Change in alcohol and other drug use during five years of continuous opioid substitution treatment.

Background: English national prospective, observational cohort study of patients continuously enrolled for five years in opioid substitution treatment (OST) with oral methadone and sublingual buprenorphine. This is a secondary outcome analysis of change in use of alcohol and other drug use (AOD) following identification of heroin use trajectories during OST.

Methods: All adults admitted to community OST in 2008/09 and enrolled to 2013/14 (n = 7717). Data from 11 sequential, six-monthly clinical reviews were used to identify heroin and AOD use trajectories by multi-level Latent Class Growth Analysis. OST outcome in the sixth and seventh year was 'successful completion and no re-presentation' (SCNR) to structured treatment and was assessed using multi-level logistic regression.

Results: With 'rapid decreasing' heroin use trajectory as referent, 'continued high-level' heroin use predicted 'continued high-level' crack cocaine use (relative risk ratio [RRR] 58.7; 95% confidence interval [CI] 34.2-100.5),'continued high-level' alcohol use (RRR 1.2; 95% CI 1.0-1.5), 'increasing' unspecified drug use (RRR 1.7; 95% CI 1.4-2.1) and less 'high and increasing' cannabis use (RRR 0.5; 95% CI 0.4-0.6). 'Increasing' crack use was negatively associated with SCNR outcome for the 'decreasing then increasing' and 'gradual decreasing' heroin use groups (adjusted odds ratio [AOR] 0.5; 95% CI 0.3-0.9 and AOR 0.2; 95% CI 0.1-0.7, respectively).

Conclusions: Continued high-level heroin use non-response during long-term OST is associated with high-level crack cocaine and alcohol use, increasing unspecified drug use, but less high and increasing cannabis use. Increasing use of crack cocaine is negatively associated with the likelihood that long-term OST is completed successfully.