Implementation of a cardiovascular toolkit in primary care increased women Veterans' engagement in behavior change programs: results from a non-randomized stepped wedge trial.

Abstract

Cardiovascular (CV) disease is the leading cause of death among U.S. women, yet women have a limited understanding of their CV-related morbidity and mortality risks. Provider-, system-, and patient-level barriers point to a need for multi-level evidence-based strategies to facilitate CV risk reduction. Supported by the Replicating Effective Programs implementation strategy, we implemented a CV Toolkit in primary care clinics for women Veterans. The objective was to evaluate the effect of CV Toolkit implementation on participation in behavior change programs that target CV risk reduction. In partnership with the Veterans Health Administration (VA) Office of Women's Health and National Center for Health Promotion and Disease Prevention, we conducted an implementation trial of a CV Toolkit at five geographically diverse VA sites between March 2017-March 2020. Using a non-randomized stepped wedge design, we evaluated the effect of CV Toolkit implementation on participation in the VA MOVE! weight management program, and on participation in health promotion and disease prevention (HPDP) programs (coaching, facilitated groups, etc.) and/or complementary integrative health (CIH) programs (yoga, meditation, etc.). We utilized a three-level (patient, site, time) non-linear fixed effect model with stratification by age (65 and older versus younger). Patient participation, utilization, and demographics were extracted from VA administrative data for all women with at least one primary care visit at a participating site from December 2016-March 2020 (n = 6009). Women were on average 46 years old; 49% were white, 32% Black, 17% Hispanic; and over a third had CV risk factors and/or mental health diagnoses. For women 65 years and older (n = 540), active toolkit implementation was associated with increased odds of MOVE! participation (OR = 1.09; 95% CI:1.030-1.152) compared to when the toolkit was not active either within or between sites. Women younger than 65 (n = 5469) had increased odds of using HPDP/CIH programs during active toolkit implementation (OR = 1.01; 95% CI:1.002-1.022). A multilevel intervention and implementation strategy were associated with improved patient-level outcomes-a rarity in implementation trials. Precision implementation may offer important next steps in understanding causality and further specifying how implementation strategies can optimize clinical and implementation outcomes. Clinical Trials.gov, NCT02991534. Registered 12-09-2016, https://clinicaltrials.gov/study/NCT02991534?cond=NCT02991534&rank=1.