LAUSR

OBJECTIVES The relationship between past drug use trajectory and long-term relapse risk after rehabilitation among heroin-dependent patients remain understudied. The primary objectives were to identify longitudinal heroin use patterns of heroin-dependent patients, to determine the associative factors with trajectories and to investigate the impact of trajectory groups on relapse after finishing compulsory rehabilitation programs. MATERIALS AND METHODS A total of 564 heroin-dependent patients were recruited from 4 compulsory rehabilitation facilities in Shanghai, China between 2007 and 2008. The baseline data was linked to participants' follow-up data on relapse from official records. Group-based trajectory model was used to identify distinctive drug use trajectory groups. The association between the identified group and heroin relapse risk was then analyzed to understand the role of past drug use trajectory on relapse. RESULTS Five trajectory groups were identified in this cohort: (1) Rapid Decrease (9.9%); (2) Persistent High (32.0%); (3) Slow Decrease (34.1%); (4) Gradual Increase (4.5%); (5) Persistent Low (19.5%). Gender, age, education, and impulsivity were found to be different between the five groups. During the 5 years after discharged from the compulsory program, 291 (59.0%) individuals relapsed. Multivariate logistic regression analysis showed that the persistent high group (OR: 2.77 [1.46-5.24]), slow decrease group (OR: 2.31 [1.32-4.06]) and gradual increase group (OR: 3.50 [1.18-10.39]) was positively associated with the heroin relapse risk when compared to the persistent low group. CONCLUSIONS Heroin use trajectories vary among heroin-dependent patients in China. The trajectories of heroin use before compulsory rehabilitation are associated with subsequent long-term relapse risk.