A new model to predict death rates among patients with COVID-19 in the ICU

What is it about?

The COVID-19 pandemic infected millions of people around the world. Some patients developed additional underlying diseases as a result and had to be hospitalized. Reports show that one-third of hospitalized patients with COVID-19 ultimately needed intensive care unit (ICU) admission. In addition, patients with COVID-19 who were in the ICU had a higher death rate than those with other lung diseases. A new study tried to find out which factors lead to higher death rates among ICU-admitted patients with COVID-19. First, they studied the medical history and chest scans of such patients. Next, they designed a model to predict the death rate among these patients based on clinical and radiologic (i.e., imaging) factors. They found that only 41 out of the total of 121 ICU patients survived. The survivors were almost 8.3 years younger than those who did not survive. Moreover, non survivors had lower blood oxygen levels and blood pressures than survivors. Hypertension and diabetes were the most common underlying diseases in ICU patients, but the former was more prevalent in non survivors. Imaging analyses showed that non survivors had a higher proportion of cardiomegaly, an enlargement of the heart. They also had more fluid between the lung and chest tissues and around the heart. The authors found that their new model could predict the death rates among these ICU patients with a 75.5% accuracy.

Featured Image

Photo by Mufid Majnun on Unsplash

Photo by Mufid Majnun on Unsplash

Why is it important?

The study found that a combination of underlying diseases, vital signs, and imaging factors could predict the death rates in ICU patients with COVID-19. Using the authors' new model, doctors will be able to identify patients with COVID-19 who are at a high risk of death and provide more resources for them. In addition, they will be able to provide these patients with early and modified treatments. KEY TAKEAWAY: A model that uses a combination of underlying diseases, vital signs, and imaging factors can predict the death rates among ICU-patients with COVID-19 with high precision.