An evaluation of generalization, renewal, and resurgence during functional communication training conducted via telehealth

Previous research suggests that differential reinforcement procedures may inadvertently strengthen problem behavior, resulting in treatment relapse (i.e., recurrence of problem behavior) when chges (e.g., intentional treatment withdrawal or treatment fidelity errors) are encountered following successful treatment. The current study evaluated one potential solution to the problem of treatment relapse based on the procedures proposed by Mace et al. (2010), which involved initially implementing treatment within a context with a minimal history of reinforcement for problem behavior rather than in the treatment context. Treatment relapse was then evaluated by conducting renewal and resurgence analyses at different points of time during treatment. Participants were four young children diagnosed with autism whose problem behavior was maintained by negative reinforcement. The children's caregivers implemented all procedures in their homes within multielement and reversal designs with coaching provided by a behavior consultant via telehealth. Interobserver agreement was collected on 34.2% of sessions and averaged 97.9% across participants. Following a functional analysis and an extinction baseline, functional communication training (FCT) was initially implemented in alternative contexts (i.e., play and neutral tasks) that had minimal history of reinforcement for problem behavior. This was conducted in order to strengthen manding and task completion without inadvertently strengthening problem behavior. I programmed for the generalization of mands and task completion by training sufficient exemplars (three alternative contexts) and programming common stimuli (picture cards, safespot, microswitch). FCT was then implemented in the treatment context (target demand), and extinction probes were conducted intermittently throughout treatment. Results demonstrated that problem behavior was reduced on average by 97.8% across all participants by the end of treatment. Treatment maintenance results showed that only minimal treatment relapse occurred. Clinically significant renewal of problem behavior only occurred with one participant, and little or no resurgence of problem behavior occurred for two participants during the first extinction probe. With minimal demand fading, resurgence was reduced during the second extinction probe for the other two participants. Manding and task completion generalized to the treatment context for most participants. These behaviors also persisted during the extinction probes showing that they were strengthened across treatment. Thus, these results showed substantial improvement in achieving maintenance than what has been achieved in previous studies evaluating treatment maintenance following FCT.