Modularized evidence-based clinical decision-making: A rescue protocol for non-responders

Background:  A vexing issue in treating children with ASD is that there is great heterogeneity in presentation and in response to evidence based treatments (EBTs).  Recognition of this heterogeneity in many areas of mental health, along with the realization that EBTs for different conditions have a variety of core principles in common, has led researchers to consider a modularized approach to the application of EBTs.  In this approach, modules (or specific behavioral strategies) from one treatment manual may be flexibly applied with those of another treatment manual to address the complex, and possibly heterogeneous, needs of the child. Since the decisions to apply different modules may unfold over time (e.g., across different treatment sessions) as more is learned about the child’s complex needs or as needs change, this modularized approach can be seen as a form of adaptive intervention. However, this treatment approach has not been studied in children with autism. In children with ASD who are minimally verbal, this type of modularized, adaptive approach may be especially useful as a “rescue protocol” for children who have been identified as non- or slow-responders to a previous trial of a manualized EBT, such as discrete trials training (DTT) or joint-attention social play (JASP) intervention.

Objectives:  The overarching objective of this presentation is to present pilot data on the application of this adaptive, modularized approach among children identified as non-responders or slow-responders to a 6 week trial of DTT or JASP. Specifically, we will present data on the usefulness of implementing a “treatment dashboard” used to organize the choices of modules in a novel, blended DTT+JASP treatment approach for these non- or slow-responding children. We refer to the blended DTT+JASP intervention as a “rescue protocol”, defined as combinations of modules and strategies from both DTT and JASP.

Methods:  Case-studies are drawn from a SMART design in which (i) minimally verbal children with ASD are randomized initially to one of two EBTs (DTT or JASP), and (ii) after 6 weeks, non- or slow-responders to initial DTT or JASP are randomized to continue on their initial treatment for an additional 6 weeks or to the rescue protocol (blended DTT+JASP). 

Results:  We will present data from illustrative case-studies of the treatment dashboard including (i) the assessments, treatment history, and child responses used by clinicians to select from a systematic menu of DTT+JASP modules in the dashboard, (ii) the rules used to make these decisions, and (iii) the benefits, challenges and obstacles to implementing such an approach, including acceptability and feasibility issues. 

Conclusions:  Minimally verbal school aged children have potential for making significant gains in spoken spontaneous language using current EBTs.  However, some will continue to make slow progress, and may benefit from combinations of treatment plans that are systematically implemented based on the child’s treatment history and progress in treatment.  A modularized, adaptive intervention approach using new behavioral health reporting systems can assist clinicians and researchers in personalizing the implementation of significant components of these EBTs.