Close

Impact of recovery interventions on opioid users. A simulation study

Abstract

Aims: (1) To characterize life trajectories for opioid users tran-sitioning between 12 mutually exclusive states that capture opioid recovery process. The states are characterized by a combination ofuse status, location (being in community, jail, special residence),being in treatment, being under criminal justice supervision; (2)to project the impact of continuum of services and communitysupport interventions on life trajectories of opioid users.Methods: Developed a microsimulation model that considers acohort of opioid users being in treatment at baseline. Individualsmove between 12 mutually exclusive states. The transition proba-bilities were estimated using an innovative Bayesian approach thatcombined published peer-reviewed literature with the estimatesfrom the GAIN data. The GAIN sample contained 979 unique opi-oid users providing at least two consecutive responses during theirbaseline, 3-, 6- and 12-month assessments. Transition depended onage, sex, number of convictions and treatment episodes. Parame-ter values corresponding to the intervention effects were obtainedfrom peer-reviewed literature.Results: The analysis of simulated trajectories showed that theresults of each: the continuity of services and community supportintervention were moderate. In a long-term (5-year) simulationmost important cohort statistics (percent incarcerated, percentnon-using, percent using in the community, etc.) have improvedby 10%. An extreme hypothetical case of a powerful continuum ofservices intervention which reached the odds ratios of 10 and 5 hasresulted in the increase of 50% in percent in recovery.Conclusions: Recovery-focused interventions should considermultiple states and state transitions in the users’ life trajectories.Multiple interventions (such as continuum of services, and com-munity support) are needed to achieve substantial reduction in useand the increase the percent of users in recovery. Validation analy-sis shows a strong heterogeneity in life trajectories across differentpopulations.

MIDAS Network Members

Citation: