Abstract

BACKGROUND: Implementation science aims to improve the uptake of evidence-based health care practices (EBPs) by defining the barriers that prevent their use, offering implementation strategies to overcome these barriers, and developing methods that allow clinicians and researchers to choose the strategies that best ad- dress the barriers they encounter. Recognizing that implementation strategy selection is often inefficient and idiosyncratic, implementation experts have called for methods to make strategy selection scientific, data- driven, and “precise”. To actualize “precision implementation”, there is a critical need to develop methods to 1) quickly and uniformly identify implementation barriers and facilitators, 2) track the use and effectiveness of im- plementation strategies, and 3) incorporate data and expert knowledge into the process of matching strategies to barriers. Without these improvements, we risk perpetuating implementation failures and health care dispari- ties. SIGNIFICANCE: Project OASIS (Optimize Approaches to Select Implementation Strategies) will support VA in becoming a “high reliability organization,” that provides Veterans with access to equitable and high-qual- ity care. The resulting methodological advancements are expected to apply outside of VA and across a wide variety of evidence-to-practice gaps. INNOVATION & IMPACT: Project OASIS will engage tremendous interdis- ciplinary expertise and data resources to address the complex, fundamental question of how to efficiently measure and select strategies to promote equitable, high-quality care. This work will leverage a one-of-a-kind dataset (38 project-years of implementation data) and techniques that are innovative in implementation science (e.g., user-centered design, machine learning). Project OASIS will advance the science of precision implemen- tation and our understanding of how to make decision aids that incorporate both machine learning and expert opinion. This work will provide innovative and practical approaches that will support broader public health and health equity goals. SPECIFIC AIMS: 1) Develop and test an implementation strategy selection Decision Aid (DA); 2) Develop and validate a survey to rapidly assess implementation barriers and facilitators; and 3) Refine a survey to track implementation strategy use over time. METHODOLOGY: Aim 1) The DA will be based on our existing, 38 project-years of data, that span 8 years, 12 EBPs, and 145 VA sites. We will first use Latent Class Analysis to identify site “types” using existing data about site barriers and facilitators coded using the Consolidated Framework for Implementation Research (CFIR)—a meta-framework of 39 constructs that can help or hinder implementation. For each site type, we will use tree-based methods to identify which implemen- tation strategies (based on the Evidence-Based Recommendations for Implementation Research, ERIC, taxon- omy of 73 strategies) were associated with best performance on EBPs at the site-level. The empirical output alongside existing expert opinion will form the backbone of a user-centered DA that recommends up to 5 strat- egies which could help to address the barriers users are experiencing to EBP adoption in practice. We will con- duct usability and acceptability testing of the DA followed by a hybrid type-III, cluster-randomized trial of the DA vs. an existing, expert-only tool in 20 VA sites to improve liver cancer screening. Aim 2) We will work with CFIR creators, clinicians, and Veterans to develop a refined, understandable, and relevant CFIR survey and evalu- ate its psychometric properties with n=48 clinicians, comparing results to semi-structured interview data (the traditional gold standard to collecting CFIR data). Aim 3) We will reconvene the ERIC developers and VA lead- ers, clinicians, and Veterans to further develop a user-centered implementation strategy survey. NEXT STEPS and IMPLEMENTATION: The Project OASIS DA and surveys are in high demand, as described in our letters of support. These research products will be immediately used by VA leadership and researchers, shared on the CFIR website—accessed nearly half a million times a year by a global audience—and posted on VA re- search websites.

Dimensions for VA

Dimensions for VA is a web-based tool available to VA staff that enables detailed searches of published research and research projects.

Learn more about Dimensions for VA.

VA staff not currently on the VA network can access Dimensions by registering for an account using their VA email address.     Search Dimensions for this project

DRA: Health Systems, Cancer

DRE: TRL - Applied/Translational, Data Science

Keywords: Cancer, Implementation

MeSH Terms: None at this time.