Title: The Relationship Between Treatment Attendance, Adherence, and Outcome in a Caregiver-Mediated Intervention for Low-Resourced Families of Young Children with ASD

Background: Families of lower socioeconomic status have been underrepresented in early intervention research for children with Autism Spectrum Disorders (ASD).  The examination of barriers to intervention research participation has been studied in the broader psychotherapeutic field, but not specifically within ASD.  Such research is especially important given the benefit of early intervention on long-term prognosis. 

 Objectives:  The purpose of the study is to investigate barriers to participation by examining how family demographic factors (family income, parent education, ethnicity/race, stress etc.) predict treatment attendance and adherence in a caregiver-mediated intervention for young children with ASD.  Furthermore, the study will examine whether treatment attendance and adherence predict outcome.

 Methods: This study is a secondary analysis of a caregiver-mediated randomized controlled trial targeting core deficits of ASD including joint attention and symbolic play (Kasari et al., under review).  Caregivers and their child with ASD were randomly assigned to a caregiver-mediated condition (CMM) in which parents received twice weekly in-home sessions or a caregiver education condition (CEM) in which parents were delivered the same intervention content, but in group lecture format.  Demographic variables included parent education, ethnicity, family income, number of children in the family, and parent-reported stress.  Treatment attendance was measured as the percentage of sessions attended over the course of the 12-week intervention period while treatment adherence was measured as the degree to which the interventionist rated the caregiver’s involvement in treatment strategies.  Treatment outcome was operationalized as the amount of time in seconds caregiver-child dyads were jointly engaged during a play interaction.  Generalized linear regressions were modeled to examine 1) the influence of demographic variables on treatment attendance, 2) the influence of demographic variables on adherence, and 3) the influence of treatment attendance and adherence on dyadic joint engagement.

 Results: 147 families were assessed and randomized to a treatment condition.  30 families dropped out of the study prior to attending any treatment sessions, 20 of which were in the CEM condition (Χ²=4.01, p <. 05).  Preliminary analyses found 1) treatment attendance was predicted by an interaction between treatment condition and number of children in the family, such that receiving CEM and having more children was predictive of lower attendance (β = -8.34, p<. 05); 2) higher levels of caregiver stress were predictive of greater treatment adherence (β = -.01,p < .05); and 3) receiving CMM and greater treatment adherence were predictive of longer joint engagement (β = 103.59, p<. 001; β = 124.4, p<. 001, respectively), while treatment attendance had no effect (β = 1.01, p=.223). 

 Conclusions: Preliminary results suggest some influence of demographic variables on treatment attendance and adherence.  Furthermore, caregiver involvement was predictive of treatment outcome, while caregiver attendance was not.  These results stress the importance of considering demographic variables in research design when considering barriers to treatment attendance and adherence, and call for further analysis of the potential mediating influence these factors may have on treatment outcome.