Abstract
BACKGROUND: Care coordination tools and toolkits can be challenging to implement. Practice facilitation, an active but expensive strategy, may facilitate toolkit implementation. We evaluated the comparative effectiveness of distance coaching, a form of practice facilitation, for improving the implementation of care coordination quality improvement (QI) projects. METHODS: We conducted a mixed methods evaluation of the Coordination Toolkit and Coaching (CTAC) initiative. Twelve matched US Veterans Health Administration primary care clinics were randomized to receive coaching and an online care coordination toolkit ("coached"; n = 6) or access to the toolkit only ("non-coached"; n = 6). We did interviews at six, 12, and 18 months. For coached sites, we'ly collected site visit fieldnotes, prospective coach logs, retrospective coach team debriefs, and project reports. We employed matrix analysis using constructs from the Consolidated Framework for Implementation Research and a taxonomy of outcomes. We assessed each site's project(s) using an adapted Complexity Assessment Tool for Systematic Reviews. RESULTS: Eleven sites implemented a local CTAC project. Eight sites (5 coached, 3 non-coached) used at least one tool from the toolkit. Coached sites implemented significantly more complex projects than non-coached sites (11.5 vs 7.5, 95% confidence interval 1.75-6.25, p < 0.001); engaged in more formal implementation processes (planning, engaging, reflecting and evaluating); and generally had larger, more multidisciplinary QI teams. Regardless of coaching status, sites focused on internal organizational improvement and low-intensity educational projects rather than the full suite of care coordination tools. At 12 months, half the coached and non-coached sites had clinic-wide project implementation; the remaining coached sites had implemented most of their project(s), while the remaining non-coached sites had either not implemented anything or conducted limited pilots. At 18 months, coached sites reported ongoing effort to monitor, adapt, and spread their CTAC projects, while non-coached sites did not report much continuing work. Coached sites accrued benefits like improved clinic relationships and team QI skill building that non-coached sites did not describe. CONCLUSIONS: Coaching had a positive influence on QI skills of (and relationships among) coached sites' team members, and the scope and rigor of projects. However, a 12-month project period was potentially too short to ensure full project implementation or to address cross-setting or patient-partnered initiatives. TRIAL REGISTRATION: NCT03063294 .