Extending Long-term Outcomes Through an Adaptive Aftercare Intervention

Reinforcement interventions have pronounced effects on reducing cocaine use. We developed and evaluated a low-cost reinforcement intervention, systematically moving it through the Stages of development to dissemination and broad clinical implementation. In an ongoing project, reinforcement interventions are yielding benefits when reinforcers are provided at treatment initiation and for longer durations. However, less than half of patients remain engaged for 12 weeks with traditional reinforcement interventions, which require frequent attendance for monitoring and reinforcing abstinence. Interventions that extend into aftercare and that are acceptable to and efficacious in preventing long-term relapse are critically needed.

Reinforcement interventions are efficacious during periods they are in effect, and pilot data show that variable interval (VI) reinforcement schedules, once behavior change occurs, hold potential for maintaining gains when administered infrequently. Assessing methods to extend benefits of these interventions is of paramount scientific and clinical concern. This study will evaluate a novel approach in which reinforcement frequency varies by patient performance. In this intervention, reinforcement will be available for 24 weeks, on a progressive VI schedule, that adapts according to patient status. Patients who maintain abstinence earn maximum reinforcers as infrequently as every three weeks on average, while frequency of monitoring and reinforcing abstinence will increase in those who relapse until abstinence is re-instated.

To test efficacy, 280 patients with cocaine use disorder will be randomly assigned to: standard care (SC), SC+traditional twice weekly reinforcement, or SC+adaptive VI reinforcement. Evaluations will be completed at baseline and throughout 18 months to assess objective and self-reported indices of drug use, psychosocial problems, and HIV risk behaviors. Primary hypotheses are (1) the adaptive VI reinforcement intervention will improve outcomes relative to standard care during the treatment period and throughout follow-up, and (2) the adaptive VI reinforcement intervention will improve outcomes relative to the traditional reinforcement system. This study will also evaluate the roles of cognitive control and treatment outcome. Patients with better cognitive control are expected to maintain longer durations of abstinence across conditions. If these measures differentially relate to outcomes across treatments, such results suggest the potential of pairing reinforcement interventions to individuals most likely to benefit from them; they may also indicate possible markers of response in a treatment-specific manner. If cognitive indices mediate treatment response, future studies can refine interventions to improve cognitive processes and long-term outcomes.