This article has been reviewed in accordance with Science X’s editorial processes and policies. The editors highlight the following attributes while ensuring content authenticity:
Secondary bacterial infections of the lungs (pneumonia) are very common in patients with COVID-19, affecting nearly half of those requiring ventilator support. By applying machine learning to medical record data, scientists at Northwestern University Feinberg School of Medicine found that unresolved secondary bacterial pneumonia was the leading cause of death in her COVID-19 patients. . It even exceeds the death rate from the virus infection itself.
Scientists also found evidence that COVID-19 does not trigger a “cytokine storm.”
This research recently Journal of Clinical Investigation.
“Our study highlights the importance of preventing, detecting, and aggressively treating secondary bacterial pneumonia in critically ill patients with severe pneumonia, including COVID-19.” and Pulmonary and Critical Care Physician at Northwestern Medicine.
Researchers found that nearly half of COVID-19 patients develop ventilator-associated secondary bacterial pneumonia.
“People who were cured of secondary pneumonia were more likely to survive, whereas those who were not cured of pneumonia were more likely to die,” Singer said. suggests that the mortality rate associated with pneumonia is relatively low, but that is offset by other things that occur during ICU stays, such as secondary bacterial pneumonia.
The findings also refute the cytokine storm hypothesis, Singer said.
“The term ‘cytokine storm’ refers to overwhelming inflammation that causes organ failure in the lungs, kidneys, brain and other organs,” Singer said. “If that is true, and if cytokine storms underlie the prolonged stays seen in COVID-19 patients, we would expect to see frequent transitions to a condition characterized by multiple organ failure. is not what I saw.”
This study analyzed 585 patients with severe pneumonia and respiratory failure in the intensive care unit (ICU) of Northwestern Memorial Hospital, of whom 190 had COVID-19. Scientists have developed a new machine learning approach called CarpeDiem. It groups similar patient days in his ICU into clinical status based on electronic health record data. This new approach, based on the concept of daily rounds by the ICU team, allowed us to ask how complications like bacterial pneumonia affected the course of the disease.
These patients or their surrogates consented to enroll in the Successful Clinical Responses to Pneumonia Treatment (SCRIPT) study, an observational trial to identify new biomarkers and treatments in patients with severe pneumonia. As part of SCRIPT, an expert panel of ICU physicians used state-of-the-art analysis of lung samples collected as part of clinical care to diagnose and determine the outcome of secondary pneumonia events.
“By applying machine learning and artificial intelligence to clinical data, we can develop better ways to treat diseases such as COVID-19 and assist ICU physicians managing these patients.” Feinberg Lung and Physician in Critical Care Medicine and Northwestern Medicine.
“Importance of bacterial co-infection in the lungs as a contributing factor to death in COVID-19 patients has been underestimated,” says Northwestern University’s Center for Pneumonia Therapeutic Systems Biology, leading a successful clinical response. Study co-author Richard Wunderink, Ph.D.
The next step in research is to use molecular data from research samples and integrate it with machine learning approaches to understand why some patients are cured of pneumonia and others are not. The researchers also hope to extend the technique to larger datasets, use the model to make predictions, and bring it back to the bedside to improve care for critically ill patients.
For more information:
Catherine A. Gao et al., Machine learning links unresolved secondary pneumonia and mortality in patients with severe pneumonia, including COVID-19. Journal of Clinical Investigation (2023). DOI: 10.1172/JCI170682
Journal of Clinical Investigation