Examining linkages between disrupted care and chronic disease outcomes during the COVID-19 pandemic: a VAMC level spatio-temporal analysis

ABSTRACT Background: The global pandemic brought on by SARS-CoV-2 has profoundly impacted health and care for veterans, who are generally older, sicker and more economically vulnerable than the overall U.S. population. Veterans are likely to face lasting risks related to care disruptions. Understanding the long-term impact of these disruptions and varied responses across VA Medical Centers (VAMC) is critical to understanding (1) primary care needs moving forward, (2) identifying high risk patients for targeted interventions, and (3) reducing disparities exacerbated by care disruptions. Significance: Diabetes and hypertension are chronic conditions requiring substantial provider and patient care to manage and result in high healthcare cost. Roughly, a quarter of all veterans receiving care at the VA have diabetes and well over a third have hypertension. Diabetes and hypertension are associated with high cardiovascular risk and lead to serious complications, including stroke, heart disease, kidney failure, amputation and death. Racial, socioeconomic and geographic disparities in disease prevalence and progression are well documented; hence, it is critical that we understand the impact of the pandemic with a particular focus on "lessons learned" and health disparities that have widened. Specific Aims: Our aims are (1) To determine the long-term impact of disrupted care on chronic disease outcomes across the nation at the patient and VAMC level; (2) to identify veterans at high cardiovascular risk as a result of disrupted care and determine the extent to which disparities with respect to race-ethnic group, rural-urban residence and social vulnerability have widened during the pandemic; and (3) with input from our advisory panel, create a Power BI dashboard of cardiovascular monitoring and risk to disseminate our results. Methodology: We will create two retrospective cohorts of Veterans receiving primary care from 2017 through 2022: a diabetes and a hypertension cohort. Social vulnerability measures will be assigned at the census-tract level based on a veterans' residence. Our models are designed to investigate associations between individual-, census tract- and VAMC- level factors, health care delivery metrics, and health outcomes using complex GIS linkages and advanced spatio-temporal statistical methods. Delivery of care metrics include the extent to which cardiovascular risk factors are monitored and their levels (when monitored) early in the pandemic. Outcomes include prevalence of atherosclerotic cardiovascular disease (ASCVD), CVD risk levels, hospitalization, and mortality. Aspects of our work that set it apart from ongoing projects are (1) our ability to include complete data on inpatient hospital visits and emergency department visits when analyses are limited to South Carolina, (2) the advanced statistical modeling that enables us to account for multiple factors at multiple levels (i.e., patient, census tract, VAMC); and (3) the spatio-temporal aspects of the proposal which are critical given the spatio- temporal nature of the pandemic Next Steps/Implementation: We plan to create a dashboard in Power BI, a VA supported business intelligence tool, that allows users to display data in visual format allowing data to inform strategic decisions. Our dashboard will provide individualized information for each VAMC illustrating adjusted quarterly levels from 2017 through 2022 for (1) disrupted care metrics (i.e., CVD risk monitoring, primary care visits by mode), (2) levels of ASCVD, (3) ACC ASCVD risk levels (in those without ASCVD); and (4) mortality rates. Our use of area-level data and our focus on the VAMC level analyses will inform policy-level decision making during and following the pandemic. Lessons learned on the relationship between changes in care delivery, VAMC-level adaptability and chronic disease outcomes will inform post-pandemic care throughout the VA.