Paper Session 18: Implementation Science

Background: Changes or adaptations often occur while implementing interventions inside healthcare settings. Categorized as planned or reactive in nature, adaptions are often expected to occur at any point across the course execution. However, the field of implementation science (IS) currently has no consensus on how to best capture, evaluate, and utilize information on adaptations that occur throughout the implementation process. While much research has focused on identifying the associated gap and proposing conceptual ideas on how to best address this, to date there is a scarcity in literature highlighting the practical application of any specific implementation strategies to track and evaluate the impact of adaptations on implementation outcomes. This study aims to highlight the use of a pragmatic multi-data approach in attempting to map and evaluate adaptations and their associated impacts.
Methods: This study used a multi-data approach including transcribed interviews from site-level workers and data logs from the research staff to track, monitor, and evaluate adaptations occurring pre, during, and post implementation. Unique data entries were reviewed individually and consolidated into adaptation events based on team consensus. Researchers used qualitative analysis to identify key themes and subthemes associated with PRISM/REAIM to identify impact to implementation outcomes.
Results: 18 adaptation events were identified in the data after consolidating data sources. These adaptation events ranged from specific larger-scale unique incidents to correlated themes of similar but smaller patterns of behavior across sites. Researchers identified several subthemes associated with adaptations including adaptations to communications, changes to the implementation strategy, and modifications to the intervention itself, as well as previously unidentified adaptations from field-level work.
Discussion: Given the difficulty in validating the tracked adaptations, researchers should have a clear description of what constitutes an adaptation. Additional focus on how to measure and track each adaptation should be identified and developed throughout the entire process. Finally, teams must develop methods to collect data relative to outcomes with focus on impact with the suggestion of incorporating greater specificity for REAIM outcomes as it relates to each adaptation.
Keywords: adaptation, PRISM, REAIM, implementation science, outcomes, mixed-methods

Author - Borsika Rabin, PhD, VA Center of Excellence for Stress and Mental Health
Co-Author - Laurie Lindamer, PhD, VA Center of Excellence for Stress and Mental Health (CESAMH)
Co-Author - Brian Blanco, LCSW, VA Center of Excellence for Stress and Mental Health
Co-Author - Chad Vacco, LPCC, VA Center of Excellence for Stress and Mental Health
Co-Author - John Gault, LCSW, VA Center of Excellence for Stress and Mental Health
Co-Author - Jaimi Nam, BS, UCSD
Co-Author - James Pittman, PhD, VA Center of Excellence for Stress and Mental Health

Title: Scaling Social Determinants of Health Screening Across a Pediatric Health System: Implementation Strategies, Barriers, and Facilitators of Systemwide Adoption

Background: Social determinants of health (SDOH) contribute to inequities in pediatric outcomes. Our most recent Community Health Needs Assessment (CHNA) identified over 10,400 homeless children and more than 212,000 children lacking reliable access to food within our catchment area. To better identify and address these disparities, our system launched electronic SDOH screening in July 2022. Initial adoption was limited: from July–December 2022, 15,294 screenings were completed across 40 clinics, with an average completion rate of only 47%.

Methods: To expand both the reach and equity of SDOH screening, we scaled into new care settings while directly addressing gaps in underperforming clinics. Our multipronged strategy included: (1) additional user testing to refine the screening tool, (2) patient family focus groups to guide improvements, (3) intensive interventions in low-screening clinics through monthly leader performance reports, rounding, and 1:1 staff support, (4) iterative workflow redesigns, (5) creation of real-time dashboards to track performance and disparities, and (6) retraining and reeducation to embed SDOH screening as a core component of comprehensive pediatric care.

Results: By July 2025, SDOH screening was active across all care settings. Screening volume rose sharply year over year, with 43,425 additional screenings (+134.6%) from 2023–2024 and 47,687 more (+63.0%) from 2024–2025, for a total gain of 91,112 screenings (+282.4%) across three years. In that time, more than 210,000 unique patients were screened, and the systemwide completion rate rose from 34% to 81%.

Conclusion: A multipronged implementation approach, combining workflow expansion, targeted support for low-performing clinics, iterative testing, and leadership engagement, enabled dramatic increases in SDOH screening volume and compliance across a large pediatric system. Continued efforts are needed to raise screening in newer areas, particularly the ED (26%) , while sustaining high performance in outpatient (77%) and inpatient (90%) settings.

AI Disclosure: A language processing tool driven by AI was used for proofreading and grammar support; all content and conclusions are author-generated.

Author - Forest Melton, M.S., M.H.A., Children's Health

Background:
Evaluation of nutrition incentives to improve nutrition security and fruit and vegetable (FV) intake has increased nationally. In January 2024, Rhode Island (RI) implemented the statewide Eat Well Be Well (EWBW) FV incentive program for Supplemental Nutrition Assistance Program (SNAP) participating- households, providing $0.50 in SNAP benefits for every $1.00 spent on fresh FV (up to $25/month) at select stores, directly linked to EBT cards. We conducted a mixed methods study examining trends in spending across participating stores and perceptions about barriers and facilitators to implementing similar state-level programs.

Methods:
We used aggregated, monthly EWBW spending data from 56,800 unique RI households from October 2023- February 2025 at both retailers. One retailer also provided deidentified, loyalty-card linked total, EWBW, and FV spending data by month. We conducted in-depth interviews with 13 stakeholders including those working in food insecurity, health departments, and retail, to gather perceptions of barriers and facilitators to EWBW implementation. Descriptive purchasing trends were analyzed using R studio and in-depth interviews were transcribed and dual-coded using NVivo to identify themes.
Results:
Over the entire period, mean household spending was $33.53 (SD 44.60), with households receiving an average of $5-$8 in EWBW benefits/month (representing <$20 spent on fresh FV), with >80% of transactions occurring at 1 of the 2 participating retail chains. In any month, only 6,758 unique households (12%) spent >$25 in EWBW spending. A similar trend in total spending was observed with the single retailer data, with an average total transaction of ~$25, $7 of which came from fresh FV purchases, resulting in minimal EWBW incentives. Stakeholder interviews contextualized these findings across 4 themes, noting: 1) implementation was limited by cost, timeline pressure, and lack of community involvement, 2) store-level factors affected retailer ability to participate, 3) there was low awareness of EWBW due to communication issues, and 4) multi-pronged approaches at all stages of development would enhance effective implementation.
Discussion:
Underutilization of the statewide EWBW program is likely due to varied factors including insufficient cohesion among stakeholders to promote EWBW and address barriers including limited retailer participation. Ongoing follow-up and evaluation will determine whether EWBW utilization increases.

Author - Maya Vadiveloo, PhD RD, University of Rhode Island
Co-Author - Emily Elenio, MPH, Brown University
Co-Author - Elena Hurtado-Solberg, Brown University
Co-Author - Alexia Sanchez Contreras, Brown University
Co-Author - Alison Tovar, PhD, MPH, Brown University

Background: The Nebraska CORD 3.0 project packaged an evidence-based FHWP, Building Healthy Families (BHF), into an online train-the-trainer resource to determine if community-based implementation teams that used the training could recruit and deliver BHF to families in rural communities. The purpose of this presentation is to report the community-level adoption, implementation, and maintenance outcomes comparing communities that had access to the online resource alone (BHF-PO) or in combination with an action learning collaborative (BHF-LC).

Methods: Adoption was defined as the number of eligible communities that participated in the RFA strategy to identify and recruit rural communities that identified childhood obesity as a priority and assembled an implementation team of supporting organizations that could train, recruit, and deliver the Building Healthy Families program. Implementation fidelity, the primary outcome, was defined as the quality of delivery and measured from a direct observation protocol in all communities. Barriers and facilitators to adoption and implementation by communities were measured through in-depth interviews using an iPARIHS and RE-AIM informed interview guide and rapid qualitative analytic approaches to coding. Maintenance was defined as the number of enrolled communities that continued to offer BHF to families after the two required cohorts.

Results. Adoption: 28 of 90 rural counties were within the recruitment areas, and 7 of 8 communities that responded to the RFA adoption strategy were selected for the pilot trial. The 7 communities were assigned to either BHF-LC (n = 4) or BHF-PO (n = 3). The seven communities assembled community-based implementation teams (CITs), n=32 members across communities (BHF-LC n=17, 4 CITs; BHF-PO= 15). Implementation: BHF-LC CITs delivered more sessions (11.75/12 vs 10/12) and had stronger adherence to protocol based on learning objectives (84% vs 72%) and activities (91% vs 73%) by number of sessions completed. Maintenance: All 7 communities planned additional cohorts, with two initiating an additional cohort at the close of the trial.

Conclusion: While the RFA process identified communities with interest in delivering BHF, the BHF-LC implementation strategy led to superior implementation quality, suggesting that while the provision of resources for training and program implementation is necessary, it may not be sufficient to achieve high implementation fidelity.

Presenter - Jennie Hill, PhD, University of Utah
Co-Author - Kate Heelan, PhD, University of Nebraska Kearney
Co-Author - Caitlin Golden, University of Nebraska Kearney
Co-Author - Bryce Abbey, PhD, University of Nebraska Kearney
Co-Author - Ali Mercado, University of Nebraska Kearney
Co-Author - Gwendolyn Porter, PhD, Consultant
Co-Author - Tzeyu Michaud, PhD, University of Nebraska Medical Center
Co-Author - Paul Estabrooks, PhD, FSBM, University of Utah