Changes in Cardiac and Pulmonary Hemodynamics as Predictor of Outcome in Hospitalized COVID-19 Patients

Background COVID-19 can lead to a bilateral pneumonia overwhelming the lungs causing dyspnea and respiratory distress. Up to 20% of the infected population is hospitalized and 5% is submitted to the intensive care unit (ICU). Up to 31% of patients in ICU develop sepsis and 61% develop ARDS with a deadly outcome at ICU of 38%. While sepsis typically causes diffuse vasodilation, the pulmonary vasculature resistance in ARDS is high. Although heart failure is per definition not the cause of ARDS, the resulting elevated pressures in the pulmonary circulation affect right and left heart function. Early detection in alterations of cardiac and pulmonary hemodynamics might prompt to actions to prevent ARDS.

Primary objective To evaluate cardiac and pulmonary hemodynamic changes over time as predictor of disease progression and outcome in COVID-19 patients admitted to ICU.

Secondary objective

• Analysis of prognostic factors based on the data at initial presentation
• Performing a trajectory analysis of the time course during ICU stay to determine what leads to optimal outcome - gain insight in the pathophysiology of the cardio-pulmonary evolution of COVID-19 pts
• Feasibility study for the creation of an individualized expected data-trajectory for new cases and continuously updating its visualization in relation to the expected trajectory related to an improved outcome
• Evaluate how Machine Learning, based on manifold learning for quantifying information similarity and its temporal evolution, is able to predict outcome using rich data in a limited number of patients Primary Endpoint

Occurrence of a major event predefined as either: