Respondent-driven sampling in a syringe exchange setting

Aims: Respondent-driven sampling is a research technique, new to the Scandinavian setting, used in hard-to-reach populations, such as subjects at risk for HIV, including drug users. This study aimed to evaluate the use of respondent-driven sampling originating from syringe exchange clients, as a method to identify 'hidden' drug users without treatment or social service contact. Methods: Nine heroin and amphetamine injectors were recruited as 'seeds' and instructed to recruit up to three heavy drug users in a chain-referral process. Recruited clients were interviewed about drug use, social conditions and contacts with treatment, syringe exchange and other authorities. In order to estimate whether the recruitment managed to evolve into groups of 'hidden' drug users, clients included beyond the fourth wave of chain-referral were compared with seeds (wave 0) and clients recruited in waves 1-3. Results: Five seeds were generative, and in total, 66 clients were assessed. Except for one of the 35 variables studied, clients in waves 4-9 did not differ from clients in waves 0-3, and were no less likely to have contacts with authorities. All clients except one were injectors and syringe exchangers, and clients recruited later in the chain-referral even tended to attend the syringe exchange more frequently. Conclusions: Respondent-driven sampling originating from syringe exchange clients may have difficulty reaching beyond the population of injectors and syringe exchangers. The technique, new to this geographical setting, has potential for future studies, but particular efforts may be necessary to study out-of-treatment heavy drug users unknown to the syringe exchange program.