PROJECT SUMMARY/ABSTRACT
There is mounting concern that patients hospitalized with COVID-19 experience unexpectedly high rates of
cardiac and vascular events. Identifying which patients are at highest risk for COVID-19-related cardiovascular
events and delineating how these events affect short- and long-term outcomes may help support individualized
patient care, illuminate underlying pathophysiologic mechanisms, and accelerate the development of effective
therapies. However, little is known about how multi-dimensional risk factors, including prior medical conditions,
socioeconomic indicators, and circulating levels of biomarkers affect patient outcomes. Building on our
team's expertise in data linkage, prediction modeling, and biomarker discovery, we will create a unique
and powerful linked data resource to characterize the biological, clinical, health system, and
socioeconomic risk factors for the development of cardiovascular sequelae of COVID-19 and examine
their impact on health outcomes. To create this data resource, we have partnered with the American Heart
Association, whose COVID-19 Cardiovascular Disease Registry is actively capturing high-quality, standardized
information on all adults hospitalized with confirmed SARS-CoV-2 infection at >100 U.S. sites spanning 30
states. We will link this registry to comprehensive health care claims, a national socioeconomic deprivation
index, and detailed health care system information. In Aim 1, we will apply traditional and machine learning
approaches to the linked multicenter registry in order to identify the clinical, health system, and socioeconomic
factors that predict in-hospital major adverse cardiovascular events (MACE) among COVID-19 patients. In Aim
2, we will characterize long-term MACE (i.e., at 1 and 2 years after discharge from the index COVID-19
hospitalization) among older adults in a large multicenter registry linked with longitudinal Medicare claims, and
identify the clinical, health system, and socioeconomic factors that predict their occurrence. Based on this
work, we will create clinically implementable risk scores which will estimate, at the time of admission for and
discharge from an index COVID-19 hospitalization, a patient's risk of developing a major cardiovascular event.
In Aim 3, we evaluate the proteomic profiles of a subset of patients in the linked registry with biobanked serial
blood samples, and identify biochemical markers that predict the occurrence of MACE, both during index
hospitalization for COVID-19 and after discharge. This research will advance our collective understanding of
the biological, clinical, and socioeconomic predictors of COVID-19-related cardiovascular morbidity and
mortality. By identifying patients at greatest at risk of cardiovascular events, our work will help frontline
clinicians better individualize clinical management strategies and health systems improve care delivery during
future waves of the pandemic.
