The uncontrolled hyper-inflammation in critically ill patients with COVID-19 might be associated with a dysfunction of the cardiovascular regulatory mechanisms. We aim to estimate the involvement of cardiovascular control in limiting the risk of mortality in COVID-19 patients. From electrocardiogram and invasive arterial pressure recording, we extracted heart period (HP) and systolic arterial pressure (SAP) at rest in supine condition (REST) and during an orthostatic challenge, namely the modified head-up tilt (MHUT), in 18 COVID-19 patients (age: 62 ± 10 yrs, 15 men) admitted in intensive care unit (ICU) for pneumonia. The patients were distinguished in two groups according to the outcome: survivors (SURVs, 8 patients; age: 58±9 yrs, 8 males) or non-survivors (noSURVs, 10 patients; age: 65±8 yrs, 7 males). We assessed the asynchrony between HP and SAP via a model-free nonlinear marker in the information domain, i.e. cross-sample entropy (CSampEn). Neither demographic indexes nor time domain markers could separate the two groups and this result held regardless of the experimental condition. Conversely, noSURVs had a larger HP-SAP asynchrony when compared to SURVs. Specifically, CSampEn increased significantly in noSURVs (1.76 ± 0.42) compared to SURVs (1.34 ± 0.42) in response to MHUT, showing a missed ability of noSURVs to activate residual cardiovascular regulatory mechanisms. We hypnotize that the differences between the groups might be attributed to the atypical dysregulated immune response triggered by COVID-19 triggering a derangement in the integrative autonomic center activity in noSURVs. We conclude that measures of the derangement of the cardiovascular control might be helpful to stratify the risk of mortality in COVID-19 critically ill patients. CSampEn and MHUT could be utilized as a part of the monitoring of COVID-19 to assess and prevent the hyper-inflammation.