Home - News RSS feed - Nizhny Novgorod and Belgian scientists have found a way to predict the severity of COVID-19

21 0902 ЦИР подсъёмы к статье 07

The scale of the COVID-19 pandemic has increased pressure on health care systems around the world. Doctors need fast and accurate methods for early clinical assessment of disease severity. This determines how and when care is provided, and how quickly health professionals make the decision to transfer a patient to an intensive care unit. Many scientists are carrying out research in this area, but so far there are no simple and reliable methods for predicting the development of Covid in clinical practice.

A team of scientists from Nizhny Novgorod and Belgium has developed a way of predicting the severity of coronavirus infection in the early stages of the disease. The study was published in Frontiers in Immunology , a leading journal in the field of clinical immunology. The project involved scientists from two Nizhny Novgorod universities, Lobachevsky University and Privolzhsky Medical Research University, as well as Ghent University (Belgium). They studied a group of 60 men and women aged 18-85 years who had been hospitalised with a confirmed diagnosis of COVID-19. They also used a control group of 17 healthy persons who were in contact with the infected persons but had no complaints and had a negative PCR test result.

Using artificial intelligence methods, the researchers built a clinical prediction model. Out of 50 biomarkers - various indicators of body condition - eight were selected. Some of the parameters, such as C-reactive protein and interleukin-6, characterise inflammation in the body. Other markers determine its physiological state, e.g. fibrinogen is related to blood clotting, while glucose level characterises the body's overall energy balance. Individually, these markers are often used in the context of many diseases, including COVID-19, but this set of eight indicators is specific to Covid. It reveals the key mechanisms for the development of complications and can ultimately predict, with 83% accuracy, whether a patient will develop a severe form of the disease.


Biomarkers associated with COVID-19 severity.

The group of severe cases is characterised by increased creatinine, glucose,

MIG, monocyte counts, fibrinogen, IL-6 and C-reactive protein, and decreased MDC.

Mikhail Ivanchenko, Vice-Rector for Research at Lobachevsky University: "This disease involves many systems of the human body, so it is very important to identify the key signs that determine the degree of severity of COVID-19. Data analysis and machine learning methods play an important role in helping to narrow down a large space of laboratory data to a few simple and interpretable indicators".

The authors of the study have developed a severity calculator for COVID-19 and an Android mobile app, available on the UNN Healthy Ageing Digital Personalised Medicine website. To use the calculator, the patient's blood test results for the eight indicators mentioned above should be entered into the app. The calculator will give the probability of severe COVID-19. The higher the probability in excess of 50%, the more careful attention the patient needs to be given.

The test uses simple parameters that can be measured in most clinical laboratories around the world. This means that the results of the study have the potential to reduce the burden on clinics and reduce mortality rates.