Using the Dynamic Sustainability Framework to Adapt Family Navigation for Children with Autism Spectrum Disorder: Results from a Quasi-Experimental Trial

Background: Family Navigation (FN) is an evidence-based case management approach to helping families overcome systems and person-level barriers to care. Our previous pilot work in a population of low-income, minority children demonstrates the feasibility, acceptability, and suggests effectiveness of FN to enhance access to services for children with autism spectrum disorders (ASD).

Objectives: In the current study, we describe: 1) how we adapted FN to improve implementation; and 2) compare the implementation of this adapted version of FN to a previous trial of FN for children with ASD.

Methods: We used the Dynamic Sustainability Framework (DSF) and data from our previous pilot of FN to inform adaptations. First, we worked with an advisory group of clinicians and researchers to make specific adaptations – the goal of which was improved implementation. Then, we recruited 40 parents of children with positive ASD screens and randomly allocated them to receive an adapted form of FN or usual care. Using a quasi-experimental design, we compared implementation outcomes (utilization, feasibility, and acceptability) between the recruited population and a population of 40 parent-child dyads who participated in a previous pilot of a non-adapted version of FN. Finally, we measured the clinical outcome - time to diagnostic resolution - using survival analysis.

Results: Adaptations were made in all three domains of the DSF: Intervention (e.g. addition of motivational interviewing training), Practice Setting (e.g. change in recruitment location), and Ecological System (e.g. changes in appointment scheduling). Compared to the un-adapted version of FN, families referred to the study were more likely to have an initial visit with the FN (97.7% versus 81.4%, p<0.001), and had more interactions (phone, text, in-person) with their FN (mean of 7 interactions v. 22, p<0.001). Parents rated the relationship with their navigator (on the satisfaction with navigation questionnaire) similarly between groups (mean score 3.80 v. 3.82, p=0.8). For our clinical outcome, FN improved diagnostic resolution for children at risk of ASD at similar rates in the original and adapted FN model (95% versus 90%, p=0.15).

Conclusions: Using the DSF to adapt FN improved implementation without altering satisfaction with or effectiveness of the intervention. These data support an adaptive approach to future implementation of FN in new settings.