Prevalence and factors related to syringe sharing behaviours among female injecting drug users who are also sex workers in China
Article Outline
- Abstract
- Introduction
- Methods
- Results
- Background characteristics
- HIV-related risk behaviours
- Cognitive variables
- Variables on psychological distress
- Associations between background variables and syringe sharing in the last month
- Associations between cognitive/psychological variables and syringe sharing
- Nested hierarchical models predicting risky drug use behaviours in the last month
- Associations between condom use and injecting drug use
- Discussion
- Role of funding sources
- Acknowledgements
- References
- Copyright
Abstract
Background
Female injecting drug users who are sex workers (IDUFSWs) are at high risk of contracting HIV. They may bridge HIV transmissions from injecting drug users to clients of female sex workers.
Methods
A total of 216 non-institutionalised IDUFSWs were recruited by snowball sampling methods. Anonymous face-to-face interviews were conducted to collect data. Univariate, multivariate and hierarchical logistic regression models were fitted to investigate the associations between background characteristics, cognitive variables, psychological stress and syringe sharing behaviours among IDUFSWs.
Results
Respectively 33.8% and 27.8% of the respondents injected drugs with others’ used syringes and gave used syringes to others for drug injection in the last month. These two syringe sharing behaviours were significantly associated with inconsistent condom use during commercial sex (OR
=5.00 and 1.92, p<0.05). Over 90% of the respondents reported at least one type(s) of psychological distress included in this study. Adjusting for significant background variables, all variables that are related to the Theory of Planned Behaviour (attitude, norm, perceived control and behavioural intention) and psychological distress (except for depression) were significantly associated with injecting drugs with others’ used syringes (adjusted OR=2.08–6.25, p<0.05), whilst variables related to perceived control, behavioural intention and insomnia were significantly associated with providing used syringes to others for injection (adjusted OR=2.00–3.56, p<0.05). In two separate summary multivariate models, variables related to the Theory of Planned Behaviours and psychological distress were independently associated with injecting drugs with others’ used syringes (OR=1.98–4.02, p<0.05) and giving used syringes to others for injection (OR=2.06–3.59, p<0.05).
Conclusions
Syringe sharing behaviours were prevalent among IDUFSWs and were associated with cognitive and psychological factors. Effective integrative intervention programmes targeting IDUFSWs are warranted.
Keywords: Injecting drug users, Female sex workers, Syringe sharing, Psychological factors, China
Introduction
As of December 2007, there were 700,000 estimated HIV cases in China; 41% and 38% of them were attributable to heterosexual transmissions and injecting drug use (State Council AIDS Working Committee Office, China & UN Theme Group on HIV/AIDS in China, 2007). Female injecting drug users (IDUs) who are sex workers (IDUFSWs) bridge HIV transmissions (Gu et al., 2008, Strathdee et al., 2008) from the IDU population with higher HIV prevalence (17.8–71.9% in China) to the clients of female sex workers (FSWs) with lower HIV prevalence (0–0.5% in China) (Lau et al., 2007b, Yin et al., 2007, Zhang et al., 2002, Zhao et al., 2005). A previous study showed that over half of the female IDUs are involved in sex work (Yang et al., 2005). IDUFSWs have a major influence on the HIV epidemic in China and in many other countries (Monitoring the AIDS Pandemic, 2004, Strathdee et al., 2008, UNAIDS and WHO, 2006).
IDUFSWs are highly marginalised and are perceived to be at the bottom of the social stratification (Alantes et al., 2002, Weeks et al., 1998). IDUFSWs, as compared to IDUs who are not FSWs, are more likely to be younger, having lower education attainment, higher economic pressure (Azim et al., 2006, Paone et al., 1999), longer duration of drug use, higher frequency of drug use and higher prevalence of syringe sharing (Lau et al., 2008, Spittal et al., 2003). IDUFSWs also tended to have a higher number of sex partners, lower prevalence of condom use during commercial sex, more sex-work clients per day and higher prevalence of HIV/STD infection (Astemborski et al., 1994, Kral et al., 1998, Paone et al., 1999, Platt et al., 2005, Strathdee et al., 2008, Tran et al., 2005) than IDUs who are not FSWs. The two groups of IDUs (FSWs and non-FSWs) also differ in psychosocial status such as feelings of hopelessness and depression (Edwards et al., 2006, Paone et al., 1999).
Among IDUFSWs, factors in association with high-risk sexual behaviours include economic pressure and severity of drug dependence (Gossop et al., 1995, Gu et al., 2008). Drug dependence may compromise the ability of IDUFSWs to negotiate and to practice safer commercial sex behaviours (Gu et al., 2008, Sherman et al., 2006). Cognitive theories, such as the Theory of Planned Behaviours (TPB), have been used to explain condom use behaviours among FSWs who do not inject drugs, as well as among IDUFSWs (Couture et al., 2010, Gu et al., 2009a). It is however, uncertain whether such factors would also be applicable in understanding syringe sharing behaviours in our IDUFSW study population.
Among female IDUs who are not FSWs, background factors such as education level, frequency of drug use (Lau et al., 2008), having a male sex partner who is an IDU (Fitzgerald, Lundgren, & Chassler, 2007), personality traits and peers’ sharing behaviours (Brook, Brook, Richter, Masci, & Roberto, 2000) are significantly associated with syringe sharing behaviours. Cognitive factors have been used to explain syringe sharing among such female IDUs (Brown, 1998). For instance, low perceived efficacy on avoidance of syringe sharing for preventing HIV transmission was associated with syringe sharing behaviours among female IDUs (Lau et al., 2008). Some psychosocial factors were also associated with syringe sharing behaviours among female IDUs who are not FSWs (Brook et al., 2000, Metzger et al., 1991).
Despite the afore-mentioned knowledge about factors in association syringe sharing among IDU who are not FSWs, little is known about similar associations in the IDUFSW population. According to our literature review, there is no study which investigates factors (neither cognitive nor psychological) associated with syringe sharing among IDUFSWs. The present study hence may be the first of its kind, investigating the prevalence and factors related to syringe sharing behaviours among non-institutionalised IDUFSWs in Dazhou of Sichuan, China. It was hypothesised that variables related to socio-demographic background, history of drug use and sex work, perceived efficacy of preventive measures, cognitions derived from the TPB and psychological status would be independently associated with syringe sharing among IDUFSWs. The findings contribute to an in-depth understanding toward IDUFSWs and their risk behaviours.
Methods
Sampling and data collection
Sichuan Province is located in southwestern China and ranks 6th in HIV prevalence among all of the Chinese provinces (State Council AIDS Working Committee Office, China & UN Theme Group on HIV/AIDS in China, 2007). In Sichuan, the prevalence of HIV among IDUs and FSWs were 7.2% and 0.6% respectively (personal communication with doctor of Sichuan Provincial Centre for Disease Prevention and Control (CDC)). Whilst Dazhou, a city in Sichuan, has a population of 6.5 millions and over 5000 registered drug users. The HIV prevalence among IDUs and FSWs in Dazhou around the time of the survey was 33.2% and 2.6% respectively (Dazhou CDC, 2005). During August 2005 through April, 2006, 216 non-institutionalised female IDUs, who were of age 18 years old or above, and who mainly used heroin for injection and had engaged in sex work (trade sex for money or drugs) in the past 6 months, were recruited using snowball sampling method. Some female peer IDU educators of the local needle exchange programme served as seeds of the snowball sampling process. They were briefed about the study and referred the prospective participants to the research team. Further referrals were made by the study participants to join this study.
Face-to-face interviews were administered by well-trained staff members of the local CDC in places where privacy was ensured. Screening questions were asked to confirm respondents’ eligibility. As drug use and sex work are both illegal in China, respondents were not required to give written consent. Instead, the interviewers signed a form pledging that they had explained the details of the research to the respondents and had obtained their verbal informed consent. Such informed consent process has been used in similar studies (Gu et al., 2009b). Respondents were given about 6 US$ as a compensation for their time spent in the interview (about 20–30min). Ethics approval was obtained from the Chinese University of Hong Kong.
Measures
Background information was recorded, including: (1) socio-demographic data (age, education, ethnicity, marital status and whether being local resident of Dazhou), (2) self-reported HIV status, (3) whether you had ever utilised HIV-related services (check-up for sexually transmitted diseases (STD), free HIV testing and counselling, needle exchange programme, and methadone maintenance treatment programme), (4) variables that are related to sex work (duration of sex work, number of clients in the last week and income per sexual transaction), and (5) variables that are related to drug use (duration and frequency of drug injection and whether having a regular sex partner who was an IDU).
Cognitive variables included perceived efficacy and cognitions derived from the TPB. Two measures on perceived efficacy for preventing HIV transmission were assessed: ‘whether condom use could prevent HIV transmission?’ and ‘whether avoiding sharing syringes with others could prevent HIV transmission?’ The constructs of the TPB include attitude, subjective norm, perceived behavioural control and behavioural intentions, which are determinants of health-related behaviours (Ajzen, 1991). Respondents were asked whether they agreed with the following statements: ‘there is a low risk of HIV infection in case of sharing syringes for drug use with close friends’ (attitude), ‘your peer IDUs usually share syringes with others when injecting drugs’ (norm), ‘when craving, sometimes you cannot get new syringes for injection’ (perceived control), ‘when craving, sometimes you cannot insist on using new syringes but to get drug injection as fast as you can’ (perceived control), and ‘it is possible for you to share syringes with others in the future 6 months’ (behavioural intention).
The level of psychological distress was assessed by asking whether the respondents have the following feelings in the last week: ‘whether you hate yourself very much’, ‘whether you feel very depressed’ and ‘whether you are suffering from severe insomnia’. These items were selected as a result of some pilot interviews with some IDUFSWs in the study site.
Syringe sharing data included ‘whether you had utilised others’ used syringes for drug injection in the last month’, and ‘whether you had given used syringes to others for drug injection in the last month’. The relationship between the respondent and the person who provided the syringes for sharing and reasons for using others’ used syringes were recorded. Respondents were asked whether they used condoms consistently (every-time) during commercial sex in the last month and whether they had used a condom during the last episode of commercial sex.
Statistical methods
Univariate logistic regression models were fitted to identify background variables (socio-demographic variables, drug use and sex work-related variables) that were associated with the two dependent variables on syringe sharing. Background variables with p<0.05 (univariate analyses) were used as candidates for fitting multivariate stepwise logistic regression models (Table 3). To further investigate the associations between cognitive variables and psychological status and the dependent variables (syringe sharing behaviours), univariate odds ratios (OR) were derived. Multivariate logistic regression models were then fitted, using cognitive and psychological variables as independent variables and adjusting for the afore-mentioned multivariately significant background variables (Table 4).
Additional summary logistic regression models were fitted for the two syringe sharing dependent variables, using the nested hierarchical modeling approach. Background variables with p<0.05 in the univariate analysis (Table 3), cognitive and psychological variables with p<0.05 in the adjusted analysis (Table 4) were used as candidates for each step. Model 1 contained the significant background variables obtained from the stepwise model. In Model 2, relevant cognitive variables were stepwisely added to Model 1 to build up Model 2, and psychological variables were then stepwisely added to Model 2 to build up Model 3 (Table 5). In order to assess statistical significance for adding variables to the previous block, the −2LL statistics was used (Cohen and Cohen, 1983). SPSS for Windows version 14.0 was used for data analysis and p<0.05 was considered as statistically significant.
Results
Background characteristics
Almost all participants were of Han ethnicity (99.1%) and were local residents (90%) (Table 1). The participants were on average 28.1 years old; 17.6% received less than 6 years of formal education; 19.9% were currently married; 71.2% had ever used at least one type(s) of HIV-related services, and 7.9% self-reported to be HIV positive (69.9% were uncertain about their HIV status).
Table 1. Background characteristics and HIV-related risk behaviours of all participants.
| n % | |
|---|---|
| Background characteristics | |
| Socio-demographic information | |
| % Han ethnicity | 99.1 |
| Age groups | |
| <25 | 29.1 |
| 25–29 | 26.9 |
| ≥30 | 44.0 |
| Education level | |
| ≤Primary | 17.6 |
| Junior high | 63.9 |
| ≥Senior high | 18.5 |
| Marital status | |
| Single | 53.7 |
| Married | 19.9 |
| Other (divorced, widow) | 26.4 |
| Local residency | 94.0 |
| HIV-related information | |
| % Ever used HIV-related services | |
| STD check-up | 33.3 |
| Free HIV testing and counselling | 56.0 |
| Needle exchange programme | 38.3 |
| Methadone maintenance treatment programme | 11.6 |
| Self-reported HIV infection status | |
| Negative | 22.2 |
| Unknown | 69.9 |
| Positive | 7.9 |
| Sex work-related variables | |
| Duration of sex work ≥3 years | 50.5 |
| Had >7 clients in the last 7 days | 58.9 |
| Income <50 per transaction | 22.8 |
| Drug use-related variables | |
| Duration of drug injection ≥3 years | 80.6 |
| Frequency of drug injection ≥3 per day | 73.6 |
| Having a regular sex partner who was an IDU | 22.4 |
| HIV-related risk behaviours | |
| Condom use-related behaviours | |
| Consistent condom use with clients in the last month | 16.7 |
| Condom use with clients in the last episode | 67.0 |
| Syringe sharing in the last month | |
| Used others’ used syringes | 33.8 |
| Gave used syringes to others for injection | 27.8 |
| Either of above | 44.3 |
About half (50.5%) of the respondents had involved in commercial sex for ≥3 years; 58.9% had had >7 clients in the last week; 22.4% were currently having a male regular sex partner who was an IDU; 22.8% reported that they charged <50 RMB (about 7 US$) per sexual transaction. As for drug use, most (80.6%) of the respondents had been injecting drugs for ≥3 years and 73.6% were injecting ≥3 times per day (Table 1).
HIV-related risk behaviours
Only 16.7% of respondents reported consistent condom use with clients during commercial sex in the last month; 67.0% used a condom in the last episode of commercial sex (Table 1). About one-third (33.8%) used others’ used syringes for drug injection and 27.8% gave used syringes to others for drug injection in the last month, with 44.3% practicing either one of these two types of syringe sharing behaviours.
Among those who had injected with others’ used syringes in the last month, 56.9%, 13.9% and 40.3% obtained the syringe from their boyfriends, husbands, and friends. Frequently mentioned reasons for injecting with others’ used syringes included: ‘no new syringe was available’ (59.7%), ‘was in urgent need and did not think much about it’ (40.3%), ‘would like to share syringes with others’ (29.2%), ‘it is more convenient to use others’ syringes’ (27.8%), ‘people providing you with used syringes are not diseased’ (26.4%), and ‘used syringes were provided by acquaintance’ (20.8%). Of those who shared syringes, 44.4% had multiple sharers (data not tabulated).
Cognitive variables
Of all respondents, 85.1% gave affirmative answers to two items indicating perceived efficacy of preventive measures. With regard to variables derived from the TPB, 25% of the respondents agreed that a low risk of HIV transmission was involved in case of syringe sharing with close friends (attitude) (Table 2); 45.3% believed that most of their peers would share syringes with others (norm); 45.8% indicated that sometimes they could not get new syringes during craving, and 71.3% could not insist on using new syringes when craving (perceived control). Overall, 36.7% believed that they would share syringes with others in the future 6 months (behavioural intention).
Table 2. Distribution of cognitive variables and psychological variables.
| n % | |
|---|---|
| Perceived efficacy | |
| % Gave positive answers to two items indicating perceived efficacy of preventive measuresa | 85.1 |
| TPB-related variables | |
| There is low risk of HIV infection if sharing with close friends (attitude) | 25.0 |
| Your IDU peers share syringes when injecting (norm) | 45.3 |
| When craving, sometimes you cannot get new syringes (perceived control) | 45.8 |
| When craving, sometimes you cannot insist on using new syringes but to get drug injection as fast as you can (perceived control) | 71.3 |
| Possible to share syringes with others in 6 months (behavioural intention) | 36.7 |
| Psychological distress | |
| You hate yourself very much | 71.8 |
| You feel very depressed | 77.3 |
| You are suffering from severe insomnia | 65.7 |
aThe two efficacy-related items were: “whether condom use in sex can prevent HIV transmission?” and “whether avoidance of needle sharing in drug use can prevent HIV transmission?”. |
Variables on psychological distress
The prevalence of the three types of psychological distress (hated oneself very much, felt very depressed and had severe insomnia) was respectively 71.8%, 77.3% and 65.7%. The majority of the respondents (91.2%) reported ≥1 types of psychological distress problems (Table 2).
Associations between background variables and syringe sharing in the last month
In the univariate analyses, respondents self-reporting being HIV positive (OR=12.35, p<0.001), having been an IDU for ≥3 years (OR=2.53, p<0.05), injecting drugs ≥3 per day (OR=2.05, p<0.05) and having a regular sex partner being an IDU (OR=2.58, p<0.01), were more likely than others to have injected drugs with others’ used syringes in the last month (Table 3). The reversed was true for having ever used needle exchange programme (OR=0.48, p<0.05). In the multivariate stepwise analyses, respondents self-reported being HIV positive (OR=10.07, p<0.001) and having a regular sex partner being an IDU (OR=2.23, p<0.05) were more likely than others to have had injected drugs with others’ used syringes in the last month.
Table 3. Associations between background variables and syringe sharing behaviours in the last month.
| Injecting drugs with others’ used syringes | Giving used syringes to others for drug injection | |||||||
|---|---|---|---|---|---|---|---|---|
| Row % | ORu | ORm | (95% C.I.) | Row % | ORu | ORm | (95% C.I.) | |
| Ever used needle exchange programme | ||||||||
| No | 27.3 | 1.00 | N.S. | 30.5 | 1.00 | – | ||
| Yes | 43.9 | 0.48* | 25.8 | 0.79 | ||||
| HIV-related information | ||||||||
| Self-reported HIV infection status | ||||||||
| Negative | 20.8 | 1.00 | 1.00 | 25.5 | 1.00 | – | ||
| Unknown | 33.1 | 1.88 | 1.76 | (0.80,3.87) | 26.4 | 1.04 | ||
| Positive | 76.5 | 12.35*** | 10.07 | (2.64,38.5)*** | 47.1 | 2.59 | ||
| Had >7 clients in the last 7 days | ||||||||
| No | 37.5 | 1.00 | – | 34.5 | 1.00 | N.S. | ||
| Yes | 31.7 | 0.78 | 23.6 | 0.53** | ||||
| Drug use-related variables | ||||||||
| Duration of drug injection ≥3 years | ||||||||
| No | 19.0 | 1.00 | N.S. | 26.8 | 1.00 | – | ||
| Yes | 37.4 | 2.53* | 28.1 | 1.06 | ||||
| Frequency of drug injection ≥3 per day | ||||||||
| No | 22.8 | 1.00 | N.S. | 24.6 | 1.00 | – | ||
| Yes | 37.7 | 2.05* | 29.0 | 1.26 | ||||
| Having a regular sex partner who was an IDU | ||||||||
| No | 28.8 | 1.00 | 1.00 | 23.0 | 1.00 | 1.00 | ||
| Yes | 51.1 | 2.58** | 2.23 | (1.11,4.48)* | 43.5 | 2.58** | 2.54 | (1.27,5.05)** |
*p |
**p |
***p |
Similarly, in the univariate analysis, two background variables were associated with giving used syringes to others for drug injection in the last months (having >7 clients in the last week and having a regular sex partner who was an IDU). Only the variable whether having a regular sex partner being an IDU (OR=2.54, p<0.01) remained statistically significant in the stepwise logistic regression model (Table 3).
Associations between cognitive/psychological variables and syringe sharing
In the univariate analyses, except for perceived efficacy and the variable ‘feeling very depressed’, all variables that were related to cognitions and psychological distress were significantly associated with injecting drugs with others’ used syringes (OR=2.06–8.57, p<0.05) (Table 4). All these associations remained statistically significant (adjusted OR=2.08–6.25, p<0.05), after adjusting for the significant background variables that were listed in Table 3.
Table 4. Associations between cognitive/psychological variables and syringe sharing in the last month.
| Injecting drugs with others’ used syringes | Giving used syringes to others for drug injection | |||||||
|---|---|---|---|---|---|---|---|---|
| Row % | ORu | ORa | (95% C.I.) | Row % | ORu | ORa | (95% C.I.) | |
| TPB-related variable | ||||||||
| There is a low risk of HIV infection in case of sharing syringes with close friends (attitude) | ||||||||
| No | 29.0 | 1.00 | 1.00 | 26.9 | 1.00 | 1.00 | ||
| Yes | 48.1 | 2.27* | 2.87 | (1.44,5.72)** | 30.8 | 1.21 | 1.39 | (0.69,2.82) |
| Your IDU peers usually share syringes when injecting drugs (norm) | ||||||||
| No | 23.1 | 1.00 | 1.00 | 21.1 | 1.00 | 1.00 | ||
| Yes | 46.4 | 2.88*** | 2.16 | (1.15,4.08)* | 35.4 | 2.06* | 1.74 | (0.91,3.32) |
| When craving, sometimes you cannot get new syringes (perceived control) | ||||||||
| No | 13.7 | 1.00 | 1.00 | 15.7 | 1.00 | 1.00 | ||
| Yes | 57.6 | 8.57*** | 6.25 | (3.14,12.43)*** | 42.3 | 3.95*** | 3.56 | (1.82,6.94)*** |
| When craving, sometimes you cannot insist on using new syringes but to get drug injection as fast as you can (perceived control) | ||||||||
| No | 19.4 | 1.00 | 1.00 | 21.3 | 1.00 | 1.00 | ||
| Yes | 39.6 | 2.73** | 2.39 | (1.13,5.05)* | 30.5 | 1.62 | 1.48 | (0.72,3.04) |
| It is possible for you to share syringes with others in the next 6 months (behavioural intention) | ||||||||
| No | 22.1 | 1.00 | 1.00 | 22.6 | 1.00 | 1.00 | ||
| Yes | 53.2 | 4.01*** | 3.64 | (1.91,6.94)*** | 37.2 | 2.03* | 2.00 | (1.06,3.76)* |
| Psychological distress | ||||||||
| You hate yourself very much | ||||||||
| No | 23.0 | 1.00 | 1.00 | 20.3 | 1.00 | 1.00 | ||
| Yes | 38.1 | 2.06* | 2.36 | (1.13,4.92)* | 30.7 | 1.74 | 1.61 | (0.77,3.36) |
| You are suffering from severe insomnia | ||||||||
| No | 23.0 | 1.00 | 1.00 | 15.5 | 1.00 | 1.00 | ||
| Yes | 39.4 | 2.18* | 2.08 | (1.04,4.14)* | 34.0 | 2.82** | 2.95 | (1.16,6.19)* |
*p |
**p |
***p |
The results of the multivariate analysis, after adjusting for the significant background variables, showed that those who believed one could not get new syringes during craving (perceived control) (adjusted OR=3.56, p<0.05), those who intended to share syringes with others in the future 6 months (behavioural intention) (adjusted OR=2.00, p<0.05), and those who had severe insomnia (adjusted OR=2.95, p<0.05) were more likely than others to have given used syringes to others for drug injection in the last month (Table 4).
Nested hierarchical models predicting risky drug use behaviours in the last month
In the first summary model in Table 5, addition of relevant cognitive variables to background variables was statistically significant (Model 2 versus Model 1, −2LL=−34.9, p<0.001); addition of a psychological variable to Model 2 (background plus cognitive variables) was also statistically significant (Model 3 versus Model 2, −2LL=−5.4, p<0.05). The final model (Model 3) contained two marginally significant variables (HIV status and having a regular sex partner being an IDU), three significant TPB-related variables that were related to attitude (OR=2.32, p<0.05), perceived control (OR=2.20, p<0.05) and behavioural intention (OR=4.02, p<0.001), as well as one psychological variable (hatred toward oneself) (OR=1.98, p<0.05). These variables were independently associated with using others’ syringes for drug injection.
Table 5. Nested hierarchical models predicting syringe sharing behaviours in the last month.
| Model 1 | Model 2 | Model 3 | ||||
|---|---|---|---|---|---|---|
| ORm | (95% C.I.) | ORm | (95% C.I.) | ORm | (95% C.I.) | |
| Dependent variable=injecting drugs with others’ used syringes | ||||||
| Block I. Background characteristics | ||||||
| Self-reported HIV infection status | ||||||
| Negative | 1.00 | 1.00 | 1.00 | |||
| Unknown | 1.87 | (0.82,4.25) | 1.49 | (0.60,3.72) | 1.49 | (0.60,3.72) |
| Positive | 10.61 | (2.70,41.6)*** | 4.02 | (0.93,17.50) | 4.02 | (0.93,17.50)+ |
| Having a regular sex partner who was an IDU (yes vs. no) | 2.38 | (1.17,4.82)* | 2.13 | (0.96,4.72) | 2.13 | (0.96,4.72)+ |
| Block II. TPB-related variables (yes vs. no) | N.A. | |||||
| There is a low risk of HIV infection in case of sharing syringes with close friends (attitude) | 2.64 | (1.20,5.81)* | 2.32 | (1.09,5.81)* | ||
| Your IDU peers usually share syringes when injecting drugs (norm) | N.S. | |||||
| When craving, sometimes you cannot get new syringes (perceived control) | 5.11 | (2.36,11.05)*** | 4.02 | (2.16,10.05)*** | ||
| When craving, sometimes you cannot insist on using new syringes but to get drug injection as fast as you can (perceived control) | N.S. | |||||
| It is possible for you to share syringes with others in the next 6 months (behavioural intention) | 2.26 | (1.09,4.67)* | 2.20 | (1.08,4.97)* | ||
| Block III. Psychological distress (yes vs. no) | N.A. | N.A. | ||||
| You hate yourself very much | 1.98 | (1.05,4.52)* | ||||
| −2LL(d.f.) | 234.7 | (3) | 199.8 | (6) | 194.4 | (7) |
| Δ−2LL(d.f.) comparison with the previous model | N.A. | −34.9 | (3)*** | −5.4 | (1)* | |
| Dependent variable=Giving used syringes to others for drug injection | ||||||
| Block I. Background Characteristics | ||||||
| Having a regular sex partner who was an IDU (yes vs. no) | 2.42 | (1.20,4.86)* | 1.84 | (0.88,3.84) | 1.57 | (0.73,3.34) |
| Block II. TPB-related variables (yes vs. no) | N.A. | |||||
| When craving, sometimes you cannot get new syringes (perceived control) | 3.72 | (1.87,7.37)*** | 3.59 | (1.80,7.19)*** | ||
| Possible to share syringes with others in 6 months (behavioural intention) | N.S. | |||||
| Block III. Psychosocial distress | N.A. | N.A. | ||||
| You are suffering from severe insomnia (yes vs. no) | 2.06 | (1.02,5.13)* | ||||
| −2LL(d.f.) | 229.3 | (1) | 214.2 | (2) | 209.2 | (3) |
| Δ−2LL(d.f.) comparison with the previous model | N.A. | −15.1 | (1)*** | −5.0 | (1)* | |
+p |
*p |
***p |
Similarly in the second summary model, addition of a cognitive variable (perceived control) to the background variable was statistically significant (Model 2 versus Model 1, −2LL=−15.1, p<0.001), and addition of a psychological variable (severe insomnia) to Model 2 (background plus cognitive variables) was statistically significant (Model 3 versus Model 2, −2LL=−5.0, p<0.05). The final model (Model 3) contained two significant variables (perceived control and severe insomnia, OR=3.59 and 2.06, p<0.05), which were independently associated with giving used syringes to others for drug injection.
Associations between condom use and injecting drug use
Adjusted for significant background variables, inconsistent condom use during commercial sex in the last month was associated with injecting drugs with other's used syringe in the last month (adjusted OR=5.00, 95% C.I.=1.54–16.67, p<0.01) and condom non-use during the last episode of commercial sex was associated with giving used syringes to others for injection in the last month (adjusted OR=1.92, 95% C.I.=1.01–3.70, p<0.05) (data were not tabulated).
Discussion
This is one of the few studies focusing on IDUFSWs’ syringe sharing behaviours. High prevalence of syringe sharing (44.3%) and unprotected commercial sex (83.3%) in the last month was reported. Different types of HIV-related risk behaviours, such as risk behaviours related to drug use and sex intercourse tend to cluster with each other. For instance, IDUs who are syringe sharers, as compared to non-sharers, are more likely to have unprotected sex with different types of sex partners (Lau et al., 2007a). Similar clustering effects were observed in this study – unprotected commercial sex was associated with syringe risk behaviours among IDUFSWs. There is hence a group of IDUFSWs who are at very high risk of HIV transmission both due to injecting drug use and commercial sex behaviours. Taking into account the high HIV prevalence (33.2%) among IDUs in Dazhou (Dazhou Center for Disease Control & Prevention, 2005), IDUFSWs’ “bridging effect” of spreading HIV from the IDU population to the male sex partners of the IDUFSW therefore raises a serious concern.
IDUFSWs’ high prevalence of using others’ used syringes (34%) and giving used syringes to others for injecting (28%) was higher than that of Lau's study which was conducted in female IDUs who were not FSWs in Sichuan (21% and 22%) (Lau et al., 2008). The findings of high prevalence of syringe sharing despite high prevalence of HIV services utilisation (71.2%), and the non-significant associations between some services utilisation and syringe sharing in this study suggested that improvements for HIV prevention programmes targeting IDUFSWs are warranted.
Consistent with the results that were obtained from other studies (Fitzgerald et al., 2007, Lau et al., 2009), having a regular sex partner who is an IDU was multivariately associated with the two syringe sharing variables. Convenience and a false sense of security would increase the chance of syringe sharing among IDU couples (Lau et al., 2009) as addressed in our study, more than half of the syringe sharers shared with their boyfriend or husband, and about 30% believed that people who provided used syringes were not diseased. Gender differential between IDU partners could be another reason. Previous research, such as Platt's study in 2005, suggests that among FSWs who inject drugs with intimate partners, a gendered process whereby they may have reduced control over drug procurement and preparation, and our data may further support this (Platt et al., 2005). Relevant interventions targeting IDUFSWs therefore need to be couple-based. Such programmes virtually do not exist in China (Lau et al., 2009).
As expected, those who self-reported themselves to be HIV positive were more likely than others to use others’ syringes for injection. It is possible that the reason for self-protection was removed. It is unclear whether all these self-reported HIV positive cases are confirmed, nor has it been clear whether the syringes that were used by those self-reporting HIV positive had been used by a third HIV negative IDU. The risk for HIV transmissions under these circumstances can still be perpetuated.
This is the first study of its kind that applies the TPB to investigate syringe sharing behaviours among IDUFSWs in China, though TPB has been applied to investigate condom use during commercial sex among another sample of IDUFSWs (Gu et al., 2009a). In this study, all the five TPB-related variables were associated with drug injection with others’ used syringes in the last month, after adjusting for significant background factors. Theory-grounded interventions are more effective than non-theory-grounded programmes in reducing HIV-related risk behaviours (Exner, Seal, & Ehrhardt, 1997). TPB should be used more often to design risk reduction programme for IDUs – increase in perceived control over syringe sharing may be especially relevant as it remained significant throughout the study.
Psychological problems were prevalent among IDUFSWs. A significant proportion of FSWs and IDUs were suffering from psychological problems such as hopelessness or depression (Hong et al., 2007, Paone et al., 1999). One's self-regulatory processes may be disrupted by negative affects (Leith & Baumeister, 1996). People experiencing negative affects may develop mal-adaptive thoughts (Beck, 1967), which compromise their motivation to take care of themselves. Negative affects, such as hatred toward oneself, would deprive one's motivation to adopt a long-term perspective on avoiding the negative consequences of risk behaviours (Cohen, 2005). IDUFSWs, as shown by our results, often hate themselves, have severe insomnia problems or feel very depressed, and these variables were associated with syringe sharing. The results implied that interventions targeting syringe sharing among IDUFSWs need to take into account psychological status of the target population.
The results of the hierarchical models suggested that background factors, cognitive factors and psychological factors are relevant and may affect syringe sharing behaviours independently. Determinants of HIV-related behaviours are multi-dimensional and risk reduction programmes need to be integrative. Our findings suggested a need for increased distribution of syringes, particularly to those who are in an intimate IDU partnership. Furthermore, screening of psychological programmes and relevant counselling are warranted among IDUFSWs to improve their psychological well being, as well as to increase their protective behaviours. However, psychological services and well-trained counsellors do not seem to be readily accessible to IDUFSWs in China.
This study had a few limitations. Snowball sampling method was used in this study because random sampling was not feasible in the absence of a sampling frame of IDUFSWs. Respondent driven sampling (RDS) methods were used in some previous studies targeting IDU but not those targeting IDUFSWs, who may not have large networks among themselves, which are required by the RDS methods. The associations reported by our cross-sectional data do not necessarily suggest causal relationships. Since the data were self-reported, reporting bias might exist. However, almost all HIV-related risk behaviour studies were based on self-reported data; efforts were paid in this study to maintain anonymity and privacy during the research process. Due to the length of the questionnaires, indicator variables instead of scales were derived for the TPB. Some validated psychological measures exist and could have been used in this study. They were however not used, again, due to the length of the questionnaire. The items used in this study were selected as a result of our pilot qualitative interviews with some IDUFSWs in the study site.
The results demonstrated that syringe sharing behaviours were prevalent among IDUFSWs, who were exposed to a high risk of HIV transmission. Integrative and tailor-made HIV prevention programmes targeting IDUFSWs are greatly warranted, in order to slow down the bridging of HIV transmissions from the IDU population to non-IDU populations. IDUFSWs in different countries may face similar issues, among other problems, therefore comparative studies are hence greatly warranted.
Role of funding sources
This work received no external funding.
Acknowledgements
The authors would like to thank all respondents of the study and staff of Dazhou CDC, Sichuan.
There is no competing interest of all authors.
References
- . Drug-using sex workers in the streets of valencia: Risk in context. Journal of Ethnicity in Substance Abuse. 2002;4:1–27
- . The theory of planned behavior. Organizational Behavior and Human Decision Processes. 1991;50:179–211
- . The trading of sex for drugs or money and HIV seropositivity among female intravenous drug users. American Journal of Public Health. 1994;84:382–387
- Vulnerability to HIV infection among sex worker and non-sex worker female injecting drug users in Dhaka Bangladesh: Evidence from the baseline survey of a cohort study. Harm Reduction Journal. 2006;3:33
- . Depression: Clinical, experimental, and theoretical aspects. New York: Hoeber; 1967;
- . Needle sharing: A longitudinal study of female injection drug users. American Journal of Drug and Alcohol Abuse. 2000;26:263–281
- . Female injecting drug users: Human immunodeficiency virus risk behavior and intervention needs. Journal of Professional Nursing. 1998;14:361–369
- Cohen D. Poverty & HIV/AIDS in Sub-Saharan Africa (22 April 2005). UNDP HIV and Development Programme Paper No. 27. Available from: http://www.undp.org/hiv/publications/issues/english/issue27e.html. Accessed 12.04.09.
- . Applied multiple regression/correlation analysis for the behavioral sciences. Hillsdale, N.J: Lawrence Erlbaum Associates; 1983;
- . Determinants of intention to use condoms among clients of female sex workers in Haiti. AIDS Care. 2010;22:253–262
- . Dazhou HIV/AIDS surveillance report. Dazhou: Dazhou Center for Disease Control Prevention; 2005;
- . Correlates of exchanging sex for drugs or money among women who use crack cocaine. AIDS Education & Prevention. 2006;18:420–429
- . A review of HIV interventions for at-risk women. AIDS and Behavior. 1997;1:93–124
- . Factors associated with HIV/AIDS high-risk behaviours among female injection drug users. AIDS Care. 2007;19:67–74
- . Female prostitutes in south London: Use of heroin, cocaine and alcohol, and their relationship to health risk behaviours. AIDS Care. 1995;7:253–260
- Severity of drug dependence, economic pressure and HIV-related risk behaviors among non-institutionalized female injecting drug users who are also sex workers in China. Drug & Alcohol Dependence. 2008;97:257–267
- Using the theory of planned behavior to investigate condom use behaviors among female injecting drug users who are also sex workers in China. AIDS Care. 2009;21:967–975
- Prevalence of needle sharing, commercial sex behaviors and associated factors in Chinese male and female injecting drug user populations. AIDS Care. 2009;21:31–41
- . Depressive symptoms and condom use with clients among female sex workers in China. Sexual Health. 2007;4:99–104
- . HIV seroprevalence among street-recruited injection drug and crack cocaine users in 16 US municipalities. American Journal of Public Health. 1998;88:108–113
- Clustering of syringe sharing and unprotected sex risk behaviors in male injecting drug users in China. Sexually Transmitted Diseases. 2007;34:574–582
- Comparing prevalence of HIV-related behaviors among female injecting drug users (IDU) whose regular sexual partner was or was not IDU in Sichuan and Yunnan Provinces, China. AIDS Care. 2009;21:909–917
- Comparing HIV-related syringe sharing behaviors among female IDU engaging versus not engaging in commercial sex. Drug & Alcohol Dependence. 2008;97:54–63
- Comparing prevalence of condom use among 15379 female sex workers injecting or not injecting drugs in China. Sexually Transmitted Diseases. 2007;34:908–916
- . Why do bad moods increase self-defeating behavior? Emotion, risk taking, and self-regulation. Journal of Personality & Social Psychology. 1996;71:1250–1267
- . Risk factors for needle sharing among methadone-treated patients. American Journal of Psychiatry. 1991;148:636–640
- . AIDS in Asia: Face the fact. A comprehensive analysis of AIDS epidemic in Asia. Monitoring the AIDS Pandemic (MAP); 2004;
- . HIV risk behaviours of current sex workers attending syringe exchange: The experiences of women in five US cities. AIDS Care. 1999;11:269–280
- Impact of gender and sex work on sexual and injecting risk behaviors and their association with HIV positivity among injecting drug users in an HIV epidemic in Togliatti City Russian Federation. Sexually Transmitted Diseases. 2005;32:605–612
- . The evaluation of the JEWEL project: An innovative economic enhancement and HIV prevention intervention study targeting drug using women involved in prostitution. AIDS Care. 2006;18:1–11
- Surviving the sex trade: A comparison of HIV risk behaviours among street-involved women in two Canadian cities who inject drugs. AIDS Care. 2003;15:187–195
- . A Joint Assessment of HIV/AIDS Prevention, Treatment and Care in China. Beijing; 2007;
- Correlates of injection drug use among female sex workers in two Mexico–U.S. border cities. Drug & Alcohol Dependence. 2008;92:132–140
- . Drug use among female sex workers in Hanoi Vietnam. Addiction. 2005;100:619–625
- . AIDS epidemic update. Geneva: UNAIDS and WHO; 2006;
- . Streets, drugs, and the economy of sex in the age of AIDS. Women & Health. 1998;27:205–229
- Heterosexual transmission of HIV in China: A systematic review of behavioral studies in the past two decades. Sexually Transmitted Diseases. 2005;32:270–280
- Continued spread of HIV among injecting drug users in southern Sichuan Province China. Harm Reduction Journal. 2007;4:6
- High prevalence of HIV-1 and hepatitis C virus coinfection among injection drug users in the southeastern region of Yunnan, China. Journal of Acquired Immune Deficiency Syndromes. 2002;29:191–196
- Sexually transmitted disease/HIV and heterosexual risk among miners in townships of Yunnan Province China. AIDS Patient Care & STDS. 2005;19:848–852
PII: S0955-3959(10)00100-3
doi:10.1016/j.drugpo.2010.06.005
© 2010 Published by Elsevier Inc.
