Site-level variability in the processes of care and outcomes over time among patients with COVID-19 and myocardial injury: Insights from the American Heart Association's COVID-19 Cardiovascular Disease Registry

Background: Elevated cardiac troponin (cTn) levels in patients with COVID-19 has been associated with worse outcomes. Guidelines on best practices of those patients remain uncertain.Methods: We included patients with COVID-19 and cTn above the assay-specific upper limit of normal (ULN) enrolled in the American Heart Association's COVID-19 registry between March 2020-January 2021. Site-level variability in invasive coronary angiography, LVEF assessment, ICU utilization, and inpatient mortality were determined by calculating adjusted median odds ratio (MOR) using hierarchical logistic regression models. Temporal trends were assessed with Cochran-Armitage trend test.Results: Among 32,636 patients, we included 6234 (19.4 %) with cTn above ULN (age 68.7 +/- 16.0 years, 56.5 % male, 51.5 % Caucasian), of whom 1365 (21.6 %) had >= 5-fold elevations. Across 55 sites, the median rate of invasive coronary angiography was 0.1 % with adjusted MOR 1.5(1.0,2.3), median LVEF assessment was 25.5 %, MOR 3.0(2.2,3.9), ICU utilization was 41.7 %, MOR 2.2(1.8,2.6), and mortality was 20.9 %, MOR 1.7(1.5,2.0). Over time, we noted a significant increase in invasive coronary angiography (p-trend = 0.001), and LVEF assessment (p-trend<0.001), and reduction in mortality (p-trend<0.001), without significant change in ICU admissions (p-trend = 0.08). Similar variability and temporal trends were seen among patients with >= 5-fold cTn elevation.Conclusions: The use of invasive coronary angiography among patients with COVID-19 and myocardial injury was very low during the early pandemic. We found moderate institutional variability in processes of care with an uptrend in invasive catheterization and LVEF assessment, and downtrend in mortality. Comparative effectiveness studies are needed to examine whether variability in care is associated with differences in outcomes.