Implementing a Self-Directed Prevention-Based Program with Foster Families: Predictors of Program Initiation and Completion

Background & Purpose
Teens in foster care experience elevated rates of health risking behaviors. However, foster families experience a wide range of barriers to using prevention programs. Although prevention program developers and implementers would benefit from knowing what individual characteristics predict program initiation and completion, very little research has been conducted on predictors of program uptake and completion among foster caregivers. We report on predictors of program initiation and completion within the randomized trial of Connecting, a family-based, prevention intervention adapted from Staying Connected with Your Teen to better meet the needs of teens in foster care and their caregivers. The program is self-directed in order to be more accessible and flexible. Weekly telephone contacts with a family consultant help to troubleshoot barriers and motivate completion.

Method
Data & Sample:
Teens in foster care (n=110, age 11-16 years) who did not have serious substance use, behavior problems or criminal justice involvement were randomly assigned with their caregivers to the intervention after recruitment. Just over half of the teens were female (55%), 25% were Hispanic, 19% were Black and 10% identified as American Indian or Alaskan Native. On average teens had been in their placement for 1.67 years, but about 32% had been with this caregiver less than 1 year. About 37% were with a relative.

Analysis
Data from family consultants' (FC) and caregivers' (CG) surveys pre and post intervention are used to examine predictors of program initiation (doing any program activities) and completion (doing 60% or more). Bivariate associations and logistic regressions are used to examine associations between teen and CG characteristics at baseline and program initiation and completion. Effect sizes and statistical significance will be reported, and responses to open ended questions about barriers to completion will be shared. Teen characteristics include age, gender, ethnicity, and length of current placement. CG characteristics include relationship to teen (kincare), age, ethnicity, substance use, family size, having a partner, and employment.

Results
Of the 110 randomly assigned families, 97 (88%) initiated the program. Fifty-eight (52%) families completed 60%+ of the 94 program activities. Bivariate associations indicate that CGs who had full-time employment were more likely to complete (70% vs. 48%) as were those with a spouse/partner (68% vs 47%). Logistic regressions indicate CGs who smoke cigarettes are less likely to initiate (B=-1.77, p=.04) or complete (B=-1.63, p=.02). Lack of time and other stressors were the primary reasons given for not completing.

Discussion
Adapting evidence based programs for use in foster families requires attention to the particular issues and barriers foster caregivers face. A self-directed program with phone support overcomes some barriers, however most caregivers still report having very little time for program activities. Most indicators of time demands (family size, employment) did not predict lower implementation, although having a partner to share responsibilities did increase the likelihood of program completion. Other time demands not formally measured but mentioned included the special needs of other children in the home, caregiver illness and/or injury, and deaths in the extended family.