The pathobiology of inflammation, thrombosis, and myocardial injury associated with severe acute respiratory syndrome coronavirus 2 (SARS-CoV2) may be assessed by circulating biomarkers. However, their relative prognostic importance has been incompletely described.

We analyzed data from patients hospitalized with COVID-19 from January 2020, to April 2021, at 122 US hospitals in the American Heart Association (AHA) COVID-19 cardiovascular (CV) disease registry. Patients with data for D-dimer, C-reactive protein (CRP), ferritin, natriuretic peptides [NP], or cardiac troponin (cTn) at admission were included. cTn quintiles were indexed to the assay-specific 99th percentile reference limits. Using multivariable logistic regression, we assessed the association between each biomarker by quintile [Q] and odds of in-hospital death and a cardiovascular and thrombotic composite outcome.

Of 32,636 registry patients, 26,424 (81%) had admission values for ≥1 of the key biomarkers, of which 4,527 (17%) had admission values for all 5 biomarkers. Each biomarker revealed a significant gradient for in-hospital mortality from Q1 to Q5: D-dimer 14% to 35%, CRP 11%-32%, ferritin 11% to 30%, cTn 13% to 43%, and NPs 7% to 35% (P_trend for each <.001). After adjustment for other biomarkers and clinical variables, Q5 for NPs (OR:4.67, 95% CI: 3.05-7.14) retained the greatest relative odds for death; cTn (OR:2.68, 95% CI: 2.00-3.59) and NPs (OR:7.14, 95% CI: 4.92-10.37) were associated with the greatest odds of the CV composite. Q5 for D-dimer was associated with the highest risk of thrombotic events (OR: 9.02, 95% CI: 5.36-15.18).

Among patients hospitalized with COVID-19, cTn and NPs identified patients at high risk for an in-hospital adverse cardiovascular outcome, while elevations in D-dimer identified patients at risk for thrombotic complications.