2019 HSR&D/QUERI National Conference

1027 — A Randomized Stepped Wedge Hybrid-II Trial to Implement the Collaborative Chronic Care Model in VA General Mental Health Clinics

Lead/Presenter: Mark Bauer,  COIN - Bedford/Boston
All Authors: Bauer MS (CHOIR, Behavioral Health QUERI), Miller CJ (CHOIR & Behavioral Health QUERI), Kim B (CHOIR & Behavioral Health QUERI) Lew R (CHOIR & MAVERIC) Stolzmann K (CHOIR & Behavioral Health QUERI) Sullivan J (CHOIR & Behavioral Health QUERI) Riendeau R (CHOIR) Pitcock J (Behavioral Health QUERI) Williamson A (CHOIR) Connolly S (CHOIR) Elwy AR (CHOIR & Behavioral Health QUERI) Weaver K (VACO Office of Mental Health and Suicide Prevention)

Collaborative Chronic Care Models (CCMs) have extensive controlled trial evidence for effectiveness in serious mental illnesses, but there is little evidence regarding feasibility or impact in typical practice conditions. We determined the effectiveness of implementation facilitation on establishing the CCM in mental health teams, and its impact on health outcomes of team-treated individuals.

We used a randomized stepped wedge trial in Behavioral Health Interdisciplinary Program (BHIP) teams in outpatient general mental health clinics of nine VA facilities, using blended internal-external facilitation. Facilitation combined a study-funded external facilitator with a facility-funded internal facilitator working with a designated team for one year. We hypothesized that facilitation would be associated with improvements in both implementation and intervention outcomes (hybrid-II trial). Implementation outcomes included the clinician Team Development Measure (TDM) and proportion of CCM-concordant team care processes. The study was powered for the primary health outcome, VR-12 Mental Component Score (MCS). All Veterans treated by designated teams were included for hospitalization analyses, based on administrative data; a randomly selected sample was identified for health status interview. Individuals with dementia were excluded. For implementation outcomes, 62 clinicians were surveyed; site process summaries were rated for CCM concordance.

The population (n = 5,596) included 881 (15%) women, average age 52.2+14.5. The interviewed sample (n = 1,050) was similar, but oversampled for women (n = 210, 20.0%). Facilitation was associated with improvements in TDM subscales for role clarity and team primacy. Percentage of CCM-concordant processes achieved varied (44-89%). No improvement in veteran self-ratings, including the primary outcome, was seen. However, in post-hoc analyses MCS improved in veterans with > 3 treated mental health diagnoses versus others. Mental health hospitalization rate demonstrated a robust drop during facilitation; this finding withstood four internal validity tests.

Working solely at the clinician level with minimal study-funded support, CCM implementation yielded provider and Veteran benefits. Although impact on self-reported overall population health status was negligible, health status improved for complex individuals, and hospitalization rate declined.

Facilitating CCM implementation provides a potential model for realigning VA outpatient general mental health care with an evidence-based model that improves provider team function and Veteran outcomes.