Presentation at the XI International AIDS Conference in July, 1996, in Vancouver, BC.

Social determinants predict needle sharing behaviour among injection drug users in Vancouver

Steffanie A Strathdee, PhD; David M Patrick, MD; Chris P Archibald MD MHSc; Marianna Ofner, MHSc; Kevin JP Craib, MMath; Peter GA Cornelisse, MSc; Greg Eades; Martin T Schechter, OBC, MD PhD; Michael V O'Shaughnessy, PhD.

I would like to acknowledge my co-authors listed above, as well as the research team, community advisory board and participants of the Point Project. We also acknowledge the Laboratory Centre for Disease Control, HIV/AIDS Epidemiology Division, Health Canada, for study funding.

Since January 1994, Vancouver has witnessed an increase in the proportion of IDUs testing newly HIV-positive, from 2% to 7%. This was cause for concern, since needle exchange and outreach services were introduced here as early as 1989.

A case-control investigation took place in 1995, whose aim was to identify risk factors for HIV infection. Results from this component of the study were presented by my colleagues, Dr. David Patrick and Dr. Chris Archibald, at the XI International Conference on AIDS.

The objective of the present analysis was to identify determinants of needle sharing behaviour among IDUs participating in this case-control study.

Eligibility criteria was as follows:

• HIV-positive (HIV+) subjects were selected as cases who had a documented seroconversion in 1994/1995; 89 such subjects were recruited.
• HIV-negative subjects had two documented negative tests within the same period; 192 subjects meeting this requirement participated.
• All subjects were required to have injected drugs in the period between their two HIV tests, defined here as the inter-test interval, and were living in the Greater Vancouver region at the time of the interview.

Questionnaire Data:

Subjects meeting the eligibility criteria who agreed to participate in an interviewer-adminstered questionnaire provided informed consent and were paid $20 for their time. Interviews focused on the inter-test interval, and included questions on:

• demographics (including migration, housing);
• injection and non-injection drug use;
• borrowing and lending of used needles;
• source of needles, including needle exchange attendance and barriers to clean needle use;
• sexual activity with males and females; and
• commercial sex work.

Non-consensual sex was assessed by asking respondents whether they had "ever been forced to have sex" (defined as oral, anal or vaginal sex) and, if so, whether this occurred before age 12, between 12-17 yrs or over 18 yrs.

To assess mental health, we asked about suicidal ideation and suicide attempts prior to the date of their second qualifying HIV test. Depression was assessed using an abbreviated 7 item version of the Centre for Epdemiologic Studies depression scale.

Statistical Analyses:

Contingency table analysis was first used, followed by unconditional logistic regression models to identify determinants of borrowing. The outcome variable was "ever borrowing a used needle" within the inter-test interval, which served as the period of recall.

The following describes demographic characteristics of the study sample. The majority of participants were male, and were white or of Aboriginal descent. Contingency table analysis indicated that borrowers were significantly more likely to be common-law or married, to report homosexual activity, and to have tested HIV-positive at the end of the recall period.

With regards to drug-using behaviours, it was not surprising that borrowers were significantly more likely to report injecting more than 4 times a day, to inject more than one drug (for example, heroin or cocaine singly or in combination as a speedball), and also were more likely to report attending shooting galleries, since these behaviours are known correlates of HIV seroconversion.

However, borrowers were not significantly more likely to report difficulties accessing sterile needles, or to report having had needles confiscated, for example, by police, even though this was frequently reported.

Of greater interest were differences between borrowers and non-borrowers with respect to indicators of psychologic distress. Borrowers were significantly more likely to have an elevated CES-D depression score, and to report having attempted suicide prior to their second qualifying HIV test.

Borrowers were significantly more likely to ever report having been forced to have oral, vaginal or anal sex (with a p-value of less than 0.001). Borrowers were also more likely than non-borrowers to report this experience occurring at all age groups. For at least 53% of all respondents, this experience occurred before first injecting a drug, but it was not possible to determine the temporal sequence for the remainder of subjects.

Logistic regression was used to identify determinants of borrowing used needles, for males and females separately and combined. The following depicts the final multivariate model for males only. After controlling for HIV status and injection drug using behaviours, NCS was associated with a 2.5-fold increased odds of borrowing. Male IDUS who reported homosexual activity within the inter-test interval was also idependently associated with borrowing, which underscores the need for prevention programs aimed at male IDUs who are homosexual/bisexual. However, an elevated depression score was not significantly associated with needle borrowing among males.

The following depicts the final multivariate model for females. Confidence intervals are wider due to smaller sample sizes. Nevertheless, as was the case for males, after controlling for HIV serostatus and injection drug use behaviours, ever experiencing non-consensual sex was a strong, independent factor associated with borrowing used needles.

In addition, female IDUs living common-law or married were 4 times more likely to use a borrowed needle. While we did not specifically ask respondents whether or not their partner was a drug user, women were more likely than men to report borrowing a needle from a regular sexual partner.

An elevated depression score was also associated with borrowing among females; however, since this variable was correlated with some of the other variables in the model, it is described separately.

The final multivariate model describing independent determinants of borrowing used needles is shown here, for males and females combined, since tests for homogeneity indicated that it was valid to collapse across gender. As seen here, ever experiencing non-consensual sex remained associated with a 3.4-fold increased odds of borrowing. Homosexual activity among males, and living common-law or married, were also independently associated with borrowing, as was an elevated CES-D score.

Limitations:

Our study was limited by the fact that it was a case-control study, and subjects were required to have had two HIV tests within a specified period. As such, our results may not be generalizable to those not seeking testing.

While this study design cannot infer a causal association, the fact that sexual abuse was independently associated with needle sharing suggests that IDUs with a history of sexual abuse are in urgent need of supports, and targeted HIV prevention efforts.

Yet our questionnaire collected only minimal data on the experience of sexual abuse. It lacked detailed information on drug treatment, and the drug using behaviours of sexual partners or persons in their social network.

Conclusions:

In conclusion, we found that factors suggestive of underlying psychologic distress, rather than access to sterile needles or barriers of clean needle use, were signifcantly associated with use of borrowed needles.

Of particular importance was the finding that sexual abuse was an independent predictor of borrowing among both male and female IDUs.

We also found that the context of current sexual relationships appears to influence decisions related to needle sharing, since homosexual/bisexual males and females living with sexual partner were significantly more likely to share needles.

Our results provide support for continued availability of sterile injection equipment, and agree with others who have suggested that AIDS prevention efforts among IDUs require a broader perspective.

Early identification of factors such as sexual abuse and depression coupled with targetted counselling and supports, should be an integral component of prevention strategies aimed at IDUs in an effort to reduce needle-sharing behaviors.

In an effort to examine these associations further, we have recently launched a longitudinal study, the Vancouver Injection Drug Use Study (VIDUS), to address trends in HIV incidence and risk factors in the IDU population. Thank you.

For more information, contact:

Bonnie Devlin
Vanguard Project Coordinator
608 - 1081 Burrard Street
Vancouver, BC, Canada, V6Z 1Y6
Tel: (604)806-8306
Fax: (604)806-9044