Service , St Vincent's Hospital
c
Albion Street AIDS Centre, Sydney Hospital
d
Directorate of the Drug Offensive, New South Wales Department of Health , Australia Published online: 25 Sep 2007.
To cite this article: M. W. Ross , A. Wodak , J. Gold & M. E. Miller (1992) Differences across sexual orientation on HIV risk behaviours in injecting drug users, AIDS Care: Psychological and Socio-medical Aspects of AIDS/HIV, 4:2, 139-148, DOI: 10.1080/09540129208253085 To link to this article: http://dx.doi.org/10.1080/09540129208253085
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &
Downloaded by [Memorial University of Newfoundland] at 20:55 20 January 2015
Conditions of access and use can be found at http://www.tandfonline.com/page/termsand-conditions
AIDS CARE, VOL. 4, NO. 2, 1992
139
Differences across sexual orientation on HIV risk behaviours in injecting drug users
Downloaded by [Memorial University of Newfoundland] at 20:55 20 January 2015
M. W. ROSS,A. WODAK',J. GOLD^ &I M. E. MILLER3 National Centre in HIV Social Research, University of New South Wales, 'Alcohol and Drug Service, St Vincent's Hospital, *Albion Street AIDS Centre, Sydney Hospital, 3Directorate of the Drug Ofensive, New South Wales Department of Health, Australia
Abstract Injecting drug users (IDUs) play a disproportionate role in the spread of HIVgiven
their injecting and sexual contacts, and thereby act as conduits between these risk groups. We investigated diferences in risk behaviour and HIV seroprevalence in a Sydney sample of 1,245 IDUs. Significant diferences were observed across sexual orientation in HIV serostatus for males, with homosexual men having the highest HIV seroprevalence rate (35%), bisexual men intermediate (12%) and heterosexual men lowest (3%). Sexual HIV risk behaviours were lowest for homosexual men, intermediate for bisexual men, and highest for heterosexual men in the case of condom use: however, for numbers of partners, seroprevalence, and anal sex the trends were reversed. There were no diferences across sexual orientation fm either sex for injecting drug risk behaviours. Both male and female respondents reported having more than 50% of sexual contacts while under the influence of drugs. This study suggests that risk reduction in the sexual domain has not generalized to the injecting risk domain regardless of sexual orientation, and demonstrates that sexual risk behaviours in IDUs are lowest in homosexual, intermediate in bisexual, and highest in heterosexual IDU men. Introduction
In many western countries, the bisexual injecting drug user is positioned at the conjunction of a number of important bridges for HIV transmission. First, they are a potential conduit of HIV transmission from homosexual men into the wider drug using population; second, they are a potential conduit from injecting drug users (IDUs) to non drug-using heterosexual people (Marmor et al., 1990); third, they are independently of drug use a potential conduit between homosexual men and heterosexual men and women; and fourth, they are indirectly together with heterosexual IDUs, a conduit for vertical transmission. For these reasons, the bisexual IDU may play a disproportionately prominent rBle in transmission of HIV infection between and across groups of individuals whose behaviour places them at risk of HIV infection. These links are all the more important because a substantial proportion of male and female prostitutes also use drugs. Boles et al. (1989) found that 13% of a sample of male
Address for correspondence: Dr Michael Ross, National Centre in HIV Social Research, University of New South Wales, 345 Crown Street, Surry Hills NSW 2010, Australia.
Downloaded by [Memorial University of Newfoundland] at 20:55 20 January 2015
140 M. W.ROSS ET AL. prostitutes in the USA engaged in recreational sex with both males and females, and that 62% of the bisexuals (as compared with 37% of heterosexuals and 43% of homosexual hustlers) had injected drugs, particularly cocaine and heroin. Moreover, while male IDUs may sometimes provide homosexual sex commercially, they will more often have recreational heterosexual partners; conversely, a large proportion of female IDUs provide sexual services to men but some may also have homosexual contacts. Drug use and prostitution are commonly linked because prostitution is often used as a means of generating income to support a drug habit, or alternatively drugs are sometimes consumed to make the commercial provision of sexual services bearable emotionally. Williams (1990) provides clear evidence of the HIV transmission prospects from bisexual IDUs. He found that IDUs under 25 generally have partners who are not themselves IDUs, and that this appeared to be related to prostitution. In his sample of 131 IDUs recruited in the USA, Williams found that 15% reported sexual contact with both sexes in the past 6 months. Of those young adults who had had sex in that period, 38% had exchanged sex for money. The majority of male prostitutes reported that their sexual identity was strictly heterosexual and that they usually had relationships with young women who were frequently also IDUs and prostitutes. Williams suggests that contact between IDU and non-IDU populations is simple and direct: young female IDU prostitutes sharing injection equipment and having sex with clients at no other risk for HIV, as well as through bisexual clients of male prostitutes who are IDUs. Chu et al. (1989) found from an assessment of CDC data that bisexual men with AIDS in the USA were more than twice as likely to have injected drugs than homosexual men. This finding remained when the data were adjusted for race or ethnicity. Further, Battjes et al. (1989) noted that equipment sharing by male homosexual and bisexual IDUs with heterosexual IDUs was a common and efficient means of introducing HIV to low-prevalence areas. These studies suggest that the combination of high risk equipment sharing and sexual behaviours among IDUs is a potent means of HIV transmission. Although sexual transmission of HIV appears to be less important than transmission by sharing of injecting equipment in IDUs, sexual behaviour has been found to be more resistant to change than needle sharing (Donoghoe et al., 1989b; Hart et al., 1989). Further, a high prevalence of sexual HIV risk behaviours have been found in this population. Donoghoe et al. (1989b) reported that IDUs at a needle exchange did not use condoms for sexual activity up to 79% of the time: Jones & Vlahov (1989) reported similar figures. These studies raise the possibility that bisexual IDUs may have a special rBle in the dissemination of HIV transmission through multiple risk behaviours and multiple contacts. However, no research has been carried out on differences in HIV risk behaviour across sexual orientation in IDUs. The aim of this study was therefore to investigate differences between heterosexual, bisexual and homosexual IDUs in both sexual and injecting HIV risk behaviours. Method
This study formed part of a national HIV-IDU research project. Respondents were obtained by three forms of advertising. First, by interviewers distributing cards with study details and the telephone and address of the interview site, and the indication that IDUs would be paid $A 20 for an anonymous interview; second, by putting advertisements with the same message in employment and social security offices, needle exchanges and pharmacies which sold needles and syringes; and third by placing the same advertisement in a popular free central city magazine. Interviews took place in an unmarked building several blocks from the
Downloaded by [Memorial University of Newfoundland] at 20:55 20 January 2015
RISK BEHAVIOURS IN INJECTING DRUG USERS
141
centre of the drug-using sub-culture in the Kings Cross-Darlinghurst area of Sydney, with direct access off the street into a waiting room. These interviews were conducted in an individual private cubicle by interviewers who had extensive personal or professional experience in the area of injecting drug use. A single receptionist recorded initials of first and surnames and date of birth and where respondents were suspected or recognized as having attended previously, these data were checked to ensure that there was no double interviewing. In addition, interviews were conducted by one interviewer in the western suburbs of Sydney to obtain a broader geographical distribution of injecting drug users: these interviews were conducted at a community health centre under the same conditions. All data were entered on the interview schedule by the interviewer. At the conclusion of the interview, respondents were invited to contribute a drop of blood obtained by MicroletR from the tip of the central digit on to a strip of blotting paper, which was then coded, dried and sealed in a plastic bag for laboratory analysis. Because in an earler study in the same city (Wolk et al., 1990) self-reports of serostatus by IDUs were 100% concordant with their subsequent serological testing, those respondents who had already reported a positive HIV test were not retested, following the approach of McCusker et al. (1990). Payment took place at the completion of the interview preceding the blood collection. Interviews were conducted between May and December 1989. The study employed an interview schedule which had been piloted on over 100 injecting drug users and subsequently modified. Sections covered demographics, drug use behaviour, use of new equipment/reuse of own equipment, sharing of injection equipment, cleaning of injection equipment, disposal of used injection equipment, social context of injecting drug use, sexual history, knowledge and attitudes about HIVIAIDS, HIVIAIDS prevention behaviours, sources of HIVIAIDS information, HIVIAIDS antibody testing, and modules on treatment and prison use if appropriate. Response possibilities ranged from closed options to open-ended questions. All response possibilities were provided on showcards where appropriate (a copy of the full 36 page interview schedule is available from the first author on request). The interview took on average 1.25 hours to complete. Interview data were coded and punched for analysis by the SPSS-X package. Categorical data were analysed by xztest with Yates correction for discontinuity where appropriate, and by one way analysis of variance with post hoc Scheffk tests. Significance was set at the 5% level. HIV-1 estimations The dried bloods were sent to the National HIV Reference Laboratory (NRL) and stored according to the method of George et al. (1 989). The ELISA used throughout this study was the Genetic Systems LAV EIA (product number 0218). The kit insert was strictly adhered to for DBS elution and eluate test procedures. Eluates initially reactive by ELISA were repeated in duplicate: repeatedly reactive eluates were tested by Western Blot (WB).The WB method routinely used by the NRL was adapted to the Immunetics miniblotter system (MN45) (Maskill et al., 1989). The required modification to the published protocol was as follows. The uncut casein blocked nitrocellulose sheet was loaded into the miniblotter according to the manufacturer’s protocol. Individual lanes were cleared of residual buffer by vacuum prior to loading 100 1 of a 2:3 dilution of eluate with 1% casein. The manifold was then vigorously agitated on a platform shaker for 1.5 hours at room temperature. After this incubation a 300 ml volume of wash buffer was passed through the lanes of the miniblotter. The nitrocellulose sheet was removed from the miniblotter for subsequent incubations and washes. Quality control DBS samples used in this study were provided by the Centers for
142 M. W. ROSS ETAL.
Downloaded by [Memorial University of Newfoundland] at 20:55 20 January 2015
Disease Control, Atlanta, GA 30333, USA. These DBS represented negative, low positive and high positive serostatus respectively (identification numbers 01,02,03) and were eluted in duplicate with each elution batch. When members of an elution batch were tested by ELISA or WB, corresponding quality control samples were tested also. WB band patterns were categorized according to NRL HIV-1 WB interpretation guidelines (Maskill et al., 1989). HIV prevalence data were based on laboratory analysis of dried bloods obtained at interview for 14 positive results from 696 respondents (dried blood spot seroprevalence =2.0%), and for the remainder data were based on self-report of previous laboratory analysis (combined prevalence =6.8%; combined n =939). Results Sample characteristics
A total of 1,245 respondents who had injected drugs in the last 2 years were interviewed: these included 908 males, 331 females, and six male to female transsexuals. Characteristics of the sample are described in Table 1. Because some questions were not answered by all respondents, the ns will not always reflect the total numbers of respondents in each category. Table 1. Sample characteristics (Mean fSD or %) ~
Variable
n Demographics Age Born in Australia Some high school education Pension or benefit Pension or benefit for more than a year Employed (full- or part-time) Been in prison Years since release Number of children Number of dependent children Others financially dependent Moved to Sydney Years since moved to Sydney Lhug behaviour Been in previous drug treatment Currently in treatment If drug treatment, years since treatment ended Age first drug injection Age first injecting drug monthly or more often Frequency of injecting per month
~
~
Males
Females
908
331
27.9 2 6.7 79.2% 57.2% 59.2% 50.8% 13.5% 45.2% 3.6 f3.8 0.6f1.3 0.1f0.5 0.2 f2.6 55.4% 4.8 +. 7.7
26.3 f7.6 84.6% 53.5% 58.5% 57.6% 16.7% 19.6% 4.9 f 8.4 0.9 f2.7 0.6f2.1 0.1 20.4 54.1% 4.9 2 8.4
63.8% 30.4%
55.9% 15.7%
2.3? 4.6 18.6 f4.4
2.3? 5.5 18.1 k3.8
20.0 f4.6 49.0k 66.0
19.0f4.4 53.0 f66.0
Sexual orientation was defined by sexual partners in the past 5 years, with bisexuals reporting at least one sexual contact with males and females in this period. Unsafe sex was defined as reporting anal or vaginal sex without a condom in the past 6 months and having more than one partner in that period. This is a liberal definition and could exclude those
Downloaded by [Memorial University of Newfoundland] at 20:55 20 January 2015
RISK BEHAWOURS IN INJECTING DRUG USERS
143
cases where the HIV status of the partner is unknown or discordant and thus there is a risk of transmission. While it does accord with current advice to IDUs to practice safer sex or limit sex to one partner, it will underestimate risk by excluding oral sex and risks associated with the partner’s other (and possibly unknown) unsafe activities. Unsafe needle sharing practices were defined as having accepted a used needle or syringe from another person in the past 6 months. Analyses of variables by sexual orientation appear for males in Table 2 and for females in Table 3. The category ‘anal receptive sex, without a condom’ included insertion of objects such as sex toys or digits into the rectum, and thus heterosexual men report some anal receptive activity without a condom in Table 2. This inclusion of sex toys was noted during data analysis, and it was not possible to exclude such cases although the risk of HIV transmission using sex toys is probably significantly lower than that for unprotected anal intercourse. Use of injecting equipment by sexual orientation Variables not included in the tables were not significantly different across sexual orientation. There were no differences across sexual orientation by use, reuse or sharing of injecting equipment, or cleaning or disposal of equipment. The mean percentage of the time a used needle or syringe was used for males, heterosexual 18.1%, bisexual 16.1%, homosexual 21.1%, N.S.; for females, heterosexual 20.1%, bisexual 25%, homosexual 14%, N.S. The mean percentage of the time a needle or syringe was shared and the respondent used first was for males, heterosexual 52.7%, bisexual 45.1%, homosexual 41.5%, N.S.; for females, heterosexual 51.6%, bisexual 49.3%, homosexual 47.7%, N.S. The mean number of people respondents accepted a used needle and syringe from in the past 6 months was for males, heterosexual 1.9, bisexual 2.8, homosexual 0.9; and for females, heterosexual 1.1 ,bisexual 2.5, homosexual 0.9, both N.S. The number of different people (including the respondent) who would use the same needle and syringe before it was thrown away was for males, heterosexual 2, bisexual 1.8, homosexual 1.4, N.S.; for females, heterosexual 2, bisexual 2.5, homosexual 1.6, N.S. Mean percentage of the time respondents were ‘high, stoned or drunk’ during sex was 57.9 for males, 56.8 for females, with no significant differences across sexual orientation for either gender. The proportion of occasions that respondents reported being ‘high, stoned or drunk’ when having sex was for males, heterosexual 57.8, bisexual 59.1, homosexual 58.9 (F2=0.2, N.S.) and for females, lesbian 54.7, bisexual 66.0, and heterosexual 53.1 (F2=7.0, ptO.001).
Discussion The representativeness of samples of IDUs is problematic as the size and characteristics of the drug-using population in a community has never been defined confidently. However, this is a large sample, containing both in-treatment and out-of-treatment populations recruited systematically and is comparable on many indices with another large sample of IDUs recruited from multiple sources in a city 300 km from Sydney (Kieboom et al., 1989). Selfreport of sexual behaviours (McLaws et al., 1990) and drug use behaviours (Darke et al., 1991) is reported to be of acceptable reliability and validity, although this may be an additional source of error. However, such error is unlikely to be systematic. Because of the size of the data set and the number of variables, we report only on data relating to sexual orientation and have not included analyses of other subgroups. These data indicate that differences across sexual orientation in injecting drug users are almost exclusively limited to sexual behaviour, as there is little difference in type of drug
144 M. W. ROSS E T A . Table 2. Differences across sexual orientation in male injecting drug users (%) Heterosexual (n=719)
(n=117)
Homosexual (n=50)
18.7 26.7 47.5
23.1 42.1 35.7
22.0 38.0 35.4
Had HIV Test HIV positive Changed behaviour since test
76.2 3.2 45.9
88.9 12.1 60.4
94.0 35.4 59.6
Sexual behaviour Regular sex partners bisexual Pay for sex Paid for sex
9.1 13.1 4.0
46.4 1.8 33.9
18.8 6.1 32.7
1.5 1.5 6.2
5.1 6.8 8.6
14.0 16.0
63.3 8.8 18.5 7.2 5.4 56.1
53.8 16.2 24.8 14.5 44.4 70.1
42.0 10.0 36.0 14.0 22.0 64.0
90.4 27.4 17.7 39.6 16.3
84.8 44.7 57.1 64.5 64.5 42.9
0 0 44.0 71.9 72.4 46.7
0
62.5
85.0
**
0
53.5
78.4
**
76.0 0
80.7 66.7
0 85.0
** **
26.8
69.2
26.0
**
STDs Gonorrhoea Genital warts Non specific urethritis No STDs reported
11.0 9.5 11.3 35.0
19.7 17.1 19.7 33.3
30.0 24.0 26.0 18.0
Sex with HIV seropositive person After knew were HIV+ Took precautions Before knew
1.5 72.7 1.4
12.2 68.5 12.5
30.4 92.9 44.8
Variable
Demographics Tertiary educated Always lived in Sydney Been to prison
Bisexual
Downloaded by [Memorial University of Newfoundland] at 20:55 20 January 2015
HIV status
Oral drugs used Benzodiazapines Ecstacy LSD h g s used when have sex Heroin Other opiates Benzodiazapines LSD Ecstacy Alcohol Sexual practices past 6 months Vaginal Vaginal, withdrawal Anal receptive minus condom Anal insertive with condom Anal receptive with condom Anal insertive, withdrawal Oral receptive with male partner Oral receptive, male partner, withdrawal Oral Insertive, female partner Manual, male partner Practice unsafe sex, past 6 months
0
Significance
** ** ** ** **
2.0
** * ** ** ** ** **
** **
** ** ** ** ** **
Mean Number of HIV tests
3.0
3.8
3.7
Bi Ho>He**
RISKBEHAVIOURSIN INJECTING DRUG USERS 145 Table 2.-Cont. Heterosexual (n=719)
Downloaded by [Memorial University of Newfoundland] at 20:55 20 January 2015
Variable Number female partners, past year Number male partners, past year W time condoms used: Vaginal Anal insertive Anal receptive Oral receptive Oral insertive Knowledge of AIDS *pHo*
6.1
6.2
0
0
10.8
23.1
Ho>Bi>He**
30.9 22.0 0 0 10.2 12.1
41.7 45.6 58.4 31.5 21.5 12.5
0 70.7 75.6 22.4 19.2 13.5
Bi >He> Ho* Ho>Bi>He** Ho>Bi>He* Bi>Ho>He* Bi>He** Ho>He*
**pHo* Bi>He>Ho** Ho Bi>He** Bi>Ho He* Bi>Ho He* Bi>Ho He**
7.6
3.7
85.0
Ho>Bi
Sexual practices past 6 months Vaginal Anal insertive, without condom Fellating male partner Fellating male partner, withdraw Active oral sex with female partner Receiving cunnilingus Masturbating male Masturbating woman Being masturbated Practice unsafe sex, past 6 months STDs Genital warts Hepatitis B Pelvic Inflammatory Disease No STD Reported
** **
**
50.0
Mean Number HIV tests Male partners past year Female partners past year Number regular partners Number sexual partners IDUs % time intoxicated during sex % time condoms used, oral receptive sex *p
