New Directions from Studying Integrated Behavioral Health and Primary Care for Patients with Multiple Chronic Conditions

Context: Despite a growing commitment to integrating behavioral health (IBH) in primary care (PC), high levels of integration that improve care and outcomes are elusive. As previously reported to NAPCRG (2021), “IBH-PC” studied the effectiveness of an intervention to support BH integration in a randomized controlled trial (RCT). A small improvement in BH integration was detected; there was no observed effect on patient outcomes.

Objective: Secondary analysis to understand relationships among intervention completion, timing of study measures relative to COVID-19, level of BH integration, and patient reported outcomes.

Study Design: Post-hoc analysis of pragmatic RCT.

Setting: 42 primary care clinics with on-site BH services, in 13 states, Sep 2017-Dec 2020.

Population: 2,426 adults with multiple chronic behavioral and medical conditions.

Intervention: Clinic-level support to implement IBH using online curricula, workbooks, remote coaching, and online learning community.

Outcomes and Predictors: Change in mean Patient-Reported Outcomes Measurement Information System scores (PROMIS) and median Practice Integration Profile (PIP); timing relative to COVID-19; and completion of the implementation stage of the intervention.

Results: Higher PIP scores were associated with better patient outcomes in all 8 PROMIS domains at baseline and follow-up. This association was significant for Anxiety (CI -0.12, 0.00, P=0.05 and CI -0.09, -0.01, P=0.01, respectively), Social Participation (CI +0.01, +0.09, P=0.02 and CI +0.02, +0.10, P=0.01), and Mental Health Summary (CI 0.00, +0.09, P=0.05 and CI 0.00, +0.09, P=0.04). Additional analysis found that a longer lapse between the start of COVID-19 and collection of final outcomes was associated with smaller improvement in PIP (P=0.04). Among clinics in the intervention arm (n=20), those that completed implementation (n=13) had significantly larger improvement in PIP scores (P=0.013) than those that did not; their patients also reported more improvement in PROMIS scores for Mental Health Summary (P=0.02), Social Participation (P=0.01), and Fatigue (P=0.04).

Conclusions: Post-hoc analyses identified associations among level of integration, improved patient reported health outcomes, and intervention completion. COVID-19 may have affected intervention completion. To help the field advance, future study must identify necessary conditions for successful implementation of IBH and the relationship between integration and health.