· No evidence of cytokine storm in critically ill patients with COVID-19
· Nearly half of patients with COVID-19 develop a secondary bacterial pneumonia
· Crucial to find and aggressively treat secondary bacterial pneumonia in ICU patients
Newswise — CHICAGO --- The city of Chicago is well-known for its unique architecture and rich history. However, in COVID-19 patients who needed mechanical ventilation, almost half suffered from a prevalent secondary bacterial pneumonia of the lungs, which machine learning applied to medical records by Northwestern University Feinberg School of Medicine researchers identified as a significant cause of death. In fact, its fatality rates may surpass those of the viral illness.
The researchers also discovered proof indicating that COVID-19 does not induce a "cytokine storm," which is commonly thought to be a cause of mortality.
"Our research emphasizes the significance of preventing, detecting, and treating secondary bacterial pneumonia aggressively in severely ill patients with pneumonia, including those who have contracted COVID-19," stated Dr. Benjamin Singer, a pulmonary and critical care physician at Northwestern Medicine and an associate professor of medicine at Northwestern University Feinberg School of Medicine, who is the senior author of the study.
The researchers discovered that almost 50% of individuals with COVID-19 experience a secondary bacterial pneumonia associated with mechanical ventilation.
"According to our findings, individuals who recovered from their secondary pneumonia had a higher probability of survival, whereas those whose pneumonia did not improve had a greater likelihood of mortality," Singer explained. "Our data indicated that the fatality rate linked to the virus is relatively low, but other factors that occur during an ICU stay, such as secondary bacterial pneumonia, negate this."
As the Lawrence Hicks Professor of Pulmonary Medicine at Feinberg, Singer added that the study's results refute the cytokine storm hypothesis as well.
Singer explained, "The concept of a 'cytokine storm' refers to an excessive inflammation that leads to organ failure in various parts of the body, such as the lungs, kidneys, and brain. If this were the case and the cytokine storm were responsible for the prolonged hospitalization observed in COVID-19 patients, we would anticipate frequent progression to conditions characterized by multiple organ failure. However, our findings do not support this."
In the research, 585 patients with severe pneumonia and respiratory failure who were admitted to the intensive care unit (ICU) at Northwestern Memorial Hospital were studied, out of which 190 had contracted COVID-19. To analyze the impact of complications such as bacterial pneumonia on the course of the disease, the scientists created a novel machine learning technique called CarpeDiem, which groups similar ICU patient-days into clinical states based on electronic health record data. This unique approach, which is similar to the concept of daily rounds conducted by the ICU team, allowed the researchers to investigate the influence of complications on the disease's progression.
These patients or their legal representatives provided consent to participate in the Successful Clinical Response to Pneumonia Therapy (SCRIPT) study, an observational trial that aims to identify new therapies and biomarkers for patients suffering from severe pneumonia. As part of SCRIPT, an experienced team of ICU physicians used advanced analysis techniques to examine lung samples that were collected during clinical care to diagnose and assess the outcomes of secondary pneumonia incidents.
Dr. Catherine Gao, a physician at Northwestern Medicine, study co-first author, and an instructor in pulmonary and critical care medicine at Feinberg, said, "The implementation of machine learning and artificial intelligence in clinical data analysis can aid in the development of more effective methods for treating diseases like COVID-19 and in assisting ICU physicians in managing patients."
Dr. Richard Wunderink, who heads the Successful Clinical Response in Pneumonia Therapy Systems Biology Center at Northwestern and is a study co-author, stated, "The significance of bacterial superinfection in the lungs as a factor leading to death in COVID-19 patients has been underestimated because most hospitals either do not test for it or only assess the presence or absence of bacterial superinfection and not the efficacy of treatment."
The study's next phase will involve the incorporation of molecular data from the samples collected during the study and integrating it with machine learning techniques to determine why some patients recover from pneumonia while others do not. The researchers also aim to expand this method to encompass larger datasets and utilize the model to make predictions that can be applied to patient care to enhance the treatment of critically ill patients.
In addition to Dr. Benjamin Singer and Dr. Catherine Gao, several other Northwestern University researchers and physicians contributed to the paper. These include Nikolay S. Markov, Thomas Stoeger, Anna E. Pawlowski, Mengjia Kang, Prasanth Nannapaneni, Rogan A. Grant, Chiagozie Pickens, James M. Walter, Jacqueline M. Kruser, Luke V. Rasmussen, Daniel Schneider, Justin Starren, Helen K. Donnelly, Alvaro Donayre, Yuan Luo, Scott Budinger, and Alexander Misharin.
Funding for the study was provided by the Simpson Querrey Lung Institute for Translational Sciences and a grant (U19AI135964) from the National Institute of Allergy and Infectious Diseases of the National Institutes of Health.
RSS