4140 — Determinants of implementation success in complex inter-organizational relationships: a mixed-methods evaluation of Veteran Directed Care
Lead/Presenter: Nina Sperber,
COIN - Durham
All Authors: Sperber NR (Center of Innovation to Accelerate Discovery and Practice Transformation, Durham), Miech, EJ (Center for Health Information and Communication, Indianapolis), Clary AS (Center of Innovation to Accelerate Discovery and Practice Transformation, Durham) Perry K (Center of Innovation to Accelerate Discovery and Practice Transformation, Durham) Edwards-Orr M (Applied Self-Direction, Boston) Van Houtven CH (Center of Innovation to Accelerate Discovery and Practice Transformation, Durham) Thomas KS (Providence VA Medical Center) Rudolph J (Providence VA Medical Center)
Veteran Directed Care (VDC) provides a flexible budget for services to Veterans at risk of nursing home placement. VDC is administered through Aging and Disability Network Agencies (ADNAs), community-based providers affiliated with the US Department of Health and Human Services Administration for Community Living. As part of an evaluation considering both sides of the partnership (i.e., VA medical centers (VAs) and community-based providers), we sought to pinpoint determinants of successful VDC implementation.
Focusing on 5 different VAs that started VDC in 2017 or 2018, we conducted individual interviews with VA and ADNA coordinators about implementation at two time points: two and eight months after starting. We systematically applied over 20 determinants from the Consolidated Framework for Implementation Research (CFIR) directly to interview transcripts and scored each one for valence and magnitude. The 5-point scale ranged from +2 ("strong positive influence on implementation") to -2 ("strong negative influence"). We scored each determinant at three different levels of analysis: VA, community-based provider, and VA/community-based provider partnership, yielding over 400 data points. To find determinants of successful VDC implementation, 6 or more Veterans enrolled at eight months, from among those scored, we used Configurational Comparative Methods (CCMs), a mathematical approach for identifying necessary and sufficient conditions associated with an outcome of interest.
At the VA/community-based provider partnership level, External Change Agent and Networks and Communication uniquely distinguished programs with implementation success. Both CFIR determinants were sufficient for implementation success only at the +2 level (i.e., strong positive influence). The VA side of the VDC partnership appeared to drive the External Change Agent +2 score: while the community-based provider score could be anywhere between 0 and +2, the VA score had to be +2 for implementation success to occur. Alternatively, Networks and Communication needed to present as +2 from the community-based provider, whereas the VA score could be either +1 or +2 for implementation success to occur.
In inter-organizational implementation like VDC, each partner may require different conditions for overall implementation success.
Configurational Comparative Methods offer new ways to identify necessity and sufficiency of specific conditions for implementation success.