HIV risk behaviour correlates among injecting drug users in Greek prisons

Aims. To identify the correlates of injecting drug use within prison. Design. A national cross-sectional study, participation being voluntary and anonymous. Setting. Ten Greek prisons. Participants. A representative sample of 1000 male inmates; 861 questionnaires were completed and analysed. Measurement. A self-report questionnaire for demographics, penal history, drug use and sharing injecting equipment. Findings. Two hundred and ninety inmates (33.7%) reported injecting drugs at some time in their lives, of whom 174 (60%) had injected while imprisoned. Among those who had injected while imprisoned, 145 (83%) had shared equipment while incarcerated. Logistic regression analysis suggested that total time in prison, previous drug conviction, being a convict (as opposed to on remand) and having multiple female sexual partners 1 year before incarceration were significant HIV risk behaviour correlates. For every year of imprisonment, the risk of injection in prison increased by about 17% [OR = 1.17 (95% CI: 1.07-1.27)]. Inmates with a previous drug-related conviction were about twice as likely to inject within prison [OR = 1.97 (95% CI: 1.16-3.33)]. Finally, convicted inmates were marginally significantly more prone to inject in prison [OR = 1.58 (95% CI: 0.92-2.74)]. Conclusions. Variables related to the inmates' prison career influence HIV risk behaviours within prison. There is a need to assist IDUs in reducing the likelihood of high-risk behaviour by considering factors such as frequency of incarceration, length of time incarcerated and availability of detoxification programmes within prison.