Destructive behavior represents a comorbid condition of developmental disability for which risk increases with intellectual disability severity, communication deficits, and co-occurring autism spectrum disorder. Destructive behavior, such as self-injurious behavior and aggression, causes harm to the child and others and increases the risk for institutionalization, social isolation, physical restraint, medication overuse, and abuse. Clinicians have used functional analyses to identify the variables that reinforce destructive behavior and to develop effective, function-based treatments. Functional communication training (FCT) is an empirically supported, function-based treatment that decreases destructive behavior. Using FCT, the clinician teaches the child to use a functional communication response (FCR) to request the reinforcer maintaining destructive behavior, while placing destructive behavior on extinction. For example, if functional analysis results showed that attention reinforced destructive behavior, the clinician would provide attention when the child used the FCR ("Play with me, please") and would not provide attention for destructive behavior. Two limitations of FCT are that (a) schedules of reinforcement maintaining the FCR must often be thinned gradually to levels that are practical for caregivers to implement consistently in the home and in the community, and (b) this necessary process of reinforcement schedule thinning regularly causes destructive behavior to increase following initially effective treatment, a form of treatment relapse called resurgence. The current project aims to improve these limitations of FCT by (a) hastening the process of reinforcement schedule thinning by removing unnecessary schedule-thinning steps using the results of a progressive interval assessment and (b) mitigating the resurgence of destructive behavior by providing stimuli that highly compete with the reinforcer maintaining destructive behavior. The investigators will conduct a randomized clinical trial to evaluate the extent to which these two promising refinements to FCT improve the process of reinforcement schedule thinning, and an exploratory experiment will examine the interactive effects of these two approaches. This novel project has the potential to substantially improve standards of care guiding the treatment of severe destructive behavior and to improve the long-term outcomes for children and families afflicted by these debilitating behavior disorders.

The severe destructive behavior (e.g., self-injury, aggression) of children with intellectual developmental disorder is prevalent, often dangerous, and negatively impacts social integration and quality of life (Borthwick-Duffy, 1994; Crocker et al., 2006). Function-based interventions that rely on differential reinforcement of alternative behavior reduce such problematic behavior effectively (Greer et al., 2016; Hagopian et al., 1998; Rooker et al., 2013), but the clinical utility of this approach is hampered in two critically important ways. First, schedules of reinforcement maintaining alternative behavior must be thinned to levels that are practical for caregivers to implement consistently in the home and in the community (Greer et al., 2016; Hagopian et al., 2011). This necessary process of schedule thinning often requires substantial time and resources to complete and typically comprises the most expensive portion of routine, clinical service delivery for this referral concern. Second, and exacerbating the first, is the consistent finding that schedule thinning often produces a form of treatment relapse called resurgence (Briggs et al., 2018; Mitteer et al., 2022; Muething et al., 2020; Shahan et al., 2020). The experimenters propose two clinically indicated, and theoretically grounded, methods of accelerating the process of schedule thinning while mitigating the resurgence of severe destructive behavior.

Recent research from the investigators has shown that the quantitative theory of resurgence called Resurgence as Choice (RaC) (Greer & Shahan, 2019; Shahan et al., 2019; Shahan & Craig, 2017) accurately describes how decrements in the availability of reinforcement predict the amount of resurgence of destructive behavior during the process of schedule thinning (Shahan & Craig, 2017), a finding recently verified in a nonhuman animal study (Shahan et al., 2020). Importantly, clinical and laboratory research alike has shown, as RaC predicts, that simply thinning reinforcement in small, gradual steps, the approach most commonly used by clinicians, does not prevent resurgence-relapse inevitably occurs once the schedule reaches a break point (Briggs et al., 2018; Shahan et al., 2020; Shahan & Craig, 2017; Shahan & Greer, 2021). RaC theory states that motivation for the functional reinforcer maintaining destructive behavior (parameter a in RaC equations) plays an important role in determining whether and to what extent destructive behavior will resurge during schedule thinning (Greer et al., 2019; Shahan et al., 2019; Shahan & Craig, 2017), a prediction well-supported by recent pilot work from the investigators on (a) individualizing the starting point for schedule thinning using a progressive interval assessment (PIA) (Miller et al., 2021) and (b) providing competing stimuli to hasten schedule thinning (Fuhrman et al, 2018; Miller et al., 2021). According to RaC theory, these two clinically indicated manipulations facilitate schedule thinning by respectively (a) tailoring the initial schedule of reinforcement to each patient's unique level of motivation for the functional reinforcer and (b) dampening motivation for the functional reinforcer by delivering an alternative and competing source of reinforcement. Additionally, theoretical and empirical work in the area of behavioral economics provides independent support for these two manipulations. This project will further the clinical and theoretical understanding of how motivational variables affect resurgence as it occurs in practice, and the project has the potential to substantially improve standards of care guiding the treatment of severe destructive behavior.

The experimenters will conduct a study to identify whether quantitatively informed refinements can improve efficiency and efficacy of reinforcement schedule thinning when treating severe destructive behavior. The project has three specific aims:

Specific Aim 1: The experimenters will extend pilot work on the utility of individualizing the starting point for reinforcement schedule thinning based on the results of a PIA. The experimenters will accomplish this by conducting reinforcement schedule thinning in two distinct stimulus contexts, one informed by the results of a PIA and another not so informed.

Specific Aim 2: Basing the starting point on a PIA, the experimenters will assess the extent to which providing competing stimuli from a competing stimulus assessment quickens the process of schedule thinning when competing stimuli are available in one, but not another, unique stimulus context.

Specific Aim 3: The experimenters will examine the potential interaction effects between these two approaches by conducting PIAs with no, low, moderate, and high competing stimuli to determine the schedule duration at which schedule thinning should commence with each competing stimulus.

Inclusion Criteria:

• boys and girls from ages 3 to 17
• destructive behavior that occurs at least 10 times a day, despite previous treatment
• destructive behavior reinforced by social consequences
• stable protective supports for self-injurious behavior (e.g., helmet) with no anticipated changes during enrollment
• on a stable psychoactive drug regimen for at least 10 half-lives per drug or drug free
• stable educational plan and placement with no anticipated changes during the child's treatment

Exclusion Criteria:

• patients who do not meet the inclusion criteria
• patients currently receiving 15 or more hours per week of treatment for their destructive behavior
• Diagnostic and Statistical Manual-5 diagnosis of Rett syndrome or other degenerative conditions (e.g., inborn error of metabolism)
• a comorbid health condition or major mental disorder that would interfere with study participation
• occurrence of self-injury during study assessments that presents a risk of serious or permanent harm (e.g., detached retinas) based on routine clinical-risk assessment
• patients requiring changes to protective supports for self-injury or drug treatment, but investigators will invite these patients to participate when protective supports and drug regimen are stable

The investigators plan to share outcome data using the National Database on Autism Research (NDAR). The investigators will also share outcome data on the human participants in these two experiments who are not diagnosed with autism spectrum disorder to better understand how complex, comorbid disorders (e.g., self-injury) differ in individuals with and without autism spectrum disorder.

Within one year of the primary completion date of the studies.

In accordance with NDAR policy.