Predictors associated with injecting risk behaviour among people who inject drugs in Stockholm county prisons over ten years

Niklas Karlsson1, 2, Michele Santacatterina3, Torsten Berglund2, Sussi Wallin1, Kerstin Käll4, Anna Mia Ekström1, 5
Affiliates: 1Department of Public Health Sciences, Karolinska Institutet, Stockholm, Sweden, 2Public Health Agency of Sweden, Solna, Sweden, 3Unit of Biostatistics, Institute of Environmental Medicine, Karolinska Institutet, Stockholm, Sweden, 4Dependency Clinic, Linköping University Hospital, Linköping, Sweden, 5Department of Infectious Diseases, Karolinska University Hospital, Stockholm, Sweden

Background
People who inject drugs (PWID) are particularly vulnerable to hepatitis and HIV infection due to a range of well-known risk behaviours. Less is know about what factors that cause or influence such risk behaviours.

Method
We used a large prospectively collected surveillance data base derived from the Remand Prison Study in Stockholm during 2002-2012. All new detainees had equal opportunity to be recruited. Respondents were invited to be tested for HIV and hepatitis and/or interviewed by trained nurses who screened all new clients for injecting drug use and associated risk behaviours. The study population was defined as PWID at their first appearance in the study, including 2 151 individuals. Three risk behavioural outcomes were defined within a 12 month recall period: sharing injection drug solution, lending already used injection equipment, and receiving already used injection equipment.

Results
84% of the respondents were men. At age 16, 79% had already started using drugs (cannabis). At age 19, 53% had started to inject drugs, 72% of these used amphetamine and 25% were heroin users. The majority, 66% (N=981) had shared injection drug solution, 56% had lent used injection equipment (N=1 030) and 62% (N=1143) had received used injection equipment. Multivariate analysis found that men were significantly less likely to share injection drug solution than women (adjusted odds ratio (OR) 0.62; 95% confidence intervals (CI) 0.39-0.97). Those who were older when they started using drugs (age 25 or older) were less likely to receive used injection equipment (OR: 0.36; 95% CI 0.14-0.92) compared to younger debutantes. Similarly, those starting to inject at age 30 or later were at much lower risk of sharing injection drug solution (OR: 0.48; 95% CI 0.27-0.86) or having lent out used injection equipment (OR: 0.55; 95% CI 0.32- 0.93) than those starting before age 17. Heroin users were less likely to share injection drug solution compared to those who preferred amphetamine (OR: 0.4; 95% CI 0.21-0.76). Those who took a hepatitis test at the visit (optional) were more likely to have lent out used injection equipment compared to those who did not take a test (OR: 2.11; 95% CI 1.16-3.84). Having a housing contract and living with someone appeared to protect both against sharing drug solution (OR: 0.58; 95% CI 0.38-0.90 and OR: 0.66; 95% CI 0.46-0.95) respectively), lending used equipment to others (OR: 0.53; 95% CI 0.36-0.80 among contact owners) and receiving injection equipment (OR: 0.36; 95% CI 0.24-0.55 and OR: 0.6; 95% CI 0.43-0.84, respectively) compared to homeless PWID. People born in Sweden were less likely than PWID from Europe or other parts of the world to have lent out used injection equipment (OR: 0.7; 95% CI 0.49-0.99 and OR: 0.49; 95% CI 0.30-0.80, respectively).

Conclusion
Our study indicates that gender, age at drug and injection drug debute, type of drug preffered, testing for hepatitis, housing and country of birth play an important roles in predicting the variation in injection risk behaviours among PWID.