Ethnic- and gender-specific differences in the prevalence of HIV among patients in opioid maintenance treatment-a case register analysis.

Background: We have sought to identify ethnic- and gender-specific differences in HIV prevalence among heroin users receiving opioid maintenance treatment in the canton of Zurich, Switzerland.

Methods: We used a generalized linear model (GEE) to analyze data from the anonymized case register for all opioid maintenance treatments in the canton of Zurich. Patients who received either methadone or buprenorphine between 1991 and 2012 (n=11,422) were evaluated for gender (male vs. female), ethnic background (Swiss vs. non-Swiss), and lifetime method of drug use (ever injector vs. non-injector). We addressed missing data by multiple imputation.

Results: The overall prevalence of HIV among patients declined substantially from 33.7% in 1991 to 10.6% in 2012 in the complete dataset. In the imputed datasets, the respective prevalence dropped from 32.8% in 1991 to 9.7% in 2012. Non-injectors had a four to five times lower risk ratio (RR) compared to the reference group, 'Swiss males who ever injected'. In addition, we found a significantly higher risk ratio of HIV prevalence among females who had ever injected; this was true both for the complete dataset and the imputed dataset (Swiss RR 1.18 CI 95% 1.04-1.34, non-Swiss RR 1.58 CI 95% 1.18-2.12).

Conclusion: In this population, gender, ethnic background, and lifetime method of drug use influenced the risk of being HIV positive. Different access to treatment and different characteristics of risk exposure among certain subgroups might explain these findings. In particular, the higher risk for women who inject drugs-especially for those with an immigrant background-warrants additional research. Further exploration should identify what factors deter women from using available HIV-prevention measures and whether and how these measures can be better adapted to high-risk groups.