Jankowski, M. K., Schifferdecker, K. E., Butcher, R. L., Foster-Johnson, L., & Barnett, E. R. (2018). Effectiveness of a trauma-informed care initiative in a state child welfare system: A randomized study. Child Maltreatment, 1-12.
This study was conducted as part of a larger intervention study that sought to improve the social-emotional well-being and developmentally appropriate functioning of children and families served by New Hampshire’s Divisions for Children, Youth and Families (DCYF) through Trauma Informed Care (TIC). The larger intervention involved the installation and implementation of a new, web-based “Mental Health Screening Tool” to assess trauma exposure of children and caregivers involved with the child welfare system. Staff were trained through three monthly workshops (seven hours in total) focused on the basic principles of TIC and its application to child protection and juvenile justice practice. Consultative support within each District Office (DO) for 1/2 to 1 day per week for 3 months following training was also provided. This study was used to assess the degree to which the intervention changed TIC attitudes, skills, perceptions and behaviours of staff members working at DCYF.
This study evaluated the TIC intervention with 372 DCYF staff using randomized, matched pairs. Randomization was done by matching 10 DOs into five pairs based on case mix, size, and geography. Each pair of DOs were randomly assigned to either an Early Intervention Group (Cohort 1) or Late Intervention Group (Cohort 2) in order to assess differences between the groups at different time points. A Trauma-Informed Care survey was conducted prior to the start of the intervention (Time 1) and at two other time points between September and October of 2015 (Time 2) and September and October of 2016 (Time 3). This survey measured six outcome variables including trauma screening, case planning, family involvement, progress monitoring, collaboration, and system performance. Cohort 1 received the intervention from November 2014 to June 2015 and Cohort 2 received the intervention from November 2016 to June 2017.
Linear mixed modelling was used to determine the effect of the intervention on each of the six outcome variables. Models included two fixed factors: cohort (early vs. late intervention) and time (three survey time points) and an interaction between cohort and time. Results showed significant differences for three of the six outcome variables including case planning, trauma screening, and system performance. By Time 3, Cohort 2 improved significantly in trauma screening initial case planning, and perceptions of system performance while Cohort 1 remained steady.
While the reason Cohort 1 did not improve following the intervention is unknown, it is speculated that staff in Cohort 2 may have been particularly receptive to a TIC approach and the additional support provided via the project despite the continued opioid crisis in New Hampshire and increased number of children entering the child welfare system. Additionally, there are some limitations to this study including the use of self-report data, a small sample size and missing data. Further, the size and homogenous White Caucasian ethnic and racial demographic of New Hampshire’s population may limit the generalizability of the results.