The acceptability and feasibility of a brief psychosocial intervention to reduce blood-borne virus risk behaviours among people who inject drugs: a randomised control feasibility trial of a psychosocial intervention (the PROTECT study) versus treatment as usual

Background: While opiate substitution therapy and injecting equipment provision (IEP) have reduced blood-borne viruses (BBV) among people who inject drugs (PWID), some PWID continue to share injecting equipment and acquire BBV. Psychosocial interventions that address risk behaviours could reduce BBV transmission among PWID.

Methods: A pragmatic, two-armed randomised controlled, open feasibility study of PWID attending drug treatment or IEP in four UK regions. Ninety-nine PWID were randomly allocated to receive a three-session manualised psychosocial group intervention and BBV transmission information booklet plus treatment as usual (TAU) (n = 52) or information booklet plus TAU (n = 47). The intervention was developed from evidence-based literature, qualitative interviews with PWID, key stakeholder consultations, and expert opinion. Recruitment rates, retention in treatment, follow-up completion rates and health economic data completion measured feasibility.

Results: Fifty-six percent (99/176) of eligible PWID were recruited. More participants attended at least one intervention session in London (10/16; 63%) and North Wales (7/13; 54%) than in Glasgow (3/12; 25%) and York (0/11). Participants who attended no sessions (n = 32) compared to those attending at least one (n = 20) session were more likely to be homeless (56 vs 25%, p = 0.044), injected drugs for a greater number of days (median 25 vs 6.5, p = 0.019) and used a greater number of needles from an IEP in the last month (median 31 vs 20, p = 0.056). No adverse events were reported. 45.5% (45/99) were followed up 1 month post-intervention. Feedback forms confirmed that the intervention was acceptable to both intervention facilitators and participants who attended it. Follow-up attendance was associated with fewer days of injecting in the last month (median 14 vs 27, p = 0.030) and fewer injections of cocaine (13 vs 30%, p = 0.063). Analysis of the questionnaires identified several service use questionnaire categories that could be excluded from the assessment battery in a full-randomised controlled trial.

Conclusions: Findings should be interpreted with caution due to small sample sizes. A future definitive RCT of the psychosocial intervention is not feasible. The complex needs of some PWID may have limited their engagement in the intervention. More flexible delivery methods may have greater reach.

Trial registration: ISRCTN66453696