Blood tests help predict Covid-19 patients most likely to die 10 days in advance with 90% accuracy, says study

These tests can flag patients whose conditions might deteriorate more than ten days in advance with a 90% accuracy, allowing doctors to treat them effectively


                            Blood tests help predict Covid-19 patients most likely to die 10 days in advance with 90% accuracy, says study
(Getty Images)

A new study suggests that simple blood tests may predict which Covid-19 patients are likely to die. These tests can flag patients whose conditions might deteriorate more than ten days in advance with a 90% accuracy, allowing doctors to treat them effectively.

Doctors may be able to identify high-risk individuals by measuring the levels of three indicators or markers of death in the blood. "It provides a simple and intuitive clinical test to precisely and quickly quantifiy the risk of death," the authors wrote in their study.  By detecting high-risk patients, these tests may help reduce the Covid-19 death rate, they added.

Data from Wuhan -- the former epicenter of the outbreak -- suggests that about 14% to 19% of COVID-19 develop a severe illness. The death rate among critical patients stands at  61.5% - and this increasing sharply with age and for patients with underlying health conditions.

In this study, the team analyzed blood samples of severely sick patients in order to detect common patterns or markers. These patterns, in turn, may detect death risk. So they studied 485 patients admitted at Tongji Hospital in Wuhan, China, between January 10 and February 18. The study was a collaboration between the hospital and Huazhong University of Science and Technology, Wuhan.

They looked at their symptoms, collected their blood samples, and performed a series of tests, including those on the liver, kidney and blood clotting, and the like. Towards the end of the study, they made a note of death rates. Of the 375 patients who were analyzed further, 201 recovered from Covid-19 and174 died.

(Getty Images)

They then deployed artificial intelligence to analyze patient data and identify common markers in those who are likely to die. The analysis showed that high-risk patients had abnormal levels of three body components: lactic dehydrogenase (LDH), Lymphopenia, and high-sensitivity C-reactive proteins (hs-CRP).

LDH is an enzyme, which indicates damage to the lungs and is also seen during pneumonia. Patients who are likely to die tend to have higher levels of this enzyme in the blood. These patients may also have lower levels of cells that defend invading microorganisms: a type of white blood cells called lymphocytes. The condition is called lymphopenia and is commonly seen in Covid-19 patients. "Lymphopenia is a common feature in patients with Covid-19 and might be a critical factor associated with disease severity and mortality," they wrote in their study.

Finally, high-risk patients may also have high levels of a protein involved in the body's response to an infection, hs-CRP, according to the study.  Heightened response from the immune system results in clogging of the lungs, leading to respiratory distress.

"The three key features, LDH, lymphocytes, and hs-CRP, can be easily collected in any hospital. In crowded hospitals, and with shortages of medical resources, this simple model can help to quickly prioritize patients, especially during a pandemic when limited healthcare resources have to be allocated," the authors wrote in their study.

The experts suggest that their prediction will hold better value with more patient data. "This study has room for further improvement, which is left for future work," they added.

The study is published in Nature machine intelligence.

For more information and statistics on the coronavirus pandemic, click on the Newsbreak tracker here