Data collection
Patient data, including demographics, medical history, clinical manifestations, laboratory examinations, comorbidities, complications, and hospital stay were collected from the electronic medical records by a trained team of physicians. All the data were independently reviewed. Acute respiratory distress syndrome was defined according to the Berlin Definition 12. Acute myocardial injury was defined as blood level of cardiac biomarker cTnI increased above the 99th percentile upper reference limit13. Acute liver injury was defined as liver enzymes and/or bilirubin were more than twice the upper reference limit13. Acute kidney injury was identified according to the Kidney Disease: Improving Global Outcomes definition14. Coagulopathy was defined as a 3-s extension of prothrombin time or a 5-s extension of activated partial thromboplastin time 13.
Electrocardiogram (ECG) monitoring
A 12-lead ECG monitoring was continuously performed for 2 hours using a Holter monitor (CT-86, Baihui, Hangzhou City, China) in all enrolled participants. All patients lied on the bed quietly with no drink or food since 15 minutes before the monitoring until completion of the monitoring. Holter software was used to analyze the heart rate and arrhythmias. Sinus tachycardia was defined as an increase in sinus rhythm to >100 beats/min. Atrial tachycardia was defined as 3 or more consecutive atrial premature contraction. All ECG data were manually examined by two unsuspecting observers and arbitrated by electrophysiologists for any suspicious event classification.