News

Press Release: Blood test shows whether newborns with fever have bacterial or viral infection

Paris, France: A test that compares the levels of two molecules in the blood is able to accurately distinguish between bacterial and viral infections in newborn babies with fever in a small study presented at the European Emergency Medicine Congress today (Sunday) [1].

A fever in a baby under three months of age is considered a medical emergency, because in very young babies with undeveloped immune systems, an infection can quickly become life threatening.

However, in the emergency department, staff need to assess babies with fever to distinguish the most serious bacterial infections that require antibiotics from viral infection that typically pass with monitoring and care.

The new research was presented by Dr Tommaso Bellini, a paediatric emergency physician at the IRCCS Istituto Giannina Gaslini in Genoa, Italy. He said: “In very young babies with fever, distinguishing serious bacterial infection from self-limiting viral illness is a major diagnostic dilemma. Our standard biomarkers can help show when there might be a bacterial infection, but they also create ‘false positives’, which drives overtreatment. This can mean unnecessary lumbar punctures, antibiotics, and hospital admissions for infants who turn out to have only a virus.”

The research involved 60 babies who were brought to the paediatric emergency department of IRCCS Istituto Giannina Gaslini. Some were being assessed and treated for fever, while others were being assessed and treated for other, non-infectious conditions.

Dr Bellini and his colleagues collected blood samples from all the babies. They tested the blood samples with flow cytometry – an advanced technique that allows researchers to study the characteristics of individual cells.

They used this technique to quantify the levels of two molecules that are known to increase in response to infections. One of these, called CD64, is found on the surface of neutrophils, a type of white blood cell, and is thought to increase in response to bacterial infection. The other, called CD169, is found on the surface of another type of white blood cell, called a monocyte, and is thought to increase in response to a viral infection.

The researchers combined measurements of both these molecules into a ratio for each baby. A ratio is a simple way of comparing the amount of one thing with the amount of another. They found that this ratio could accurately distinguish between babies with no infection (average ratio score of 1.10), babies who were ultimately confirmed to have a viral infection (average 0.33) and babies confirmed to have a bacterial infection (average 16.22).

The test had a ‘sensitivity’ of 100%, meaning it correctly picked up all the cases of bacterial infection and a ‘specificity’ of 96.6%, meaning that the rate of ‘false positives’ (where the test suggested a baby had a bacterial infection when they did not) was extremely low.

Dr Bellini said: “These two marker molecules show opposite signatures – during a viral infection one is high while the other is low, and during a bacterial infection the reverse is true. Combining them into a single ratio separated bacterial from non-bacterial infection with very high accuracy in our initial cohort of 60 babies.

“This assessment is currently run on flow cytometry, which needs specialised equipment and trained staff. It is not yet a bedside test, and not widely available outside of specialist labs. Translating it into a fast, low-cost, point-of-care format is our next research goal.”

The research is continuing and will ultimately include data on 75 babies. Dr Bellini and his colleagues are also beginning a larger study called DISCERN (Discriminating Infection via Surface CD-marker Expression Ratios iN infancy). This is a multicentre, prospective study assessing the accuracy of the test in 300 to 500 babies.

Dr Maša Sorić is a member of the EUSEM abstract selection committee. She is head of the emergency department at University Hospital Merkur, Zagreb, Croatia, and was not involved with the research. She said: “For parents, taking a very young baby with a fever into the emergency department can be a very worrying time. We have to treat every case as potentially serious until we know otherwise. At the moment, this can mean invasive tests, treatment with antibiotics and spending a long time in hospital. This is a small study, but the findings are very positive. If these results can be reinforced in a larger trial, they could mean that we have a much better way of knowing which babies probably have a bacterial infection that needs antibiotics, and which babies probably do not. This could ultimately improve treatment and reduce the need for unnecessary treatment.

“There is also a broader issue with bacteria becoming resistant to antibiotics, partly as a result of over-use of these treatments. If we can ensure we only use antibiotics when they are needed, we may also reduce the growth of antibiotic resistance, helping to maintain the effectiveness of these vital medicines.”

(ends)

[1] Abstract no: OA091, “CD64 and CD169 as Translational Biomarkers for Early Bacterial-viral Discrimination in Febrile Infants ≤90 Days” by Tommaso Bellini, Oral abstract session: Paediatric abstracts session, Sunday 27 September, 11.00-12.30 hrs CEST, Room 142.

Funding: No external funding. The study was conducted as part of Dr Bellinin’s doctoral research at the University of Genova (DINOGMI).