AIMS: Cardiovascular risk factors and comorbidities are highly prevalent among COVID-19 patients and are associated with worse outcomes. We therefore investigated if established cardiovascular risk assessment models could efficiently predict adverse outcomes in COVID-19. Furthermore, we aimed to generate novel risk scores including various cardiovascular parameters for prediction of outcomes in COVID-19.
METHODS AND RESULTS: A total of 441 consecutive patients diagnosed with SARS-CoV-2 infection were followed-up for 30 days after hospital admission for a combined endpoint, which consisted of first manifestation of either all-cause mortality (ACM) and/or venous/arterial thromboembolism, and secondary endpoints comprising ACM, mechanical ventilation and venous/arterial thromboembolism. Discrimination performance of DAPT, GRACE 2.0, PARIS-CTE, PREDICT-STABLE, CHA2-DS2-VASc, HAS-BLED, PARIS-MB, PRECISE-DAPT scores for selected endpoints was evaluated by ROC-analysis which showed that GRACE 2.0 score performed best in predicting combined endpoint and ACM. Novel risk assessment models including age, reduced right ventricular function, moderate/severe tricuspid regurgitation, arterial hypertension, D-dimer, C-reactive protein, procalcitonin, troponin I, lactate dehydrogenase and/or N-terminal prohormone of brain natriuretic peptide, were generated and could successfully predict combined endpoint (AUC=0.830), ACM (AUC=0.827), venous/arterial thromboembolism (AUC=0.774) and need for mechanical ventilation (AUC=0.862).
CONCLUSION: Novel risk assessment models including age, laboratory parameters, cardiovascular risk factors and echocardiographic parameters showed good discrimination performance for adverse outcomes in COVID-19. The newly developed scores outweighed discrimination performance of established cardiovascular risk assessment models.