3 - (SYM 139) Implementation Outcomes from the Reaching Families Clinical Trial

We will share findings from the Reaching Families multisite cluster randomized clinical trial (RCT), which investigated the impact of training providers on clinical reasoning and implementation of evidence-based procedures for improving treatment engagement. Across three school-based mental health organizations in Los Angeles and South Carolina, we randomized 144 professionals (30 supervisors, 114 therapists) to a Coordinated Knowledge System (CKS) condition and a Practice Guidelines (PG) control condition. Families participating in mental health services were screened for risk of low treatment engagement around their fourth session (N = 1,352). Of 365 youth who screened positive and met eligibility criteria, 231 consented and were allocated to one of the two conditions, and 221 youth received at least one session of any intervention. Average participant age was 12.78 (SD = 3.44); race was 43% Latino/a, 35% Black, 20% White and 2% other; and 54% of students identified as male, 64% as female.

The CKS condition used a system of knowledge resources synthesizing findings from 40 RCTs on engagement interventions, with prompts to guide clinical reasoning of supervisors and therapists to identify a focus, consider relevant practices with supporting evidence, plan for practice delivery, implement the practice in session, and review the implementation in subsequent supervision. The PG condition used set of guidelines for best practices based on the same 40 RCTs that informed the CKS condition but did not include a set of decision resources.

We transcribed 1,019 digitally recorded events (627 supervision meetings, 392 treatment sessions) into 270,835 discrete text components, which were then coded using a hierarchical compound code system for activities and objects (e.g., activity = “selects” and object = “psychoeducation”), yielding roughly 38,000 coded excerpts. Results showed significant large effects of CKS relative to the PG control group on supervisor and provider instances of considering, identifying, or selecting as a focus an engagement problem present in the youth (odds ratios = 74.76, 42.47, and 85.53 respectively), and similar effects for considering, selecting, and applying matching evidence-based procedures, (odds ratios = 12.11, 10.96, and 6.75 respectively). Results demonstrate the benefit of a coordinated decision support system to manage a complex but ubiquitous challenge in community practice.