Building Community Capacity for Evidence-Based, Parent-Mediated Early Intervention: Effectiveness of a Train-the-Trainer Approach

Background: Parent-mediated naturalistic developmental behavioral interventions (NDBI) have demonstrated positive child and family outcomes for young children with ASD (Zwaigenbaum et al., 2015; Burrell & Borrego, 2012). Despite demonstrated benefits, however, there has been little dissemination of parent-mediated NDBI into community settings, where the majority of children receive services (Stahmer et al., 2016). Community providers with expertise in both the NDBI strategy use and partnering with parents are needed to effectively deliver NDBIs. However, based on the range of mechanisms and avenues through which children with ASD may receive services, community providers have variable educational backgrounds and ASD specific experience. As communities struggle to serve the growing number of young children with ASD with appropriate intervention, more information is needed on effective provider training that adequately equips interventionists to work effectively with families and meet the specialized needs of children with ASD.

Objectives: The objective of this presentation is to demonstrate training outcomes from a train-the-trainer study of an evidence-based, parent-mediated NDBI for community early interventionists (Project ImPACT for Toddlers or PI for T; Ingersoll & Dvortscak, 2010).

Methods: A multiple probe design was used to examine effectiveness of the training model. Fourteen leaders from 12 publicly-funded community agencies participated. Agencies included school-based early intervention (N=2), infant and early childhood services (N=6), speech language (N=1), ASD specific agencies (N=1) and a Children’s Hospital ASD clinic (N=1), and family health programs (N=1). Agency leaders received 12 weeks of training in PI for T from the research team. Leaders were trained in groups of 4-5, per the multiple probe design. After agency leader training was complete, each leader returned to their individual agency to deliver training to 2-3 interventionists (45 providers total). Intervention fidelity in PI for T was measured across the training period (before, during and after) at multiple time points. A total of nineteen fidelity criteria were rated on a 1-5 scale, where 1 = does not implement and 5 = implements competently and consistently.

Results: Agency leaders (n=14, 100%) were successfully delivered training at their agencies, utilizing both online (n=3) and in-person approaches (n=11). Intervention fidelity data are variable but demonstrate growth for strategies that providers were not implementing at baseline. On average, providers were using 21% of strategies correctly at baseline strategies (M=3.5 or above), including limiting distractions in the environment (baseline M = 4.1) and adjusting animation (baseline M=3.6). After training, on average, providers were using 74% of strategies correctly; waiting for a child response (baseline M=3.2, post-training M=3.8) and making comments/avoiding questions (baseline M=3.3, post-training M=4) demonstrated the most change. Specific support for sustainment beyond the training period (e.g., assistance securing public funding) facilitated agencies’ continued use of the approach. Differences in post-training fidelity based on provider and agency characteristics will also be explored.

Conclusions: A train the trainer model is an effective method to build community capacity for evidence-based early intervention across a range of service settings. It may be possible to deliver training more efficiently by tailoring training to gaps in provider knowledge.