Using interrupted time series data to assess the efficacy of the CARE program model for residential care

The current study examined the impact of a setting-level intervention on the prevention of aggressive or dangerous behavioral incidents involving youth living in group care environments. Eleven group care agencies in a... [ view full abstract ]

The current study examined the impact of a setting-level intervention on the prevention of aggressive or dangerous behavioral incidents involving youth living in group care environments. Eleven group care agencies in a southeastern U.S. state implemented Children and Residential Experiences (CARE), a principle-based program that helps agencies use a set of evidence-informed principles to guide programming and enrich the relational dynamics throughout the agency. All agencies served mostly youth referred from child welfare. The 3-year implementation of CARE involved intensive agency-wide training and on-site consultation to agency leaders and managers around supporting and facilitating day-to-day application of the principles in both childcare and staff management arenas. Agencies provided data over 48 months on the monthly frequency of behavioral incidents most related to program objectives. Using multiple baseline interrupted time series analysis to assess program effects, we tested whether trends during the program implementation period declined significantly compared to the 12-months before implementation. Mixed effects negative binomial regression models were constructed to carry out the analyses, adjusting for between-agency variability at the start of the intervention and over time. Results showed significant program effects on incidents involving youth aggression toward adult staff, property destruction, and running away. Effects on aggression toward peers and self-harm were also found, but were less consistent. Staff ratings of positive organizational social context (OSC) predicted fewer incidents, but there was no clear relationship between OSC and observed program effects. Findings support the potential efficacy of the CARE model and illustrate that interventions at the setting level may contribute to preventing coercive caregiving patterns and increasing opportunities for healthy social interactions.