Severe COVID-19 With Myocardial Injury Did Not Translate Into Frequent Cardiovascular Death

Severe COVID-19 can injure the heart, but this multicenter cohort suggests the downstream clinical story is often driven elsewhere. Among 1,328 consecutive patients admitted with severe COVID-19 and biochemical or echocardiographic evidence of myocardial injury, individual cardiovascular complications remained uncommon, and cardiovascular death was not a dominant pathway to mortality.

The study drew from 16 centers in the CARDIO-COVID consortium and focused on patients sick enough to require intensive care-level attention. ICU mortality reached 29.4%, a sobering figure in its own right. Yet the factors most tightly linked to death were not isolated cardiac events. Age, SOFA Cardiovascular score, elevated troponin, intubation, high vasoactive-inotropic score, high PEEP, and high FiO2 carried the strongest association with mortality, pointing toward a physiology of global critical illness rather than a single-organ cardiac catastrophe.

Cardiac biomarkers still mattered.

Elevated troponin and myocardial dysfunction remained prognostic signals, but the pattern looked more like strain than direct viral cardiotoxicity. In critically ill patients, myocardial injury can reflect supply-demand mismatch, inflammatory stress, vasopressor exposure, hypoxemia, and septic physiology. This study gives that interpretation clinical weight: cardiac injury tracked with severity, while the event types usually feared at the bedside did not rise to dominate outcomes.

Rare Events, Cumulative Burden

The investigators examined a broad set of cardiovascular complications, including myocardial infarction, stroke, pulmonary embolism, myocarditis, ventricular tachycardia or ventricular fibrillation, and new-onset heart failure. Each individual complication occurred in fewer than 5% of patients. Viewed separately, none of these events appeared common enough to explain the mortality signal on its own.

Viewed together, the burden was more substantial. Any cardiovascular complication occurred in 17.6% of the cohort, a reminder that severe COVID-19 can still destabilize the heart and circulation even when no single diagnosis dominates the chart. The critical nuance is that this cumulative burden did not show a significant association with mortality. In other words, the cardiac events were real, but they were not the main drivers of death in this population.

That matters at the bedside because severe COVID-19 often prompts a search for one decisive cardiology explanation when the patient deteriorates. These data argue for a broader frame. Myocardial injury in this setting may function as a marker of systemic failure, not necessarily as the proximate cause of death. The distinction changes how clinicians interpret troponin elevation, ventricular dysfunction, or hemodynamic instability in the intensive care unit.

Septic Shock Physiology Appeared to Sit Behind the Sickest Hemodynamics

Patients with high vasoactive-inotropic scores and high SOFA Cardiovascular scores had a significantly higher incidence of secondary bacterial pneumonias. That pattern points toward septic shock physiology as a major part of the hemodynamic collapse in the sickest patients. Respiratory failure and septic shock emerged as the primary causes of death in investigator-adjudicated data, reinforcing the idea that the terminal pathway usually ran through multiorgan dysfunction rather than a discrete cardiovascular event.

This is clinically important because myocardial injury in critical illness can be misleading if interpreted in isolation. A rising troponin may reflect the cumulative burden of hypoxemia, vasopressor exposure, afterload stress, and inflammatory myocardial depression. A patient who needs higher vasoactive support and escalating ventilatory settings may be showing the physiology of systemic shock, with the heart acting as an injured organ within that broader failure pattern.

What the Machine Learning Model Added to the Picture

A light-GBM machine learning model was used to identify variables associated with mortality. The model did not replace bedside reasoning, but it helped rank the variables most closely aligned with death in this cohort. Age, SOFA Cardiovascular score, elevated troponin, intubation, high VIS, high PEEP, and high FiO2 emerged as the strongest signals.

The model’s output is consistent with the clinical reality of severe COVID-19 in critical care units: mortality risk is shaped by respiratory failure, vasopressor dependence, and the extent of organ dysfunction. Myocardial injury appears embedded within that syndrome rather than standing apart as a separate lethal cardiac process. In practical terms, an elevated troponin in these patients should prompt careful hemodynamic assessment, but it should not automatically be interpreted as evidence that a primary cardiovascular complication is driving the outcome.

Interpreting Troponin in Severe COVID-19

For clinicians, the main takeaway is not that the heart is irrelevant. It is that the heart injury seen in severe COVID-19 may often be a biomarker of stress rather than the endpoint itself. Troponin elevation, myocardial dysfunction, and even overt cardiovascular complications deserve attention, yet their presence does not appear to explain most deaths in this cohort.

This has implications for escalation decisions, prognostic counseling, and the framing of cardiac workup in critically ill patients. When respiratory failure and septic shock dominate the final trajectory, a narrow focus on the myocardium can miss the broader physiology. The study supports a more integrated view of severe COVID-19, in which the heart is vulnerable but not usually the principal terminal organ.

Limits of the Signal and the Questions Left Open

The cohort was large and multicenter, but the study still describes a hospitalized, critically ill population during severe COVID-19 rather than the full spectrum of SARS-CoV-2 infection. The findings also rely on observational associations, not randomized comparisons. Even so, the central pattern is hard to ignore: cardiovascular complications were present, yet they were rarely the direct story behind death.

The study closes a familiar loop from the pandemic’s early months. Myocardial injury in severe COVID-19 is clinically meaningful, but in this cohort it behaved more like a marker of multisystem organ dysfunction than proof of direct myocardial destruction by SARS-CoV-2. Future work will need to separate transient myocardial strain from structural cardiac injury with greater precision, especially in patients whose hemodynamics are already being shaped by ventilation, vasopressors, and sepsis.


Reporting by Dr. Aishwarya N., MBBS for Medicine World Foundation, with additional data provided by Nature Reviews Clinical Oncology.