Long-term follow-up of recovered patients with COVID-19

M Cortinovis, N Perico, G Remuzzi - The Lancet, 2021 - thelancet.com
M Cortinovis, N Perico, G Remuzzi
The Lancet, 2021thelancet.com
colleagues must be congratulated for developing a comprehensive and well designed score
for integrating thrombotic and bleeding risks in patients with ACS. Compared with previous
scoring approaches (appendix), the PRAISE score is based on one of the largest study
populations and shows superior performance when externally validated. The variables
included in the score are easily accessible from patient discharge information. Therefore, the
score is simple, intuitive, and easy to implement in everyday clinical practice, also thanks to …
colleagues must be congratulated for developing a comprehensive and well designed score for integrating thrombotic and bleeding risks in patients with ACS. Compared with previous scoring approaches (appendix), the PRAISE score is based on one of the largest study populations and shows superior performance when externally validated. The variables included in the score are easily accessible from patient discharge information. Therefore, the score is simple, intuitive, and easy to implement in everyday clinical practice, also thanks to the PRAISE score calculator available online. Of note, previous scores have often been derived from cohorts of patients admitted for both stable coronary artery disease and ACS. 4–6 As these two clinical entities are likely to be associated with different risk profiles, the training and validation of the PRAISE score on cohorts of only patients with ACS should be considered as an additional strength. Although the ability of machine learning based models to overcome limitations of traditional regressionbased risk prediction systems remains a subject of debate, D’Ascenzo and colleagues’ study7 does not allow conclusions to be drawn in this regard, since a comparison with conventional statistical models is lacking. 8–11
The study of D’Ascenzo and colleagues surely represents a considerable step forwards in enhancing the risk stratification of patients with ACS. Nevertheless, some caution must be applied when analysing the results. There might be confounding by indication due to the observational design of the derivation cohorts. Also, the prediction of events based on clinical features (eg, age, haemoglobin concentration, eGFR, and LVEF) that correlate with both outcomes (ie, ischaemia and bleeding events) might mean that the model does not always lead to therapeutic decisions that ultimately improve prognosis. Whereas the PRAISE models were
thelancet.com