Aims: To investigate whether opiate substitution therapy (OST) and needle and syringe programmes (NSP) can reduce hepatitis C virus (HCV) transmission among injecting drug users (IDUs).

Design: Meta-analysis and pooled analysis, with logistic regression allowing adjustment for gender, injecting duration, crack injecting and homelessness.

Setting: Six UK sites (Birmingham, Bristol, Glasgow, Leeds, London and Wales), community recruitment.

Participants: A total of 2986 IDUs surveyed during 2001-09.

Measurement: Questionnaire responses were used to define intervention categories for OST (on OST or not) and high NSP coverage (100% or higher versus fewer than 100% needles per injection).  The primary outcome was new HCV infection, measured as antibody seroconversion at follow-up or HCV antibody-negative/RNA-positive result in cross-sectional surveys.

Findings  Preliminary meta-analysis showed little evidence of heterogeneity between the studies on the effects of OST (I2 = 48%, P = 0.09) and NSP (I2 = 0%, P = 0.75), allowing data pooling.  The analysis of both interventions included 919 subjects with 40 new HCV infections.  Both receiving OST and high NSP coverage were associated with a reduction in new HCV infection (adjusted odds ratios (AORs) = 0.41; 95% CI, 0.21 to 0.82 and 0.48; 95% CI, 0.25 to 0.93, respectively).  Full harm reduction (on OST plus high NSP coverage) reduced the odds of new HCV infection by nearly 80% (AOR = 0.21; 95% CI, 0.08 to 0.52).  Full harm reduction was associated with a reduction in self-reported needle sharing by 48% (AOR 0.52; 95% CI, 0.32 to 0.83) and mean injecting frequency by 20.8 injections per month (95% CI, -27.3 to -14.4).

Conclusions: There is good evidence that uptake of opiate substitution therapy and high coverage of needle and syringe programmes can substantially reduce the risk of hepatitis C virus transmission among injecting drug users.  Research is now required on whether the scaling-up of intervention exposure can reduce and limit hepatitis C virus prevalence in this population.

See also comment on this paper by C Day and L Topp, p.1989-1990.